An Improved FastSLAM Using Resampling Based on Particle Fission Propagation

FastSLAM is a SLAM algorithm based on particle filter. In order to solve the problem of particle degradation caused by resampling of FastSLAM algorithm in simultaneous localization and map construction of mobile robot, A method of attracting fission resampling is proposed in this paper . The heavy particles are fission at the time of resampling, and the parameters such as relative distance and attraction radius are introduced to screen the offspring particles produced by fission, and the weights of the filtered offspring particles and the parent particles are calculated uniformly. The results of simulation show that the filtering accuracy of this method is higher and the running time is optimizing compared with particle filter and fission bootstrap particle filter.