Reliability Evaluation Method of Active Distribution Network Based on Optimization Model

Among the existing reliability assessment methods of active distribution networks, a simulation method is difficult to obtain exact index calculation results, and an analytical method is difficult to deal with the optimal load shedding problem of dynamic islands. In this paper, a reliability evaluation method of active distribution networks based on an optimization model is proposed. The method takes into account the randomness and correlation between photovoltaic and load, fault transfer and the limitation of power flow back transmission. Firstly, a multi-state model accounting for correlation of photovoltaic and load is established, and random scenarios generated from the multi-state model, are clustered based on k-means clustering algorithm. For a specific scenario, a mixed integer linear programming model for reliability index evaluation is proposed. By solving the models independently defined for all the clustered scenario, various reliability indexes of the system can be obtained by weighting the independent results with their scenario probabilities. Case studies are conducted to validate the method and reveal its merits against traditional reliability evaluation methods. The proposed method is valuable in compatibly modelling system reliability considerations into an optimization framework of system planning.