Traditional methods based on particle filter for mobile robot SLAM(simultaneous localization and mapping)always induce particles degradation. Focusing on the particles degradation of the traditional particle filter and the need of a large number of particles to improve the precision of robot location，the AFAS(artificial fishing-swarm algorithm) is introduced into the particle filter method. This method updates the particle’s prediction again basing on the AFSA which adjusts the particle distribution concentrate around the robot’s true pose and improve the accuracy of SLAM. Through the MATLAB simulation，the results show that the method can locate the robot quickly and accurately，and improve the mapping precision.