A Heuristic Path Planning Algorithm for Inspection Robots

Abstract

In this paper a novel heuristic path planning algorithm based on cross-entropy optimization is proposed for substation inspection robots. We seek to provide a low-cost, fast and effective solution for such robots. The optimization problem is modeled based on a topology map of the substation, it turns out to be non-convex yet requires real-time solution. A classification algorithm is applied to simplify the task points to reduce the computational cost and to speed up the path planning. The cross-entropy optimization algorithm is used to derive the shortest path among the simplified task points. Then, a path resolution algorithm is implemented to resolve the simplified path to the original task space and eventually get the real inspection path. Simulative and experimental studies have verified feasibility and effectiveness of the proposed method. Particularly, the algorithm implemented on an embedded computer delivers a solution for 153 task points in only three seconds, while very low repetition rate (about 0.2%) is approached.