A cluster-based dispatch strategy for electric vehicles is proposed

with which electric vehicles cluster by arrival time

charging time

departure time and maximum delayed charging time. The dispatch strategy is formulated for each cluster

which is capable of reducing the intraday fluctuation of distribution network active power and absorbing superfluous photovoltaic power during noon period. With user satisfaction taken into consideration

the cluster-based dispatch model can make a rational charging and discharging strategy for each individual electric vehicle based on the statistical model. Considering the uncertain factors in the distribution network

the interval number is adopted to describe the uncertainty of the charging power of electric vehicles

while interval optimization is used for power flow calculation in the distribution network. In the end

the validity of the proposed strategy and the necessity of interval optimization are verified by examples. This work is supported by National Natural Science Foundation of China (No. 51177143) and National Basic Research Program of China (973 Program) (No. 2013CB228206).