In order to avoid trapping into local minimum and further improve optimization ability of simple immune genetic algorithm, a novel mUlti-population competition-based immune genetic algorithm (MPCIGA) is presented firstly. In the proposed MPCIGA, there are three populations, two of them play the role of learners and evaluators, and they stimulate each other's surviving ability through the relative fitness; the remaining population plays the role of hall of fame and evolves according to the absolute fitness and population concentration. Finally, in order to solve the path planning in mobile robot autonomous navigation, the MPCIGA is used to finish the path search in complicated environments. Compared with basic immune genetic algorithm (BIGA), the simulation results show that the proposed MPCIGA not only solves the mobile robot path planning, but also the length of planned path is shorter, which shows the validity of the proposed MPCIGA.