From the perspective of power suppliers in electricity market,the multi-time scheduling model is built considering the uncertainty of photovoltaic power prediction,photovoltaic is consumed and tracked by using battery and air-conditioning demand response resource.The model ensures the economy and the sufficiency of demand response resources that can consume the error of the value of photovoltaic forecast.The uncertainty of photovoltaic power is taken as chance constraint in the day-ahead scheduling model and converted into certainty equivalent condition when solved,the minimum unbalance and minimum regulation are taken as objects in the intra-day scheduling model.Based on the simulation results,considering the error of the value of photovoltaic forecast in the day-ahead scheduling can effectively improve the consumption of new energy and the benefits of supplier is verified.The higher confidence degree causes the better photovoltaic tracking.The actual daily benefit is related to the cost of intra-day scheduling.