Optimizing Unmanned Vehicle Navigation: A Hybrid PSO-GWO Algorithm for Efficient Route Planning

Abstract

This study aims to address the route-planning problem for autonomous systems, which plays a significant role in the operation of unmanned vehicles. A hybrid algorithm has been proposed to enhance the performance of metaheuristic algorithm approaches used to solve the specified problem. In the hybrid algorithm, the simplicity and powerful global search capabilities of the Particle Swarm Optimization (PSO) algorithm are combined with the strong exploration and local minimum avoidance features of the Grey Wolf Optimization (GWO) algorithm. The proposed hybrid approach seeks to achieve both computational accuracy and efficiency in processing time. Using the hybrid approach, routes were calculated in an unknown environment with the help of sensors. The performance of the hybrid algorithm was compared with that of the standalone PSO and GWO algorithms. The comparison evaluated the algorithms based on their execution time for finding the optimal route, the length of the calculated route, the required number of iterations, and their ability to escape local minima. The results were simulated using a custom-built interface, demonstrating a significant advantage in terms of route calculation time. Furthermore, the local minimum problem inherent in the PSO approach was successfully mitigated, while the iteration count and processing time were improved compared to the GWO approach. This approach can be particularly beneficial in disaster management scenarios, where autonomous unmanned vehicles can assist in efficiently planning routes for search, rescue, and resource delivery in unknown or obstructed environments.