Prediction of Short-Term Solar Irradiance Using the ProbSparse Attention Mechanism for a Sustainable Energy Development Strategy

Abstract

Sustainability refers to a development approach that meets the needs of the present generation without compromising the ability of future generations to meet their own needs. Solar energy is an inexhaustible and renewable resource. From the perspective of resource utilization, solar power generation has a high degree of sustainability. Therefore, solar power generation is one of the most important ways to transform the energy structure and promote the sustainable development of the economy and society, and it is of great significance for promoting the construction of a resource-conserving and environmentally friendly society. However, solar energy resources also exhibit strong unpredictability; therefore, this paper proposes a novel artificial intelligence (AI) model for short-term solar irradiance prediction in photovoltaic power generation. Leveraging the ProbSparse attention mechanism within an encoder-decoder architecture, the AI model efficiently captures both short- and long-term dependencies in the input sequence. The dingo algorithm is innovatively redesigned to optimize the hyperparameters of the proposed AI model, enhancing model convergence. Data preprocessing involves feature selection based on mutual information, multiple imputations for data cleaning, and median filtering. Evaluation metrics include the mean absolute error (MAE), root mean square error (RMSE), and coefficient of determination (R2). The proposed AI model demonstrates improved efficiency and robust performance in solar irradiance prediction, contributing to advancements in energy management for electrical power and energy systems.