Title: Multidimensional meteorological data analysis based on machine learning

Authors: Jianxin Wang; Geng Li

Addresses: China Meteorological Administration (CMA), Henan Key Laboratory of Agrometeorological Support and Applied Technique, Zhengzhou, Henan, China; Anyang National Climatological Observatory, Anyang, Henan, China ' School of Physics and Electronic Information Engineering, Henan Polytechnic University, Jiaozuo, Henan, China

Abstract: Multidimensional meteorological data has very important application scenarios, and how to effectively analyse and use it is a challenging problem. This paper proposes a multi-dimensional meteorological data analysis method based on an improved Bayesian neural network. This paper considers the example of wind power forecasting for wind farms. The input data can be divided into two categories, which are the multidimensional meteorological data and historical data of wind power. First, the raw multidimensional meteorological data are pre-processed using Principal Component Analysis (PCA). Then, the processed meteorological data and historical wind power data are fed through the Long- and Short-Term Memory (LSTM) network to achieve data feature extraction and further data dimensionality reduction. At last, they are input to the improved Bayesian neural network to achieve data fitting. This paper selects the data of 12 wind farms in a certain region of China for simulation experiments. Our proposed method is compared with BP-Neural Network and Support Vector Machine (SVM) to evaluate its performance. The experimental results show that the method proposed in this paper has good performance.

Keywords: meteorological data; multidimensional data; Bayesian neural network; wind power forecast.

DOI: 10.1504/IJCAT.2023.132098

International Journal of Computer Applications in Technology, 2023 Vol.71 No.3, pp.244 - 250

Received: 30 Apr 2022
Received in revised form: 30 May 2022
Accepted: 09 Jun 2022
Published online: 11 Jul 2023 *

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