Research on Indoor Mobile Robot Localization Method Based on Improved MCL

This paper proposes an indoor mobile robot localization algorithm based on improved MCL. The algorithm solves the kidnapped robot problem by introducing the average weight of particles to judge whether the localization fails suddenly; introducing KLD sampling to adaptively determine the particle set threshold according to the distribution of particles in the state space to reduce the algorithm computation, and designing the adaptive resampling rule to improve the particle degradation problem and improve the a posteriori confidence level. The odometer and lidar fusion localization simulation experiment is carried out with the two-wheeled differential mobile robot as the research object. The experimental results show that the algorithm proposed in this paper can enable the indoor mobile robot to recover from global localization failure, and can adjust the number of particles online over time to improve the stability, real-time, and accuracy of localization.