Path Planning of Inspection Robot Based on Improved Intelligent Water Drop Algorithm

Abstract

In response to the limitations observed in the Intelligent Water Droplet (IWD) algorithm for path planning, including its weak problem-solving capability and susceptibility to local optimization, this paper presents an enhanced Intelligent Water Droplet algorithm. The improved algorithm incorporates the distance factor between nodes and the target point into the original algorithm’s path probability, enhancing its problem-solving prowess and expediting algorithmic convergence. Simultaneously, a roulette-based probability selection method is introduced to circumvent local optimization during the solution process. Additionally, the algorithm is coupled with the Floyd algorithm to refine the planned path, reducing the number of inflection points to align with the motion characteristics of the inspection robot. Simulation results underscore the effectiveness of the enhanced Intelligent Water Droplet algorithm in mitigating the impact of local optimization in path planning. In comparison to the original IWD algorithm, the optimized path exhibits a 17.42% reduction in length, a 58.3% decrease in the minimum number of iterations required for path convergence, and a 36.3% reduction in the number of path inflection points. Furthermore, the total path length is decreased by 7.7% following path optimization via the Floyd algorithm.

Keywords

• Path planning
• intelligent water drop algorithm
• local optimality
• Floyd