Abstract: Photovoltaic power generation is vulnerable to environmental factors such as temperature and irradiance. In recent years, haze (with high concentration of PM_2.5) has caused serious pollution, greatly reducing the power generation of the photovoltaic system. Therefore, it is of great significance for the photovoltaic market to predict the photovoltaic power generation in the haze weather. In this paper, based on the annual photovoltaic data of a household photovoltaic roof in Shanghai, the relationship between the PM_2.5 concentration and the power generation loss index is fitted and analyzed with controlled variables and the similar days for haze analysis. According to the principle of iteration, the algorithm for photovoltaic power generation prediction is optimized, and the formula for photovoltaic power generation prediction under haze is given to modify the photovoltaic revenue prediction model. The results show that the optimized algorithm can improve the accuracy and stability of the prediction results. Through the revenue analysis of three photovoltaic economic models, the iterative optimization algorithm can improve the accuracy of photovoltaic revenue forecast.