Research on the correction method of irradiance forecast based on the characteristics of the Western Pacific Subtropical High

With the rapid expansion of the photovoltaic power generation industry, accurate irradiance forecasting is crucial. Meanwhile, the Western Pacific Subtropical High has a significant impact on weather in China. This study focuses on the influence of subtropical high on irradiance prediction errors. This article uses data from a certain region in southern China from 2020 to 2023 to find that the characteristics of subtropical high pressure are closely related to the bias of irradiance prediction, especially the area and intensity of subtropical high pressure. On this basis, numerical models are used to simulate and forecast the subtropical high, and the results show that its performance improved as the start time approached. Based on the historical prediction bias situation, this article constructs a bias correction model based on Light-GBM. Data analysis and case studies show that the model effectively improves the accuracy of irradiance prediction. Future research can focus on broader data collection and more advanced algorithm improvements.