A new genetic algorithm named niche pseudo-parallel genetic algorithm (NPPGA) is presented for path evolution and genetic op timization of autonomous mobile robot. The NPPGA is an effective improvement to maintain the population diversity as well for the sake of avoiding premature and strengthen parallelism of the population to accelerate the search process combined with niche genetic algorithms and pseudo-parallel genetic algorithms. The proposed approach is evaluated by robotic path optimization, which is a specific application of traveler salesman problem (TSP). Experimental results indicated that a shortest path could be obtained in the practical traveling salesman problem named "Robot tour around Pekin", and the performance conducted by NPPGA is better than simple genetic algorithm (SGA) and distributed paralell genetic algorithms (DPGA).