Intelligent Neural Network with Parallel Salp Swarm Algorithm for Power Load Prediction

Abstract

The neural network runs slowly and lacks accuracy in power load forecasting, so it is optimized using meta-heuristic algorithm. Salp Swarm Algorithm (SSA) is a novel meta-heuristic algorithm that simulates the salp foraging process. In this paper, a parallel salp swarm algorithm (PSSA) is proposed to improve the performance of SSA. It not only improves local development capabilities, but also accelerates global exploration. Through 23 test functions, PSSA performs better than other algorithms and can effectively explore the whole search space. Finally, PSSA is used to optimize the weights as well as the thresholds of the neural network. Using the optimized neural network to predict the power load in a certain region, the results show that PSSA can better optimize the neural network and increase its prediction accuracy.