Abstract
In this paper, we extend the FastSLAM algorithm for fusing sensor data of laser range finder (LRF) and sonar. In the case of sensor fusion, the computation time of updating map is increased. For solving this problem, the process of updating map is moved into the resampling step of particle filter. This process is not executed for all particles, is executed only for particles living next time. The computation time is decreased by this extension. Proposed algorithm is implemented on a vehicle robot and evaluated in real environment. The experimental results show that the execution time is improved.
