Mobile robot localization based on particle filter in the environment topology map

For mobile robot localization problem, this paper presents a localization method based on particle filter in the environment topology map. First, a environment topology map was built. The method is based on SOM (Self-organizing Map), by continuously adding new neurons to achieve the growth of network, in order to describe the environmental characteristics. Then, in the environment topology map, particle filter was used for mobile robot localization, according to the topology map of the distribution of neurons in the particle distribution, by increasing the number of neurons to increase the number of particles. Simulation results show the effectiveness of the method.