Short- Term Power Prediction Method for Photovoltaic Power Generation Based on Elman Neural Network for Aspen Swarm Optimization

• Published in: 2023 6th International Conference on Energy, Electrical and Power Engineering (CEEPE) pp. 1063 - 1067
• Main Authors: Cai, Xinlei, Zhu, Jinzhou, Liu, Jiale, Meng, Zijie, Huo, Yuhang, Yu, Yang
• Format: Conference Proceeding
• Language: English
• Published: IEEE 12.05.2023
• Subjects: aspen swarm search algorithm; Data models; Elman neural network; integrated similarity; Mathematical models; Neural networks; photovoltaic power prediction; Photovoltaic systems; Power engineering; Prediction algorithms; Predictive models
• DOI: 10.1109/CEEPE58418.2023.10166666

The volatility and randomness of photovoltaic (PV) power generation lead to the difficulty of achieving ideal PV power prediction accuracyand improving the prediction accuracy of PV power generation is an effective way to suppress the adverse effects of PV grid connection. In order to improve the prediction accuracy of Elman neural network the weights and thresholds of output-implicit implicit-undertaking and implicit-output of Elman neural network are searched for optimality by using the aspen swarm search algorithm and a short-term power prediction method of PV power generation based on the aspen swarm search algorithm optimized Elman neural network (BSAS-Elman) is proposed. Firstly a model of Elman neural network is built in MATLAB; then the power of a PV plant is predicted under three weather conditions: sunnycloudy and rainy using the PV power data of a PV plant and meteorological data as inputs;finally the prediction results of BSAS-optimized Elman neural network model are compared with the prediction results of Elman neural network and the actual power output data of the PV plant are compared. The results show that the aspen swarm search algorithm can improve the prediction accuracy of the Elman neural network model and realize the effective prediction of PV power short-term prediction