In order to effectively handle the challenges of safe operation brought by the intermittent power supply

virtual micro-network

flexible load and other uncertain variables in the active distribution network(ADN)

a two-level optimal scheduling model based on probabilistic power flow is proposed

which is constituted by day-head and real-time optimal scheduling model. The day-ahead scheduling determines the operation plan of each unit to minimum the operation cost of ADN based on the power forecasting results. Real-time scheduling is performed for the volatility of each unit in the actual operation and the results of the short-term forecasting. The plan made by day-ahead scheduling is adjusted in accordance to the probabilistic power flow

which is calculated regarding the uncertain units as random variables. Node voltage and branch power in ADN still satisfy the constraint condition when fluctuating in triple standard deviation. Therefore

the safety margin of ADN is improved. The gravitational search algorithm and particle swarm optimization algorithm are applied to solve the scheduling model. In the end

the effectiveness of the scheduling model is verified by case study.