Short-Term Wind Power Load Prediction Based on APSO-LSTM

Accurate prediction of wind power load can improve the stability and reliability of power grid. In this paper, an adaptive optimal PSO- LSTM neural network algorithm (APSO-LSTM) is proposed to accurately predict wind power load. Firstly, the whole particle swarm is divided into three subpopulations, and the particle evolution is carried out according to the properties and renewal mode of the particles in each subgroup, so as to ensure the diversity of the population. Secondly, adaptive algorithm is used to optimize the learning factor of PSO to improve the convergence of the algorithm and the quality of the optimal solution. Experiments show that the APSOLSTM neural network algorithm proposed in this paper has high precision for wind power load prediction.