Optimization of VMD-LSTM load power forecasting model based on OOLSSA algorithm
Year: 2024
Type: article
Abstract: For short-term load power forecasting models, improving forecast accuracy has been one of the important research directions. This paper proposes an improved sparrow search algorithm based on orthogonal opposition learning (OOLSSA). The method first optimizes the VMD model parameters by the OOLSSA algorithm, then performs VMD decomposition of the original load power data, and then adjusts the LSTM model parameters based on the OOLSSA algorithm. This algorithm has a better optimization convergence effect, which can more accurately screen out model parameters, optimize the hybrid VMD-LSTM prediction model, improve model performance, and reduce human intervention. By comparing the experimental results of the improved OOLSSA-VMD-OOLSSA-LSTM model with the independent LSTM model, the VMD-LSTM model, and the VMD-SSA-LSTM model of the unimproved sparrow search algorithm, the experimental results are obtained in line with the expectations and provide a new research idea for power prediction.
DOI: 10.3934/jimo.2024197
Authors: Chengxiang Li, Mingyu Yao, Lingling Lv, Yu Wu, Ziqiang Hao
Language: English
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4405921447