In recent years

frequent pollution weather has a significant effect on transmission process of solar radiation in the atmosphere

which makes photovoltaic power forecasting face severe challenges.To solve this problem

a soft sensor model of aerosol optical thickness(AOT) is established based on air quality monitoring data and meteorological data.Dual-band atmospheric radiative transfer model is used to calculate the irradiance

and a multi-time-scale photovoltaic power forecasting method is proposed

which is driven by day-ahead and intra-day air quality forecast information.On this basis

the power prediction software of photovoltaic power generation is developed and experimental verification is performed.The results show that

the proposed method can effectively reflect the influence of pollution weather on irradiance and realize accurate prediction of photovoltaic power

and the forecasting software has good engineering application value.