Composite Inertia Weighted APSO for Autonomous Vehicle Path Planning

초록

Efficient path planning is essential for autonomous vehicle navigation. This paper proposes a Composite Inertia Weighted Adaptive Particle Swarm Optimization (APSO) algorithm to improve convergence reliability and solution quality. The proposed method introduces a non linear inertia weight update mechanism that combines exponential decay with a quadratic distance based component to dynamically balance global exploration and local exploitation. This adaptive strategy enables the algorithm to effectively navigate complex environments and generate collision free, smooth, and efficient trajectories. The performance of the proposed algorithm is evaluated using benchmark functions and autonomous path planning simulations. Results demonstrate that the composite approach significantly outperforms traditional metaheuristics, achieving a reduction in path length of up to 6.05% over standard PSO, 8.90% over ABC, and 2.48% over Dijkstra. Furthermore, the algorithm increases convergence speed by 34.9% and 20.8% compared to PSO and Dijkstra, respectively, proving its efficacy for high-dimensional, real-time optimization in complex environments. © 2020 IEEE.

키워드