Title: Recognising the kind of cloud using a new fuzzy knowledge-based system

Authors: Mehdi Neshat; Mohhammad Ahmadi

Addresses: Department of Computer Science, College of Software Engineering, Shirvan Branch, Islamic Azad University, Shirvan, Iran ' The Advanced Research Center of Hydro-Meteorological, Tehran, Iran

Abstract: Nowadays, expert systems play a major role in better doing of complex tasks and giving advice to the experts because expertism is a specialised knowledge. Overall, expert systems are used to solve the problems for which there is not an accurate knowledge and a particular algorithm. Understanding the atmospheric phenomena and their role in human life are the most important and affecting issues in human societies. In meteorology, it is important to identify the type of clouds. By monitoring from the Earth's surface (seeing bottom view of the cloud) and using satellites (seeing top view of the cloud), we can identify the variety of clouds. A fuzzy inference system with the specialists' knowledge of meteorology is designed in this paper and its aims are detection of the cloud type through extracting knowledge from satellite images of the cloud upper portions. The used data are extracted from the reputable website of UCI called cloud dataset. This dataset is gathered by Philip Collard in two ranges of IR and VISIBLE. Using the experts' knowledge, this system determines the type of cloud with an accuracy level of 88.25% ± 0.5 and according to experts' opinion; the results are suitable and acceptable.

Keywords: cloud datasets; fuzzy logic; fuzzy expert systems; clouds; fuzzy KBS; knowledge-based systems; meteorology; cloud types; fuzzy inference systems; FIS; knowledge extraction; satellite images.

DOI: 10.1504/IJRIS.2016.082989

International Journal of Reasoning-based Intelligent Systems, 2016 Vol.8 No.3/4, pp.168 - 180

Received: 28 Jan 2016
Accepted: 18 Sep 2016

Published online: 17 Mar 2017 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article