Short-term load forecasting based on improved manta ray algorithm to optimize neural network

Abstract

Accurate short-term load forecasting can reduce the shutdown reserve and rotating reserve of generator units, which directly affects the safety, stability and economic benefits of power grid. Aiming at the shortcomings of sensitive initial value and easy to fall into local optimization of BP neural network, a load forecasting method based on improved manta ray algorithm to optimize BP neural network is proposed. The location update formula of manta ray foraging is improved through Levy flight strategy, the local and global search performance of the algorithm is balanced, its ability to jump out of the local optimization is enhanced, the weight and threshold of BP network are optimized, and the optimization model for short-term load forecasting is established. The results show that the prediction accuracy and convergence speed of the improved prediction model have been greatly improved. It has engineering practical value.