International Journal of Innovative Computing and Applications (60 papers in press)
Regular Issues
Energy Sensing Streaming Media Data Transmission Protocol Based on Implicit Markov Algorithm in WSNs 
by Guozhong Li
Abstract: Streaming media transmission protocols can be divided into traditional streaming media push protocol and streaming media pull protocol. Traditional streaming media push protocols such as RTP, whose server determines the channel state according to the RTP feedback from the client, and then decides to send data packets suitable for the current channel state. The pull protocol of streaming media sends data packets according to the content of the client when the transmission rate meets the requirement on the contrary. Streaming media pull protocol greatly reduces the complexity of servers in streaming media transmission technology, it can also support the application of different adaptive algorithms in the transmission process compared with the traditional universal server transmission algorithm mechanism. Therefore, this kind of pull protocol cannot only improve the quality of service of streaming media transmission, but also meet the requirements of different channel states and different clients.
Keywords: WSNs; Data aging; Mesh area; Energy aware; Data transmission protocol.
Cardiac Arrhythmia Classification using Sequential feature selection and Decision tree classifier method
by Durga S, Esther Daniel, Deepa Kanmani S, Jinsa Mary Philip
Abstract: Cardiac arrhythmia is referred to as a condition in which the hearts normal functionality is restricted resulting in cardiovascular diseases. Effective and well-timed monitoring is very much essential to save human life. During the past few years, keeping track of when and how arrhythmias occur has gained a lot of significance as it leads way towards many life-threatening issues like stroke, sudden cardiac arrest and also heart failure. This paper provides an evaluation of various classification algorithms based on feature selection techniques that improve the performance of the cardio monitoring system. The pre-eminent features are sorted out using feature selection methods. The feature selection methods enable to decide the features that can contribute to improve the performance. The paper also gives efficient combinations of distinct classification algorithms along with feature selections which improves the accuracy. Few popular machine learning algorithms namely Na
Keywords: Cardiac arrhythmia; Classification algorithms; Accuracy; Decision tree classifier; Sequential feature selection.
Analyzing Smart Power Grid against different Cyber Attacks on SCADA system
by Tafseer Akhtar, Brij Gupta
Abstract: The integration of information and communications technology (ICT) with traditional power grid is transforming the power grid into a more reliable and smarter cyber physical system (CPS). Supervisory Control and Data Acquisition (SCADA) system is generally used to implement the cyber layer for power grid. This system is responsible for real time monitoring and control of power grid and power delivery network. SCADA system with the help of other devices create smart power grid environment, but it also makes power grid vulnerable against different cyber attacks as it needs connection with different kinds of open networks. Smart power grids are extremely critical infrastructure and attack of any kind on it, may affect socio-economical condition of the region. In this paper we present a co-simulation environment for smart power grid using OMNeT++ and openDSS. And different cyber attack scenarios are also taken as a part of the developed framework. Co-simulation involves integration between network simulator and power generation simulator as both works as cyber layer and physical layer respectively. The main contribution of the paper is behavioral analysis of the smart power grid environment while attack performed on cyber layer of the grid and store the response of power grid during these attacks. The findings of the proposed work are vulnerabilities realization of smart power grid, as during cyber attack grid is unstable and power generation is significantly dropped. The Proposed framework is developed by using open source simulators only.
Keywords: Co-simulation; Smart power grid; SCADA system; Cyber attack scenarios.
Research on Intelligent City Parking Guidance Method Based on Ant Colony Algorithm
by Aifeng Chen
Abstract: In order to get the most satisfactory parking space at the fastest speed, an intelligent urban parking guidance method based on ant colony algorithm is proposed. The main factors affecting the selection of parking spaces in parking lots are analyzed, including walking distance, driving distance, walking time, driving time and so on. Each factor is set as multiple attributes of berth, and the optimal berth selection model of smart city is established. Ant colony algorithm is used to solve the model, obtain the optimal parking space, and realize intelligent guidance of intelligent city parking. The simulation results show that the proposed method is feasible and effective.
Keywords: Ant Colony Algorithm; Intelligent City; Parking; Intelligent Guidance.
Business Intelligence & Data Analytics Framework: Case Study of Humanitarian Organizations Refugees Registration System
by Amal Ballout, Mays Al Hasan, Salam Fraihat, Ammar Elhassan
Abstract: Business Intelligence and Analytics has gained prominent focus among organizations with information systems that collect and process vast amounts of data. Voluminous, unprocessed data does not lend itself to offering useful insights for businesses, especially with basic statistical methods and traditional reporting techniques. In this work, we design a Business Intelligence and Data Analytics Framework for Refugee Registration System serving over six million refugees to collect, collate and filter demographic data. The proposed reporting mechanism leverages the power of interactive dashboards to offer informative and intuitive reports and visualizations that are accessible and interpretable by stakeholders.
Keywords: Business Intelligence; Analytics; Demographics; Reporting; Dashboards.
Research on Coordinated Control Method of Urban Traffic based on Neural Network
by Lede Niu, Mei Pan
Abstract: Due to the complexity of data in urban traffic coordination and control, there is a lack of relevance between data nodes of existing coordination methods, which leads to urban traffic holdup and congestion.A coordinated urban traffic control method based on neural network is proposed.The traffic flow prediction model is constructed to predict and calculate the urban traffic flow, green signal ratio, phase and period.The coordinated control method based on neural network is used to fine-tune the traffic signal, green signal ratio, phase and cycle, so as to improve the traffic capacity at the intersection during the peak period, and finally realize the coordinated control of urban traffic.The simulation results show that the proposed method can effectively alleviate the traffic capacity at intersections, and the time required for the method to run the whole process is less, which indicates that the proposed method is effective and reliable.
Keywords: Neural network; Urban traffic control; Signal light coordination; simulation.
An Efficient Image Compression Using Pixel Filter for Social media applications
by Ramesh Makala, Ranganath Ponnaboyina, Gowtham Mamidisetti
Abstract: Image data transfer has increased rigorously in present times in social networking sites, mobile apps and live streaming video applications. This phenomenon puts enormous effect on internet bandwidth and speed of image transfer and loading. Image compression deals with this issue by reducing image sizes while maintaining quality aspects. In general filters are applied on a block and deal with quality enhancements. We introduce a new filter which is applied on each pixel and compresses it. Proposed method classifies pixels into different buckets based on filter. Inverse process tries to restore pixel values back from buckets. With experimental results, we show compression and quality aspects variations based on filter selection.
Keywords: Color Map; Image Compression; Filters; Quality Enhancement.
PERFORMANCE ANALYSIS OF RECONSTRUCTED IMAGE USING COLOR DEMOSAICING ALGORITHM FOR DIGITAL IMAGES
by Allwin Devaraj
Abstract: Demosaicing algorithm is emerging as one of the most promising techniques in Digital Image Processing because of providing high-quality image enhancement with low power and high speed. Edge detector, weighting block of anisotropic type, and a compensator with filters are the components of the proposed algorithm. Spatial filters and Laplacian methods are used in the filter-based compensation methodology. These improve edge detection, and the blurring effect is reduced. Hardware sharing reduces the cost of the system. Portable and fast systems are required to produce the test results in a short span. For the above-mentioned purpose, the system is designed in VLSI technology. Matlab and Xilinx tools are used for the system design. A novel strategy to address shading contrasts in multiview video successions is proposed, and it utilizes a thick coordinating based worldwide enhancement structure. The proposed vitality work guarantees to safeguard nearby structures by controlling deviations from the first picture angles. When compared with the existing works 8% and 90.6% of power reduction is observed in the proposed work. It also improves the average color signal-to-noise ratio quality by more than 1.6db.
Keywords: Very Large Scale Integration; Peak Signal-to-Noise Ratio; Field Programmable Gate Array; Green Interpolator.
Modified Adaptive Inertia Weight Particle Swarm Optimization for Data Clustering
by Vikash Yadav, Indresh Kumar Gupta
Abstract: Data clustering is widely applied in many real world domain including marketing, anthropology, medical science, engineering, economics, and others. It concern with partition of unlabelled dataset objects into clusters (groups) based on similarity measure. Partitioning approach of dataset objects must follow that, intra-cluster distances are smaller and inter-cluster distances are larger. In current work particle swarm optimization (PSO) is employ for clustering. Sometime PSO may stuck into local optima, to overcome the PSO algorithms trapping in local optima a modified adaptive inertia weight particle swarm optimization (MAIWPSO) is developed for data clustering based on fitness value of particles. K-means, PSO and MAIWPSO for clustering have been simulated on six standard dataset namely iris, thyroid, heart, breast cancer, crude oil and pima. Simulation results confirm MAIWPSO is better approach for clustering against K-means and PSO.
Keywords: Data clustering; K-means clustering; Particle swarm optimization; fitness function.
Instance-based novelty detection in passive sonar signals
by Victor Hugo Da Silva Muniz, João Baptista De Oliveira e Souza Filho, Eduardo Sperle Honorato
Abstract: In submarines, sonar operators have the main task of identifying potential threats, named as contacts in the military jargon. The principal tool exploited when dealing with such situations is the passive sonar system. Automatic contact classification models may relieve the huge sonar operator workload but require mechanisms capable of identifying any contact not considered during system development. This paper discusses the development of a hierarchical instance-based detector of unknown contact classes for passive sonar signals, focusing on practical strategies for its hyperparameter tuning and performance assessment. Experimental data exploited in system evaluation comprises the acoustic noise irradiated by 28 ships belonging to 8 classes. These ships were submitted to different operational conditions in several runs conducted in an acoustic range. The kNN algorithm has performed best, achieving a novelty detection rate of 78.0%, associated with an average known case identification rate of 95.0%, considering a five unknown class evaluation scenario.
Keywords: passive sonar; novelty detection; automatic classification systems; instance learning; kNN; k-nearest neighbours.
Constrained Neural Classifier Training Method for Flaw Detection in Industrial Pipes Using Particle Swarm Optimization
by Gilvan Silva, Edmar Souza, Eduardo Simas, Paulo Farias, Maria Albuquerque, Ivan Silva, Cláudia Farias
Abstract: A novel method for constrained training of multi-class artificial neural network classifiers is proposed in this work. The traditional training procedure is usually based on mean square error minimization and thus, all classes of interest are considered as having the same relevance for system performance. This is not always the case for real-world applications in which the class relevance may be unbalanced. In this paper, cost functions designed to introduce classification performance constraints for specific classes are presented and particle swarm optimization is used as global optimization method. The proposed method is applied to a non-destructive evaluation decision support problem using pulsed eddy currents signals. Experimental results obtained from thermally insulated industrial pipes indicate the efficiency of the proposed method in comparison to neural networks trained from the traditional back-propagation algorithm.
Keywords: Artificial Neural Networks; Particle Swarm Optimization; Pulsed-Eddy Current Evaluation; Signal Processing.
Convolutional neural networks applied in the detection of pneumonia by x-ray images
by Luan Silva, Leandro Araújo, Victor Souza, Raimundo Neto, Adam Santos
Abstract: According to the World Health Organization (WHO), pneumonia kills about 2 million children under the age of 5 and is constantly estimated as the leading cause of child mortality, killing more children than AIDS, malaria, and measles together. The application of deep learning techniques for medical image classification has grown considerably in recent years. This research presents three implementations of convolutional neural networks (CNNs): ResNet50, VGG-16, and InceptionV3. These CNNs are applied to solve the classification problem of medical radiographs from people with pneumonia, as a manner to assist in the disease diagnosis. The three architectures used in this research obtained satisfactory results. The ResNet50 outperformed InceptionV3 and VGG-16, achieving the highest percentage of training and testing precision, as well as superior recall and f1-score. For the normal class, the f1-score related to ResNet50 was 88.42%, compared to 81.54% for InceptionV3 and 81.42% for VGG-16. For the pneumonia class, this metric was 95.10% against 92.82% for InceptionV3 and 92.54% for VGG-16.
Keywords: Deep learning; Pattern Recognition; Convolutional neural networks; Pneumonia; X-ray.
An EfficientOppositional Crow Search Optimization-based Deep Neural Network Classifier for Chronic Kidney Disease Identification
by Pramila Arulanthu, Eswaran Perumal
Abstract: The Internet of Things (IoT) enables the total gathering of patient data and patient information that can incite logically exact and minute finish of prosperity events. Distributed computing along with the Internet of Things (IoT) thought is another example of gainful regulating and treatment of sensor data. This paper displays a stage-dependent on IoT and Cloud-based capable choice emotionally supportive network for identification of Chronic Kidney Disease (CKD). Additionally, a Deep Neural Network DNN classifier is utilized for the assurance of CKD.An Oppositional Crow Search (OCS) optimization approach is acquainted with selects the necessary features and takes out the undesirable features and also it enhances the process of DNN.The presented structure gathers the patient information utilizing the IoT gadgets joined to the client which will be put away in the cloud alongside the related therapeutic records from the UCI vault.The exhibitions of proposed strategies are gotten by utilizing not many of the assessment measurements for example accuracy, specificity, execution time,and sensitivity. A similar investigation demonstrates that our proposed system achieves better arrangement sensitivity, accuracy, execution time, and specificity estimate when contrasted and different classifiers. The proposed model produces the 97.71% accuracy, 98.88% sensitivity and 93.44% of specificity for the analysis and the exploratory outcomes exhibit that the created OCS-DNN is outflanked.
Keywords: OCS; DNN; Optimization; Cloud; IoT and CKD.
Detection of Atrial Fibrillation from Cardiac Signal using Convolutional Neural Network
by Saumendra Kumar Mohapatra, Mihir Narayan Mohanty
Abstract: Multi-disciplinary research including engineering and medicine paves the way of modern research. In this work, authors have taken attempt to classify long duration ECG signal. The data is of 30-60 seconds and is collected from form '2017 Physionet/Computing in Cardiology Challenge database'. Use of this database for analysis of long duration signal in terms of data mining is the novelity. This analysis is performed for ambulatory ECG diagnosis and monitoring. As the pre-processing of the signals Savitzky-Golay (SG) filter is used. The filtered signals are classified with the designed deep neural network model. Convolutional neural network (CNN) is used for data classification. The classification task carried for four classes: normal, atrial fibrillation, alternative rhythm, noisy. A ten layer CNN model is designed to classify these four classes of signals. Sensitivity, specificity, and accuracy, these measuring parameters are used for the performance evaluation. Result found from the proposed method is promising one as compared to earlier methods and exhibited in result section. The accuracy with this data is 95.89%. For the comparison purpose RBFN, SVM, MLP, and PNN methods have been verified and presented.
Keywords: ECG; Neural Network; Deep learning; Convolutional Neural Network.
Fuzzy Logic Control based High Step Up Converter for Electric Vehicle Applications
by Ravindranath Tagore Yadlapalli, Anuradha Kotapati
Abstract: At present, the Electric Vehicles (EVs) have a dominant role for visualizing the pollution less environment. Morover, it is important to design efficient power conditioning units for fulfilling the desired tasks. With advancements in the technology, the Fuel Cells (FCs) can provide a better solution for powering the EVs. However, their downside is the production of very low output voltage. Therefore, this necessitates a high step ratio DC-DC converter for providing the desired dc link voltage for DC-AC converter. This paper focuses on fuzzy logic control (FLC) based high gain DC-DC converter for EV applications. The fuzzy controllers are superior in dealing the nonlinearities and plant parameter variations without the need of strict mathematical modeling of the converter system. The converter is simulated with simple voltage mode control as well as FLC based voltage mode control. The MATLAB simulation results are compared for the above two control strategies with an emphasis on the steady state and dynamic regulations.
Keywords: Fuel cells; electric vehicles; power converters; control strategies.
A Prediction Method For The Service Trajectory Of Badminton Moving Video Based On Fuzzy Clustering Algorithm
by Lin Zhu
Abstract: Due to the weak regularity of target's motion attitude in badminton video, the existing methods can't segment the image accurately, which leads to the low accuracy of predicting the service track of badminton video. A method of predicting the service track of badminton based on fuzzy clustering algorithm is proposed. The fuzzy clustering algorithm is used to segment the video image of badminton movement, obtain the contour of the serving arm in the video image, calculate the body state ratio and compactness of the serving arm of the moving object, construct the overall matching similarity function of the moving object, and realize the tracking and trajectory prediction of the moving object in the video image. The experimental results show that the proposed method has good noise iteration performance, clearer target of the processed motion video image, and can accurately predict the serve trajectory in the badminton motion video.
Keywords: Fuzzy clustering; Badminton; Moving video; Service trajectory; Morphological operator; Marginal profile.
An Empirical Evaluation of Strategies Based on the Triangle Inequality for Accelerating the k-Means Algorithm
by Marcelo Matte, Maria Nicoletti
Abstract: The k-Means clustering algorithm has a long history of success in a wide range of applications in many different research areas. Part of its success is due to both, the simplicity of the algorithm, which helps its quick implementation and the good results it produces. Despite success, however, the original k-Means has some shortcomings. One of them relates to the processing time required for the algorithm to finish the iterative process that, given a set of data instances, and an integer value k, induces a clustering having k clusters of the given instances. This article presents an empirical evaluation of three strategies found in the literature, with the purpose of accelerating the k-Means processing time.
Keywords: k-Means; optimization; triangular inequality; clustering; Machine Learning; acceleration.
A High Precision Stereo-Vision Robotic System for Large-sized Object Measurement and Grasping in Non-structured Environment
by Guoyang Wan, Guofeng Wang, Kaisheng Xing, Tinghao Yi, Yunsheng Fan
Abstract: Handling and loading of large-sized objects represent a challenging task in industrial environments, especially when the object is a metal object with reflective surface features. Active stereo vision technology is not good for measuring the posture of reflective metal object, a high-precision pose measurement system based on passive stereo vision is proposed to automatic measurement, handling and loading of large objects by industrial robots in industrial environments. The system adopts advanced coarse and fine stereo vision positioning strategy, and realizes the high-precision positioning of the measured target for the premise of ensuring stability. For coarse positioning, an improved multi-models template matching method based on machine learning is proposed to robust recognition of the multiple objects in complex background. A RANSAC-based method for the ellipse fitting and multi-points plane fitting is proposed for the 6-DOF pose of object accurately obtained in fine positioning step. Compared with the classical CAD-views method, experiments show that the method proposed in this paper has better performance in positioning accuracy and recognition robustness.
Keywords: Industrial robot; machine vision; stereo vision; 3D measurement; template matching; coordinate transformation.
A Robot-Soldering Workstation Combined with the Deep Learning and the Template Matching Technology
by Guoyang Wan, Tinghao Yi, Guofeng Wang, Kaisheng Xing, Yunsheng Fan
Abstract: To improve the stability and precision in PCB soldering application, an unmanned robot workstation is proposed to solve the problem of automatic soldering of different model chips. In the first step, a novel 2D hand-eye calibration method is developed to acquire a high-precision transformation model from the coordinate system of the vision system to the robot working coordinate system. Then a robust classification and location method is developed to realize high precision positioning of PCB objects, which is based on deep neural network combining with the traditional template matching. The proposed method solves the problems of poor positioning accuracy and low recognition rate when traditional machine vision detects PCB objects with the same shape but different models. And the automation of chip soldering is realized. The experimental results show that the proposed algorithm displays excellent robustness.
Keywords: deep learning; template matching; hand–eye calibration; coordinate transformation; object detection.
Optimization of Plagiarism Detection using Vector Space Model on CUDA Architecture
by Jiffriya Mohamed Abdul Cader, Akmal Jahan Mohamed Abdul Cader, Hasindu Gamaarachchi, Roshan G. Ragel
Abstract: Plagiarism is a rapidly rising issue among students that occurs during submission of assignments, reports and publications in universities and educational institutions because of the easy accessibility of abundant e-resources on the Internet. To mitigate plagiarism among students, many tools are available for natural language plagiarism detection. However, they become inefficient when dealing with prolific number of documents with large content due to the time they consume. Therefore, we have proposed a way for software-based acceleration on text-based plagiarism detection using a suitable model on CPU/GPU. For the evaluation in CPU, initially software-based serial vector space model was implemented on CPU and tested with 1000 text-based documents particularly, students assignments, where it consumed 1641s for plagiarism. As computation time of the plagiarism detection is a bottleneck of the performance while treating a prolific number of text-based sources with different sizes, we focus on accelerating and optimizing the model with the number of documents. Therefore, this research intends to implement and optimize the vector space model on the Graphics Processing Units (GPU) using Compute Unified Device Architecture (CUDA). In order to speed-up, the model was developed on GPU using CUDA, tested with the same dataset which gained 45x speed up compared to CPU, and it optimized further showed 389x faster than serial implementation
Keywords: Graphics Processing Units (GPU); Computer Unified Device Architecture (CUDA); Plagiarism Detection; Vector Space Model.
A Robust and secure watermarking technique based on projecting the original
image using rotating views
by Djalila Belkebir
Abstract: Due to their high computational complexity, frequency domain-based steganography techniques are unsuitable for applications that require real-time constraints. These applications are dedicated for devices that are designed with limited source of power. In this paper, we propose low-computational complexity, robust and secure watermarking technique, that is based on the strategy of projecting pixels from the original image into a new image while rotating the view. The pixels positions in the new image are then used to hide the secret message. A comprehensive performance evaluation has been carried out against a vast variety of other related watermarking techniques. Numerous image quality metrics have been employed to measure the performance of the proposed technique. In addition, histogram analysis and RS steganalysis tests have been carried out for security assessment. The outcomes confirmed that the secret message remains robust and secure against different detection attacks.
Keywords: Direct image projection; LSB; Steganography; Watermarking; Projection; CRC32.
Special Issue on: Security, Privacy and Trust in Cognitive-inspired Computing and Applications
Structural Equation Modeling and Numerical Simulation on the Entire Process of E-commerce Model
by Xiaoxi Fu, Xuecheng Sun, Wentao Li, Mark Roger
Abstract: Nowadays, most enterprises introduce e-commerce to enhance efficiency, broaden the market and restructure their organization. Therefore, many scholars have taken notice on how to introduce and implement e-commerce model in traditional enterprises successfully. This study tried to explore the complex relationship between original enterprise operation and e-commerce model using numerical simulation. Through the simulation of the entire process of e-commerce mode introduction and implementation, we found that traditional operation could promote the implementation of e-commerce at initial phase. With the implementation of the e-commerce model, the original operational base will become an obstacle. To cope with this, companies need to break the inherent boundaries and industry concepts and simplify the organizational structure to promote the full utility of e-commerce model.
Keywords: Enterprise operation; E-commerce model; Numerical simulation; Structural equation modeling.
DOI: 10.1504/IJICA.2021.10035131
Safety Evaluation on Central Separation Opening of Reconstructed Freeway Based on Surrogate Safety Assessment Model
by Wei Hou, Zijun Du, Zhaoxin Liu, Xu Wang
Abstract: In order to reduce the risk of driving at the opening of the median strip of the old road of the freeway, the driving safety was quantified by taking the opening length of the central section of the old road and the main line flow as independent variables. Based on the VISSIM micro-simulation software, a typical simulation scenario is established. The vehicle speed, the position of the vehicle's change point, the number of conflicts, the severity of the conflict, and the conflict points are analyzed by the Surrogate Safety Assessment Model (SSAM). The influence of the opening length of different median strips and the influence of flow on safety is quantitatively evaluated, so as to obtain a reasonable opening length of the median strip. The research results show that the flow has a greater impact on the number of collisions, and the length of the open section has a significant impact on the severity of the conflict. The conflict points are found to be concentrated at the front of the 400m open section. The results of the study provide a theoretical basis for the opening length of the median strip and the setting of traffic signs.
Keywords: safety evaluation; conflict; trajectory; median strip; freeway; single side widening.
An Improved Sensorless Strategy on the basis of Improved PI regulator
by Haigang Zhang, Piao Liu, Xiaoqi Sun, Xiangsheng Kong, Xuan Chen, Bulai Wang, Heng Wan
Abstract: The sensorless control strategy was implemented by replacing the conventional PI regulator with a improved fuzzy PI regulator, meanwhile a adjusted sliding mode observer was established for replacing the traditional rigid position sensor based on the principle of non-position sensor. The signals of rotor volocity and location were extracted from the counter electromotive force for reducing the error and complexity of the calculation of the angular and rotational speed. Compared with conventional PI observer, the regulate system of the PMSM has a preferable influence on the steady-state and dynamic response. A Matlab platform has been built to validate this method, and the experimental results showed that the fuzzy PI sliding mode observer had strong robustness and restrained the chattering to a certain extent. On the other hand it also could improve the stability, rapidity and dynamic performance of the SMO system.
Keywords: Fuzzy Control?Non-position Sensorless?PMSM?Sliding Mode Observer.
Cryptologic Characteristics of Circulunt Matrices
by Haiqing HAN, Qin Li
Abstract: A 4
Keywords: Circulant Matrix; Branch Number; Orthormorphic Matrix; Cryptologic Characteristic; Symmetrical Permutation.
Application of EPC Internet of Things and Radio Frequency Identification Technology in Logistics
by Xiaobei Wang, Yaya Wang
Abstract: E-commerce is growing stronger, more and more flexible and diverse application to various industries of society, Initiating innovation in shopping and trading methods. In particular, major changes have taken place in the circulation of goods. Then, How to make electronic money safely and reliably purchase real-world items in a virtual network? Logistics plays a key role in this. Realizing informationization, automation and intelligence has become a new trend in the development of modern logistics. The continuous development of network development provides a broad development prospect and technical support for logistics. Modern logistics and the Internet complement each other, and support the commercial application of modern networks. How to apply existing network technology to effectively improve the logistics level is an important topic of our research. This paper focuses on the practical application of EPC and RFID technology in the field of logistics. The Internet is the foundation of the Internet of Things, IOT using wireless communications, RFID technology, computer network technology etc, construct a huge system that covers the world. In order to meet the identification and efficient identification of individual products, the Automatic Identification Laboratory of the Massachusetts Institute of Technology proposed the concept of EPC (Electronic Product Code), each item is assigned a unique identification code. The EPC code is usually stored in an electronic tag of a silicon chip material, and the tag is attached to the identified article. When an EPC tag is embedded in an item, the item and the product electronic code in the EPC tag are in one-to-one correspondence. It is identified, transmitted, and queried by high-level information processing software?On the basis of the Internet, a logistics model that provides various information services for the supply chain is formed[1] [6]. As a wireless version of the barcode, RFID technology has its own advantages and features non-contact automatic identification. It mainly identifies the object objects in the logistics by issuing specific RF signals and obtains related item information. The perfect integration of electronic tags, EPC codes and Internet technologies has created the Internet of Things known as the next generation Internet.
Keywords: EPC; RFID; Computer network technology; Internet of Things.
The Effects of Green Product Trust and Perception on Green Purchase Intention in China
by Pinghao Ye, Liqiong Liu, Linxia Gao, Quanjun Mei
Abstract: According to the theory of planned behavior (TPB), we created a model of factors affecting customer green purchase intention, with such factors including green product experience, green purchase behavior, subject social norms, perceived environmental protection value, green product trust, and green product perception. Then, a questionnaire was developed based on the model. 307 respondents were involved and the feedback was analyzed, using a structural equation model, to verify the data reliability and validity. Results show that green product trust and green product perception exerted significant positive effects on green purchase intention; green product experience had significant positive effects on green product trust; and subjective norms and perceived environmental protection value positively affected green product perception. These findings provide valuable information for studying factors affecting the green purchase intention of customers.
Keywords: Green product trust; Green product perception; Green purchase intention; Perceived environmental protection value.
Terrain Frames Classification Based on HMC for Quadruped Robot (February 2017)
by Zhe LI, Yibin LI, Xuewen RONG, Hui ZHANG
Abstract: As a multi-body nonlinear rigid-flex system, the quadruped robot must maintain the correct perception and classification capabilities for the external environment. This ability is necessary to help quadruped robots make path planning, gait adjustment and attitude control while maintaining complex interactions with the external environment. This paper proposes a terrain classification algorithm based on HMC (HMRF-MAP-CNN) as the basis for robot motion control strategy selection. Different from the classification method based on image features, the terrain-based classification method has higher accuracy and better computational efficiency. The raster map terrain frames classification method proposed in this paper includes a rugged terrain recognition algorithm based on HMRF and a rugged formation algorithm based on convolutional neural network. In the process of solving the actual terrain classification problem, the algorithm firstly uses HMRF to classify the obtained terrain frames into two categories, flat and rugged, and then use CNN to filter, according to the causes of rugged terrain frames. Through the simulation experiment and comparative analysis, the superiority of HMC terrain frame classification algorithm is confirmed.
Keywords: quadruped robot; terrain classification; HMRF; MAP; CNN.
Energy efficiency optimization for heterogeneous cellular networks based small cell power allocation
by Ziyu Pan, Jie Yang, Nan Guo
Abstract: During the study of wireless green communication, energy efficiency (EE) optimization of heterogeneous cellular networks (HCNs) is always the hot topic. This paper studied the EE of HCNs in homogenous Poisson Point Process (HPPP) and proposed a new EE optimization method based on small cell transmitting power allocation. First, it modeled the HCNs by using HPPP. Second, it derived the expression coverage probability and achievable rata in Rayleigh channel situations, and deduced the expression of EE in closed form. Finally, it optimized the small cell transmitting power via convex-optimum method in order to maximize EE for HCNs. The simulation results show that the small cells transmitting power has significant effects on the HCNs EE. Its essence is through reasonable setting of small cells transmitting power to available help improve its efficiency?
Keywords: Heterogeneous Cellular Network; energy efficiency; small cell; transmission power.
Research on the Evaluation System of Smart Tourism Projects Based on User Experience
by Yipeng Wang
Abstract: In recent years, with the rise of smart tourism, many tourist attractions invest a lot of money to build smart tourism projects to make tourism more intelligent. In order to evaluate the problems in the use of such projects and point out the direction of improvement, this paper proposes a smart tourism evaluation index system based on user experience. The evaluation index system probes into the actual feedback data of the users in the tour process from users sensory experience, program experience, content experience and related experience. Then I utilize the fuzzy comprehensive evaluation method to process the collected data to obtain the evaluation index under the framework of the index system. In the application of this evaluation index system?it is compared with the commonly used evaluation index system in industry to point out the advantages and disadvantages of smart tourism projects and the direction of improvement more directly and effectively.
Keywords: Smart tourism;User experience;Evaluation index system.
Development of Microcomputer Protection and Measuring-control Device for 35 kV Distribution Line
by Tianpeng Zhang, Yafang Zhai, Zhanwei Wu
Abstract: According to the research status of 35 kV microcomputer protection device, a microcomputer protection and measuring-control device was designed to address common electrical failures during the running of 35 kV distribution line. In this device, a 32-bit high performance embedded microprocessor was used as the control center, which could realize real-time detection on electric parameters such as current, voltage, active power, reactive power, and provide the distribution line different protections, include current direction protection, over-current acceleration protection, zero sequence current direction protection and under frequency load shedding. This paper mainly focused on the overall structure of the device, gave design proposals of analogue acquisition module, switch processing module, communication module and power module. In addition, it also analyzed various protection logic such as three-phase current direction protection with low-voltage blocking, over-current acceleration protection, three-phase zero sequence current protection and under frequency load shedding protection, moreover, partial test data was also showed. In general, this device combined functions of protection, measurement, control and communication, had a high application value.
Keywords: Distribution Line;Microcomputer Protection;Measuring-control Device.
Research on the Sponge City Smart Platform of Wenzhou High-speed Railway New City
by Shuang Han, Meng Li
Abstract: Using the new technology of Internet of things, big data and cloud computing, an intelligent sponge city platform with the functions of monitoring, data transmission, early warning and management is built. The hardware system based on NB-IoT and STM32 and the software system of T-S fuzzy neural network and BP neural network are designed. BP neural network is used to calculate and analyze the water quality data measured by the sensor, the calculation results are fed back to the drainage pipeline and drainage pump station facilities to control the water quantity and quality of the outlet. In addition, T-S fuzzy neural network is used to reasonably predict the changes in the water depth of the low-lying areas of the city from the flow data measured above, and make a warning response to possible waterlogging as soon as possible. The relevant staff can use the timely monitoring data of the smart platform to take necessarily remedial measures through the special management application. The results show that the employment of the smart platform has reduced the annual runoff of Wenzhou City by 40%, the concentration of the suspended solids in river rainwater has been reduced by more than 40%, and the concentration of the chemical oxygen demand, ammonia nitrogen, total nitrogen and total phosphorus has been reduced by more than 30%, which reduces the economy loss by 5 million yuan and improved the aqueous environment to some extent.
Keywords: sponge city?smart platform?NB-IoT; BP neural network; T-S fuzzy neutral network; underground rainwater storage tank; urban waterlogging.
An Efficient and Reliable Approach Based on Adaptive Threshold for Road Defect Detection
by Xiaoliang Jiang, Xiaojun Yang, Xiaokang Ding
Abstract: With the rapid development of economy, automatic detection of road crack becomes a hot research study. However, it still has very big challenges due to the shading, intensity inhomogeneity and divots in the crack image. Traditional detection algorithms use the characteristic of the image such as intensity and texture to segment the crack regions, so they cannot achieve satisfactory performance. In this article, we introduce an automatically image-based method for road crack detection. Firstly, a new mask image is constructed based on the gray-mean of original image. Secondly, calculating the difference between the original image and the mask image and selecting the highest value to extract the targets from the image background. Finally, the redundant interference is removed by morphological operation. Experimental results concentrate on public datasets show that our method is more reliable than previous approaches.
Keywords: Road crack detection; Image segmentation; Otsu; Adaptive threshold.
Special Issue on: Artificial Intelligence for Sustainable Future Computing
A Review on Existing Learning Techniques Applied in Solving Optical Character Recognition Problem
by Vijaya Krishna Sonthi, Nagarajan S, Krishna Raj N
Abstract: Optical Character Recognition is treated as the classical problem in Image Processing that can be widely used for application areas related to Computer Vision. This process is highly dependent on the languages, as well as various formats of information like hand written/printed/sculptured documents as input. The chief operations that can be performed while handling this task are training the system and predicting the characters based on the trained knowledge. Conventionally, classification and regression mechanisms are used for this purpose. Due to the recent advents in technological progress, the same mechanism was served with the help of modern Machine Learning and Deep Learning concepts. The main agenda of this paper is to create awareness among the researchers by discussing the basic functionalities of above said techniques.
Keywords: Machine Learning; Deep Learning; Optical Character Recognition; Pattern Recognition; Image Processing;.
DOI: 10.1504/IJICA.2021.10033832
A NOVEL APPROACH FOR THE SOLUTION OF GENERALIZED FUZZY ASSIGNMENT PROBLEM
by E. Melita Vinoliah, K. Ganesan
Abstract: In Engineering and other fields, generalized assignment problem plays a vital role. It has widespread applications in routing problems, knapsack problems and other complicated models. In real world problems, the available data may not be known with certainty. Hence to model and solve practical problems, we must deal with uncertainty and vagueness. Fuzzy sets play a vital role to tackle these uncertainty and vagueness. The generalized fuzzy assignment problem became popular and has gained its importance too. In this paper, we consider a generalized fuzzy assignment problem which has restrictions both on tasks and on persons with respect to his/her efficiency / qualification. We propose a unique approach for the solution of generalized fuzzy assignment problem with restrictions without converting the problem to a corresponding crisp form. Costs for handling jth task by the ith person are taken to be trapezoidal fuzzy numbers. The trapezoidal fuzzy numbers are first represented in its parametric form. In view of the decision makers preference, a new ranking method and arithmetic operations are used to bring a desirable solution. A numerical example is given to illustrate the proposed method.
Keywords: Generalized fuzzy; assignment problem; trapezoidal fuzzy numbers; fuzzy arithmetic; fuzzy ranking; extremum difference method; judging matrix; uncertainty and vagueness; generalized assignment problem; trapezoidal fuzzy number; multilevel fuzzy generalized assignment problem,.
DOI: 10.1504/IJICA.2021.10036813
A 600mV +12 dBm IIP3 CMOS LNA with gm smoothening auxiliary path for 2.4 Ghz wireless applications
by Sharath Babu Rao, V. Sumalatha
Abstract: CMOS Low Noise Amplifiers (LNAs) were used in IEEE 802.11 ac/g/n/ad RF Receivers which finds a large scale of applications in gyroscopes, accelerometers, biomedical applications such as neural, nerve, cardiac, biological probes, and other mission sensors. Additionally, LNAs were found in an environment, to amplify weak signals with in-sufficient radiation from WI-FI transmitters operating in the 2.4/5 GHz range, which include cordless telephones, Bluetooth devices, wireless keyboards, and radio equipment. Non-Linearity limitations of the LNA come up with producing altered (or modulated) amplitudes resulting from higher-order frequency component (Especially third-order harmonics) interaction with fundamental frequency (first-order harmonic) components. For radio-frequency (RF) applications where signal frequencies are high, consequences of intermodulation distortion results from the nonlinearities of the MOS Transistors used in the LNA. As it is impracticable to eradicate nonlinearities completely, Second-order nonlinearities were added in out of phase with the third-order nonlinearities in Derivative and Modified Derivative methods found in the literature can able to suppress the third-order nonlinearities up to an Input Intercept Point(IIP3) of 8 dBm only. In this article proposed LNA smoothens the nonlinearity dependence on the transconductance, can able to increase the IIP3 performance up to 12 dBm using the common source differential architecture with PMOS Loads in the auxiliary path used to suppress the nonlinearities in the LNA circuitry. Because of battery operation or limited power supply capacity, these application areas require low power operation. The proposed LNA offers gain (S21) of 12 dB operates at ultra-low voltage supply headroom of 600mV with good stability deemed as a favorable requirement for low power wireless applications.
Keywords: 802.11 b/g/n; Local Area Network (LAN); Wireless LAN(WLAN); Wireless Sensor Networks (WSN’s); Wireless Personal Area Networks (WPAN’s); Low Noise Amplifier (LNA); Generic Process Design Kit (GPDK); Berkeley Short-Channel IGFET Model version 3 (BSIMv3); Cascode LNA; Folded Cascode LNA. Modified derivative superposition (MDS) technique.
DOI: 10.1504/IJICA.2022.10035002
Special Issue on: Smart Computational Intelligence and Optimisation Method Applications to Electrical Power Engineering
XGBoost Regression Model Based Electricity Tariff Plan Recommendation in Smart Grid Environment
by Dayal Kumar Behera, Madhabananda Das, Subhra Swetanisha, Janmenjoy Nayak
Abstract: Electrical Power System is an interconnected network of power generation, transmission and distribution of electricity from a power plant to the consumer. Smart Grid is an advancement over the power grid that facilitates an efficient and reliable two-way delivery system. It acts as a backbone infrastructure to enable various cutting-edge models like smart city, efficient smart metering and tariff retailing. Power System deregulation enables the power industry to provide residential customers to choose retailing electricity plan. This allows competition among retailers or traders and also minimizes the energy expenditure with quality of services. Few researches have been proposed for electricity tariff plan recommendation using Collaborative filtering approach. We have proposed an XGBoost regression model for Electricity Tariff Plan Recommendation. Firstly, proposed regression model with basic statistical features is compared with Support Vector Regression (SVR), Decision Tree (DT), Bayesian Ridge and KNN regression model. Secondly, performance of the proposed model is extensively studied by combining the features from other user-based, item-based and matrix factorization based techniques. In this research, dataset shared in the project Smart Grid Smart City (SGSC), Australia is used for conducting experimental analysis. A rating inference approach is designed to infer the choice of electricity consumer for a specific retailing plan. The proposed model achieves better performance as compared to other baseline methods.
Keywords: XGBoost Regression Model; Recommender System; Smart Grid; Power System; Electricity Tariff Plan Recommendation.
Frequency regulation of hybrid power system using firefly algorithm
by Tulasichandra Sekhar Gorripotu, Pilla Ramana
Abstract: In the present work, Firefly Algorithm (FA) based cascaded PD-PID with filter controller (PD-PIDF) is implemented for the hybrid power system to keep the frequency within the limits. At first, a two-area hybrid power system is proposed, considering thermal and distributed power units in area-1, and hydrothermal units in area-2. The cascaded PD-PIDF controller parameter values are optimized using the integral time multiplied absolute error (ITAE) criterion. The proposed system is analyzed under three cases by applying step load perturbations (SLPs) and noise signals at possible areas. The superiority of cascade PD-PIDF controller is shown by comparing with recently published results. Finally, sensitivity analysis is performed to demonstrate the cascaded PD-PIDF controller capability while varying system parameters and conditions.
Keywords: Cascaded PD-PIDF controller; distributed power system; firefly algorithm (FA); frequency regulation; hybrid power system.
Speed Control of Battery and Super capacitor Powered EV/HEV using PID and Fuzzy Logic Controller
by NITESH TIWARI, Shekhar Yadav, Sabha Raj Arya
Abstract: State of art of this paper is to obtain the fine speed control of EV/HEV and analysed the behaviour of the PID controller and FLC. FLC based drive scheme is more suited for achieving the fine speed control on comparing to the traditional PID controlled drive scheme. FLC based EV motor drives perform well during normal and abnormal working conditions including environmental parameter variation and high resistance fault condition on comparing to PID controller based EV motor drive scheme. The battery bank is utilized as a power source for the conventional EV but this paper proposed the combination of the battery bank (followed by the UBBC) and super-capacitor bank (followed by the BBBC) as a power source for HEV. A super-capacitor is provided power at the time of sudden or peak demand and extracts power at the time of regenerating breaking. All the results are verified with the help of simulation software named as MATLAB and Simulink.
Keywords: Electric vehicle (EV); hybrid electric vehicle (HEV); unidirectional buck-boost converter (UBBC); bidirectional buck-boost converter (BBBC); proportional integral derivative (PID); fuzzy logic controller (FLC); battery; super-capacitor (SC).
Implementation of beta- Chaotic Mapping to Improved Elephant Herding Optimization to Dynamic Economic Dispatch Problem
by JAGANNATH PARAMGURU, Subrat Kumar Barik
Abstract: In this paper, a chaotic function is mapped to an improved novel swarm intelligence technique for performing dynamic economic operation of conventional generators. Practical constraints like ramp rate limit, valve point loading effect and transmission losses are taken into account for the operation of the generators. This modified optimization algorithm is proposed for solving economic load dispatch problem based on the herding behavior of the elephants under the leadership of the matriarch. Here two-updating processes involved for the optimum operation, which depend upon the response of each elephant in a clan and the separating behavior of male elephant. The existing technique is modified and certain ?-Chaotic function is being mapped to improve exploration and exploitation superiority. This proposed technique is applied and compared for four test system to find the dominance over other applied techniques. A comparative analysis is illustrated with other techniques for this proposed method to deduce the supremacy and enhancement.
Keywords: Dynamic Economic Dispatch (DED); Elephant Herding Optimization (EHO); Chaotic Elephant Herding Optimization (CEHO); Ramp rate limit.
A novel modified random walk grey wolf optimization approach for non-smooth and non-convex economic load dispatch
by ARUN SAHOO, Tapas Kumar Panigrahi, Gopal Krishna Nayak
Abstract: In practice, economic load dispatch (ELD) problems have non-smooth and non-convex cost functions which are subjected to numerous non-linear equality and inequality constraints involving multiple decision variables. It makes the problem computationally labyrinthine to solve via any analytic method. Therefore, this study proposes a coming of age improved version of conventional grey wolf optimization (GWO) technique to solve the ELD problem. In the improved GWO, the leadership hierarchy of the grey wolf is ameliorated by taking the random walking behaviour of the grey wolfs into consideration. The algorithm aims to modify the existing leaders with the best leaders in order to overcome the drawbacks of the conventional GWO. The improved GWO guarantees better exploration and exploitation of the search space. The non-linear constraints of generating units like ramp rate constraints, influence of valve-point loading and Prohibited Operating Zones (POZs) are taken into account for both lines with and without losses. The obtained results are compared to that of other contemporary algorithms for demonstrating the superiority of the suggested one. This technique provides the optimum dispatch with a faster convergence rate in comparison to the conventional GWO and some other existing methods.
Keywords: Economic Load Dispatch (ELD); Constraints; Grey Wolf Optimization (GWO); Modified Random Walk Grey Wolf Optimization (MRWGWO); Valve Point Loading.
Special Issue on: IWSSIP 2020 Latest Scientific and Theoretical Advances in Systems, Signals and Image Processing
Performance Evaluation of Energy Reconstruction Methods for the ATLAS Hadronic Calorimeter Using Collision Data
by Juan Marin
Abstract: Particle accelerators are machines that collide particle beams next to the speed of light. The Large Hadron Collider, the largest and most powerful beam collider, operates at a center-of-mass energy of 13~TeV with a 40~MHz collision rate. In the ATLAS experiment, optimal filter techniques can be applied on energy estimation even in the occurrence of noise and signal overlapping. In this context, the present work compares the performance of the energy estimation in the main hadronic ATLAS calorimeter, for two methods: the baseline algorithm OF2 (Optimal Filter) and COF (Constrained Optimal Filter). The work proposes an adaptive estimator of the energy signal pedestal in the ATLAS experiment at the LHC. For performance comparisons, the statistics from the energy estimation distributions are employed. The results show that the COF method presents a better performance than the OF2 method in terms of estimation error. The proposed signal baseline estimator properly improves the COF efficiency.
Keywords: High-energy physics; Signal estimation; Optimal filtering; Signal pile-up.
Light Leaf Spots Segmentation Algorithm based on Color Difference Vectors
by Sheng Miao, Weili Kou, Lihui Zhong
Abstract: Leaf spots image segmentation and quantifying is a crucial step of disease image recognition, which is very important for monitoring plant diseases and subsequently diagnosis. In the leaf spots segmentation, compared with dark spots, light leaf spots segmentation is hinder due to unobvious color characteristics, and easily confused with leaf veins. This study specially focuses on the light leaf spots segmentation and vein removal. First of all, a mask extraction model was constructed to get leaf spots mask, includes three steps: L * a * b color model mapping, color difference vectors construction and color difference vectors segmentation. Secondly, veins mask and spots mask have been extracted from healthy leaves and spots leaves by mask extraction model respectively. And vein ratio of vein mask was calculated to evaluate whether vein removal is needed or not. Thirdly, vein area and roundness were calculated for both vein mask and spots mask, and remove areas with high similarity to vein features from spots mask. Leaves with light spots of 12 plants types were selected to test proposed algorithms, the experimental results show: compared to the manual segmentation method for identifying light leaf spots, the presented algorithm has higher average accuracy (86%) than both traditional watershed (77%), Barbedo 2017(82%) and CNN (81%). Moreover, this algorithm is easy to be used in other plant species.
Keywords: Light leaf spots; Image segmentation; Vein removal; Color difference vector.
Detecting Acute Leukemia in Blood Slides Images Using a CNNs Ensemble
by Maíla Claro, Rodrigo Veras, Luis Vogado, André Santana, Vinicius Machado, Justino Santos, João Tavares
Abstract: Leukemia is a disease that has no defined etiology and affects the production of white blood cells in the bone marrow. Young cells or blasts are produced abnormally, replacing normal blood cells (white, red blood cells, and platelets). Consequently, the person suffers problems in transporting oxygen and infections combat. Acute leukemia is a particular type of leukemia that causes abnormal cell growth in a short period, requiring a quick start of treatment. Classifying the types of acute leukemia in blood slide images is a vital process, and a system of assisting doctors in selecting treatment becomes necessary. This article presents an ensemble approach composed of three convolutional neural networks (CNNs) - Alert Net-RWD, Resnet50 and InceptionV3. These CNNs, individually, demonstrated effectiveness in differentiating blood slides with Acute Lymphoid Leukemia (ALL), Acute Myeloid Leukemia (AML), and Healthy Blood Slides (HBS). We verified that the union of these three well-known CNNs improves the hit rates of current techniques from the literature. The experiments were carried out using 18 public data sets with 3,293 images, and the proposed CNNs ensemble achieved an
accuracy of 96.17%, and precision of 96.38%
Keywords: Acute leukemia diagnosis; model ensemble; convolutional neural network.
Evaluation of Banknote Identification Methodologies Based on Local and Deep Features
by Leonardo P. Sousa, Rodrigo M. S. Veras, Luis H. S. Vogado, Laurindo S. Britto Neto, Romuere R. V. Silva, Flávio H. D. Araujo, Fátima N. S. Medeiros
Abstract: There are many people with disabilities; it is estimated that 39 million people are blind and 246 million have limited vision, giving a total of 285 million visually impaired people. Information and communication technologies can help disabled people achieve greater independence, quality of life, and inclusion in social activities by increasing, maintaining, or improving their functional capacities. This paper presents a significant evaluation of local and deep features for an automatic methodology for identifying banknotes. To determine the best local features, we evaluated a set of four point-of-interest detectors, two descriptors, seven ways of generating the image signature, and six classification methodologies. To define the deep features, we extract features using three pre-trained well-known CNNs. Additionally, we evaluated using a hybrid descriptor formed by the combination of local and deep features. In this situation, the features were selected according to their gain ratios and used as input to the classifier. Experiments performed on US Dollar (USD), Euro (EUR), and Brazilian Real Banknotes (BRL) obtained rates of accuracy of 99.96%, 99.12%, and 96.92%, respectively.
Keywords: accessibility; visually impaired; banknote recognition; assistive technologies.
Multilevel CNN for Anterior Chamber Angle Classification using AS-OCT Images
by Marcos Ferreira, Geraldo Braz, João Almeida, Anselmo Paiva, Rodrigo Veras
Abstract: Glaucoma is the second leading cause of blindness and the first leading cause of irreversible blindness. The main types of the disease are open-angle and angle-closure glaucoma. In people with angle-closure, the anterior chamber angle is narrow, which leads to a rising intraocular pressure, and consequently, optic nerve damage, causing vision loss. Since it is irreversible, an early diagnosis is essential. So, angle classification is fundamental for diagnosis.
Anterior segment optical coherence tomography is one of the imaging tests used to diagnose the disease. In addition to the fact that no eye contact is necessary, this test provides a fast way to analyze the anterior chamber angle, to classify it as open or closed. We propose the anterior chamber angle classification method based on visual feature extraction, using deep neural networks in this work. In a multilevel architecture, different pre-trained CNNs are adjusted to extract deep features and train two classifiers. The best model extracted visual features from the anterior camera angle in the experiments, and achieved a sensitivity value of 1.000 as the best result.
Keywords: Angle-Closure Glaucoma; Transfer Learning; Deep Features; Multilevel Networks.
Transfer Learning Based Lung Segmentation and Pneumonia Detection for Pediatric Chest X-Ray Images
by Vandecia Fernandes, Gabriel Bras, Lisle Faray De Paiva, Geraldo Braz Junior, Anselmo Cardoso De Paiva, Luis Rivero
Abstract: Pneumonia is the leading cause of morbidity and mortality in under-five children, especially in developing countries. Accordingly to UNICEF and World Health Organization, a child dies of pneumonia every 39 seconds, and pneumonia kills more children than any other infectious disease, accounting for 15% of all deaths of children under five years old. In regions with a high prevalence, the early detection and treatment of pneumonia can significantly reduce children's mortality rates. Commonly, a chest x-ray is a diagnostic exam. Nevertheless, it is a problematic image for reading and interprets, requiring an expert physician. So, it is essential to provide computational methods to help exam interpretation or enhance important information. This paper proposes a transfer learning method to segment lung regions on the Chest X-Ray dataset to extract ROI for pneumonia detection. The results are promising and reach 0.917 of dice using U-Net combined with InceptionV3 in a chest x-ray dataset without lung annotation. For pneumonia detection, the method achieves 0.954 precision.
Keywords: Pneumonia; Transfer Learning; Deep Learning; Segmentation.
Wind Turbine Fault Detection: A Semi-Supervised Learning Approach With Two Different Dimensionality Reduction Techniques
by Fernando Sá, Rafaelli Coutinho, Eduardo Ogasawara, Diego Brandao, Rodrigo Toso
Abstract: The quest to save the environment has led many countries to change their mix of energy sources, with most countries focusing on wind energy as a clean, alternative source. Since wind turbines are at the center of this revolution, ensuring their continuous operation by preemptively detecting and correcting faults is a key to success. Towards that end, making use live operational signals from the turbine, various data-driven methods for fault prediction using traditional machine learning have been proposed. As with all traditional machine learning solutions, these require careful feature selection and hyperparameter tuning for each turbine, which is not simple to scale. This work adopts a novel, automatic, end-to-end AutoML approach covering aspects from features and hyperparameters selection to fault prediction using semi-supervised support vector machines guided by the multi-objective optimization framework Non-dominated Sorting Genetic Algorithm II (NSGA-II). Experiments were carried out using a dataset containing unlabelled records of five 2.0MW wind turbines. In the results, we analyze and compare the approaches with respect to fault detection performance. We found that our AutoML approach using NSGA-II for feature selection offers up to 9% improvement in solution quality over the state-of-the-art while being fully automated and requiring no costly and time-consuming feature engineering.
Keywords: Feature Selection; Fault Detection; Wind Turbine; Semi-Supervised Learning.
Heuristic-based approaches for fracture detection in borehole images
by Maira Moran, Eduardo Vasconcellos, Jordan Cuno, Mauro Biondi, Jose Riveaux, Maury Correia, Esteban Clua, Aura Conci
Abstract: The success of an oil well drilling process depends, among other factors, on the borehole stability, which is closely related to fractures. Borehole imaging tools are commonly used to identify this kind of feature. Visually, fractures can be identified as sinusoidal curves in the 2D data obtained by the Ultrasonic Borehole Imager (UBI). Their automatic detection is not trivial. Previous works in literature proposed solutions for this problem. Most of them are based on extensive curve searches with a high computational cost. A recent work uses a more efficient method based on a heuristic algorithm. However, it results in many false fractures. In this work, we propose a set of 11 new heuristics-based innovative methods, composed of two different algorithms and three pre-processings, to identify the sinusoidal curves related to fractures in UBI data. Moreover, we include a post-processing step to reduce the number of incorrect fractures. We perform an evaluation of these proposed methods in real data, comparing them to the only heuristic-based method previously presented in the literature. Our achieved results are very promising, with the best methods presenting fewer false curves and a higher rate of correctly detected fractures.
Keywords: Oil drilling; Ultrasonic Borehole Image; Fractures; Sinusoidal curves; Heuristics; Variable Neighborhood Descendent; Iterated Local Search; Detection.
Special Issue on: Recent Advances in Intelligent Systems
Multimodal Multi-objective Differential Evolution Algorithm Based on Spectral Clustering
by Shenwen Wang, Xiaokai Chu, Jiaxing Zhang, Na Gao, Yao Zhou
Abstract: In recent years, in the face of the same problem in industrial production and life, decision-makers often hope to have a variety of different solutions to deal with. In other words, we hope to locate more different Pareto solutions under the condition of finding Pareto front. However, there are few researches in this field. For this reason, we propose a multimodal multi-objective differential evolution algorithm based on spectral clustering (SC-MMODE), which mainly uses some mechanisms to divide the solutions in the decision space into several mutually independent subpopulations. First, SC-MMODE uses a spectral clustering algorithm to control the decision space and form multiple sub-populations with good neighborhood relations. Secondly, a special crowding distance mechanism is used to balance the distribution of solutions in the decision space and objective space. In addition, the classical differential evolution algorithm can effectively prevent premature convergence. Then, in 17 test problems, the SC-MMODE algorithm and some new multimode multi-objective algorithms are tested simultaneously. Finally, through the analysis of experimental data, the SC-MMODE algorithm can find more Pareto optimal sets in the decision space, so it can effectively solve such problems.
Keywords: Multimodal multi-objective; Spectral clustering; Decision space;Differential evolution algorithm;Special crowding distance.
Enhancing cuckoo search algorithm with complement strategy
by Hu Peng, Hua Lu, Changshou Deng
Abstract: Cuckoo search algorithm(CS) is a simple and effective swarm intelligence algorithm. It searches for new solutions by the global explorative random walk and the local random walk, the greedy strategy is used to choose a better solution. However, the CS uses a single strategy and fixed parameters, in which performance of the CS in balancing exploitation and exploration is inadequate, which results in poor convergence performance of the CS. To cope with this problem, a novel CS variant with complement strategy (CoCS) was proposed by us, in which the new solution is generated by two strategies in a random manner. One of the strategy is an improved Levy flights, and the other is adaptive to determine the step size according to the fitness value of the step size and the number of current iterations. The algorithm also uses an improved random walk. The proposed CoCS, the standard CS, and other excellent CS variants were tested on CEC 2013 test suite. Experimental results prove that the CoCS is superior to these competitors.
Keywords: Cuckoo search algorithm; Global optimization; Complement strategy; Dynamic parameter adjustment.
Accelerating Artificial Bee Colony Algorithm Using Elite Information
by Xinyu Zhou, Yanlin Wu, Shuixiu Wu, Maosheng Zhong, Mingwen Wang
Abstract: In nature, the foraging behaviour of bee colony is always guided by some elite honeybees with the aim of maximizing the overall nectar amount. Being inspired by this phenomenon, we propose an improved artificial bee colony (ABC) variant by using elite information. In our approach, as the main way of generating new offspring, two novel solution search equations are developed based on utilizing elite information, which has the advantages of accelerating convergence rate. Moreover, to preserve the search experience of the scout bee phase, a new reinitialization method is proposed based on using elite information. Extensive experiments are conducted on the CEC 2013 and CEC 2015 test suites, and other four relevant ABC variants are included in the comparison. The results show that our approach has better performance in terms of convergence speed and result accuracy.
Keywords: Artificial bee colony; Solution search equation; Elite information; Exploration and exploitation.
Density peaks clustering algorithm based on kernel density estimation and minimum spanning tree
by Tanghuai Fan, Xin Li, Jiazhen Hou, Baohong Liu, Ping Kang
Abstract: Rodriguez et al. reported a clustering algorithm that can achieve rapid searching of density peaks. This algorithm requires objective function without iterative optimization and reduced quantity of parameters, resulting in facile yet effective clustering. However, this algorithm cannot automatically determine the local density of samples based on data size; the allocation process is easy to produce allocation errors and subsequent problems, resulting in a poor final clustering effect. Aiming at solving the shortcomings of density peak clustering (DPC) algorithm in local density and allocation strategy, this paper proposes a density peak clustering algorithm based on kernel density estimation and minimum spanning tree (MST). The proposed DPC algorithm adopts the Gaussian kernel density to estimate the local density of samples, coordinating the relationship between the part and the whole; Proposing a new allocation strategy, which combines the idea of minimum spanning tree to generate a tree from the data set according to the principle of high density and close distance; the degree of polymerization is defined and calculated before and after disconnecting one edge of the tree, and the edge is disconnected with the larger degree of polymerization, which until the expected number of clusters is met. The experimental results make known that the proposed algorithm has better clustering result.
Keywords: density peak clustering algorithm; minimum spanning tree; local density; degree of polymerization.
Elite subgroup guided particle swarm optimization algorithm with multi-strategy adaptive learning
by Runxiu Wu, Lulu Wang, Shuixiu Wu, Hui Sun
Abstract: In order to address the premature convergence of the standard particle swarm optimization (PSO) algorithm, this paper proposes an elite subgroup guided particle swarm optimization algorithm with multi-strategy adaptive learning (EGAPSO). In order to enhance the particles ability to escape from the local extremum point, the social cognitive part of originally learning only from the global optimal particle is changed to the part of adaptively choosing to learn from the global optimal particle and the particle in the elite subgroup in the model. Meanwhile, in order to make the algorithm more universal, a variety of learning strategies such as elite particle opposition-based learning, subspace Gaussian learning and mean center learning with different search characteristics are adaptively selected in the evolutionary process. Combination of the two improved measures can not only increase the universality of the algorithm, but also enhance the diversity of the population, which effectively helps the algorithm escape from the local optimum and avoid the premature convergence. Simulation results on the typical test function set and test results of comparison with other classical and newly improved PSO algorithms show that the proposed algorithm performs better in optimization and stability.
Keywords: PSO algorithm; elite subgroup; subspace; opposition-based learning; mean center.
Agent-based modeling approach for Internet rumor propagation and its empirical study
by Tinggui Chen, Bailu Jing, Jingtao Rong, Zepeng Wang
Abstract: This paper analyzes the current research status of Internet rumor propagation at home and abroad, proposes an expanded SPNR rumor propagation model based on SIR infectious disease model, carries out simulation analysis with agent method, and uses two methods of data set verification and case verification to verify the SPNR model.Firstly, the infection status of rumor spreaders is divided into positive and negative, and the positive and negative infection rates are dynamically set according to the proportion of surrounding infected persons. At the same time, the forgetting mechanism is also increased, and a more applicable SPNR rumor spreading model is proposed. Secondly, the simulation of rumor spreading process is realized based on agent method, and the influence analysis of parameters affecting rumor spreading is carried out by numerical simulation method, which provides the basis for the formulation of rumor control strategy. Finally, from the perspective of data set validation and case validation, the simulation results of SinaWeibo and SPNR model are compared to verify the applicability of SPNR model in the process of rumor propagation.
Keywords: Network rumors; epidemic models; complex networks; information dissemination.
Optimization of Cost 231-Hata Model Based on Deep Learning
by Qinxia Huang, Cheng Zhang, Jing Liu, Shilin Wu
Abstract: Based on the data set of question A in the 16th "Huawei Cup" mathematical modelling competition, this paper uses deep learning algorithm to optimize the Cost 231-Hata model of wireless communication. Firstly, the feature parameters of Cost 231-Hata model are analysed, and the corresponding features are found in the data set. Secondly, two new reference features are extracted according to the geometric relationship between base station and cell location. Then, the principal component analysis is used to reduce the dimension of the data set, and six features that are highly correlated with the target are extracted from multiple reference features. Finally, as these six features are taken as the input of neural network, a wireless propagation model based on deep learning is constructed by using error back propagation algorithm. The results show that the prediction accuracy of this model is higher than that of the traditional Cost 231-Hata model.
Keywords: wireless communication; Cost 231-Hata model optimization; feature engineering; deep learning.
Simulation Research on Trajectory Tracking Control System of Manipulator Based on Fuzzy PID Control
by Ruyi Ma, Zeshen Li, Du Jiang, Juntong Yun, Ying Liu, Yibo Liu, Dongxu Bai, Gongfa Li
Abstract: In view of the influence of external environment interference, parameter changing and random noise in the process of robot arm trajectory tracking, as well as the problems of fuzzy PID control, such as fuzzy rules are not easy to set, quantization interval and parameters are affected by subjective consciousness, which leads to the problem of insufficient performance of manipulator system. In this paper, firstly, in the MATLAB simulation environment, based on the D-H parameter model, the linkage model of the four degree of freedom manipulator is established, and the homogeneous transformation matrix of the manipulator is calculated. Then, through the application of fuzzy PID control in the Simulink toolbox, the optimal parameters of the controller are found by using the proportion factor self-adjusting strategy, and the corresponding trajectory tracking and forward inverse kinematics simulation are studied. The trajectory tracking curve of the manipulator is obtained. The research method and experimental results in the paper are of great significance for the simulation of articulated manipulator.
Keywords: manipulator; trajectory tracking; fuzzy PID control; simulink.
A Novel Differential Evolution with Staged Diversity Enhancement Strategy
by Wei Li, Yafeng Sun, Ying Huang
Abstract: Differential evolution (DE) algorithm is a simple and efficient evolutionary computing technology. Although DE has achieved good results in many fields, inappropriate parameter combinations can easily lead to the problem of premature convergence. In response to this problem, this paper proposed an effective DE with staged diversity enhancement strategy (SDESDE), which can increase the diversity of the population. In the early stage of SDESDE evolutionary process, SDESDE emphasizes the balance search strategy, and use the diversity enhancement strategy to avoid getting trapped in the local optima in the middle stage. In the later stage, the faster convergence strategy is adopted. Besides, an adaptive mechanism is added to enhance the control of population diversity at different stages to close to the global optima faster and improve the efficiency of search. The proposed SDESDE algorithm is compared with four representative DE and experimental results demonstrate that the proposed algorithm not only has better performance in maintaining population diversity but also has highly competitive in overall performance.
Keywords: Differential Evolution; Staged Strategy; Diversity Enhancement; Adaptive Mechanism.
A hybrid firefly algorithm based on modified neighbourhood attraction
by Rongfang Chen, Jun Tang
Abstract: Some studies reported that firefly algorithm (FA) had high computational time complexity. To tackle this problem, different attraction models were designed including random attraction, probabilistic attraction, and neighbourhood attraction. This paper concentrates on improving the efficiency of neighbourhood attraction. Then, a hybrid FA based on modified neighbourhood attraction (called HMNaFA) is proposed. In our new approach, the best solution selected from the current neighbourhood is used for competition. If the best solution wins the competition, the current solution flies towards the best one; otherwise a new neighbourhood search is employed to produce high quality solutions. Experiments are validated on several classical problems. Simulation results show HMNaFA surpasses FA with neighbourhood attraction and several other FA algorithms.
Keywords: Firefly algorithm; modified neighbourhood attraction; generalized opposition-based learning; neighbourhood search.
A genetic algorithm based robust approach for type-II U-shaped Assembly line balancing problem
by Guangyue Jia, Honghui Zhan, Yunfang Peng
Abstract: U-shaped assembly lines are widely used to implement Just-in-time manufacturing. U-shaped assembly line balancing problem is important to improve productivity. Most of research ignore uncertainty such as operation times. This study applies robust optimization method to deal with type-II U-shaped assembly line balancing problem (UALBP-2) under uncertainty. A mathematical programming model is proposed with interval task operation times, and a genetic algorithm is developed to deal with it. A robust solution is defined as the most frequent solution falling within a pre-specified percentage of the optimal solution for different sets of scenarios. The experimental results are compared with the expected solution to verify the feasibility and effectiveness of the robust method.
Keywords: U-shaped assembly line; mathematical programming; robust solution; genetic algorithm.
|
Articles marked with this Open Access icon are freely available and openly accessible to all without any restriction except the ones stated in their respective CC licenses.
Our Blog 
Follow us on Twitter 
Visit us on Facebook