A Computational Time Analysis of Dhouib-Matrix-SPP versus Particle Swarm Optimization Metaheuristics for Grid-based Path Planning

Actually, path planning is one of the most fundamental aspects of mobile robots study. The objective is to determine the shortest feasible trajectory from a starting point to a goal location while avoiding obstacles. Particle Swarm Optimization (PSO) has been widely applied to this problem. However, it is often complex, requiring careful parameter tuning and extensive computational resources, in spite of that it suffers from high computational complexity, sensitivity to parameter tuning, and local optima stagnation. To overcome these limitations, the new Dhouib-Matrix-SPP (DM-SPP) method is proposed, which is rapid, straightforward, and does not require parameter adjustment. Simulation experiments on four case studies (I-shaped, U-shaped, T-shaped and Randomly shaped) demonstrate that DM-SPP consistently outperforms the ranking Particle Swarm Optimization (rPSO) metaheuristic and the artificial potential field-based Particle Swarm Optimization (apfrPSO) metaheuristic in terms of computational time: DM-SPP is 66 time rapider than the rPSO metaheuristic and 31 time rapider than the apfrPSO metaheuristic. These findings indicate that DM-SPP is a powerful and scalable approach for mobile robot path planning.