Dynamic Path Planning Based on Improved Particle Filter Optimisation for Patrol Robots

Abstract

A dynamic path planning algorithm based on improved particle filter optimisation was proposed to address unexpected situations, such as dynamic uncertain obstacles and temporary change in task points, that patrol robots often encounter in power monitoring systems. The computational structure of the traditional particle filter was improved by introducing iterative convergence computing, and the sampling filtering operation was performed several times in each time step to expedite the search process of the optimal path. Furthermore, the proposed algorithm was combined with the receding optimisation strategy to update the local path in real time in the form of rolling windows to ensure the predictability of planning. The simulation results revealed that the proposed algorithm can effectively handle dynamic task scenarios and satisfy the requirements of real-time planning and security.