Optimized ant colony system algorithm for path planning in radiation environments

Abstract

A new ant colony optimization algorithm, PRM-ACSO, is proposed based on the PRM algorithm’s path search network and the ACS algorithm, aimed at path planning for personnel in static radiation environments. By integrating the random path network of the PRM algorithm, this approach significantly reduces the number of iterations and runtime. Simultaneously, based on the initial dose rate distribution of the radiation field, a dynamic evaporation coefficient is introduced to adjust the pheromone evaporation rate adaptively, and a dead ant path pheromone update strategy is incorporated to enhance the ability to resolve deadlock issues. Considering the unique characteristics of the radiation environment, the cumulative radiation dose is incorporated into the heuristic function to strengthen the guidance toward the goal. Furthermore, based on the number of iterations, the pheromone weight parameter α and the heuristic weight parameter β are dynamically adjusted. During each iteration, the distance between the current node and the starting point is considered to define a pseudo-random proportion coefficient, improving the convergence speed between the starting and ending points and enhancing the current node’s exploration capability concerning surrounding paths. Simulation results show that paths obtained using PRM-ACSO are more stable and converge faster, with the algorithm demonstrating more pronounced advantages in large-scale scenarios with complex obstacles