Prediction of stable power generation period of photovoltaic power station based on time series ARIMA model

Abstract This paper presents an algorithm based on big data and artificial intelligence technology to predict the stable generation period of photovoltaic power stations. Due to the influence of solar radiation fluctuations at sunrise and sunset, the power generation status frequently toggles between states, resulting in significant data variability. The algorithm integrates deep learning models with data from weather stations and forecasts to evaluate power generation stability between sunrise and sunset, analyzing the causes of any instability. Experiments show that the proposed algorithm is superior to the traditional algorithm in prediction accuracy and calculation speed, and is suitable for practical industrial applications. It can automatically find the right time point, eliminate the unstable periods of AC output, and reduce energy waste. At the same time, the preliminary auxiliary analysis of the data is carried out to determine the stable work, weather reasons, or fault problems. It can enhance the operational stability of photovoltaic power stations and increase the reliability of power grids, holding significant practical application value.