In the context of the increasing maturity of on-orbit service and the growing application requirements, taking the on-orbit servicing spacecraft as research object, the satellite-oriented on-orbit servicing mission planning problem was studied, and the way how to rationally arrange and deploy on-orbit servicing resources was discussed. The problem was decomposed into two sub-problems, namely on-orbit service resource allocation and on-orbit service path planning, and the corresponding bi-level mathematical models were established. An on-orbit service mission planning method was designed, including chaotic genetic algorithm based on multi-group parallel evolution and NSGA-Ⅱ+GSDE algorithm based on global coordinate transformation. The feasibility and effectiveness of the algorithm were verified through comparative analysis of simulation results.