Mobile Robot Localization Method Based on Adaptive Particle Filter

摘要

This paper puts forward the mobile robot localization method based on adaptive particle filter (ADF). This method defines the particle number by designating sampling error boundary, and regulates the particle number dynamically by following the uncertain intensity of robot state. When the uncertain intensity of state space is low, ADF uses fewer particles, and when the uncertain intensity is high, ADF uses more particles. The weight of particles is updated with distance similarity function. The simulation result shows that ADF improved the systematic computational efficiency and precision of localization in mobile robot localization.