Abstract: The bidirectional regulation ability of EVs can be planned to relieve the peak-valley pressure. However, the disordered charging and discharging of EVs may bring problems such as reverse peak-regulation and variables explosion of decision-making. To solve these problems, we proposed a clustering-peak-task decomposition model based on different travel chains by using the IFCM algorithm, and solved the problem that a large proportion of electric vehicles participate in peak regulation optimization under the premise of improving the accuracy of the scheduling model. Firstly, the IFCM clustering algorithm was adopted to aggregate EVs. Secondly, to minimize the control cost, the clustering results were applied to the power system peak regulation, and the charging and discharging tasks of each cluster were obtained. Finally, an individual task assignment model was established to obtain the optimal charging and discharging strategy for a single EV. Example analysis shows that, by using IFCM algorithm to improve clustering performance, the model effectively reduces the dimension of decision-making variables for peak regulation, and ensures the economy of users on the premise of meeting the demands of peak regulation.