A Minimum Error Increment ELM Algorithm Based on Catch Fish Optimization Algorithm for Power Load Forecasting

Power load forecasting constitutes a pivotal component within the framework of power system planning and scheduling. In conventional ELM models employed for power load forecasting, the generation of input weight matrices and bias vectors is often overlooked. This oversight can lead to pathological characteristics in the overall model, thereby introducing uncertainty with respect to the number of hidden layers. This, in turn, poses challenges to the system with regard to computational cost and forecasting performance. In this paper, we propose a minimum error increment ELM algorithm based on the fishing optimization algorithm (CFOA-MEI-ELM) to address these issues. Through experimental verification, we assess the efficacy of this novel algorithm.