Electric logistics vehicles restrict their effective promotion in the field of logistics and distribution due to the limited battery capacity, long charging time and inadequate supporting facilities. To this end, based on clustering non-dominant sorting algorithm (AP-NSGA-II) to solve the problem of multi-target path optimization of electric logistics vehicles, a charging strategy is established, through the design of weighted AP clustering distribution clusters, to avoid the randomness and blindness of the initial population, at the same time, the size of distribution points within clusters reduces the running time and complexity of non-dominant sorting algorithms, and according to the distribution and distance relationship of charging stations, electric logistics vehicles to carry out part of the charging strategy. Finally, the effectiveness of the algorithm is proved by simulation experiments, and the differences between the full charge and partial charging conditions of electric logistics vehicles are compared.