Path Planning Optimization of Smart Vehicle With Fast Converging Distance-Dependent PSO Algorithm

Path planning is a crucial technology and challenge in various fields, including robotics, autonomous systems, and intelligent transportation systems. The Particle Swarm Optimization (PSO) algorithm is widely used for optimization problems due to its simplicity and efficiency. However, despite its potential, PSO has notable limitations, such as slow convergence, susceptibility to local minima, and suboptimal efficiency, which restrict its application. This paper proposed a novel strategy called the Distance-Dependent Sigmoidal Inertia Weight PSO (DSI-PSO) algorithm to address slow convergence in path planning optimization. This innovative algorithm is inspired by neural network activation functions to achieve faster convergence. In DSI-PSO, each particle computes a distance metric and leverages a sigmoid function to adaptively update its inertia weight. Beyond improving convergence speed, this approach also addresses path-planning challenges in autonomous vehicles. In intelligent transportation systems, effective path planning enables smart vehicles to navigate, select optimal routes, and make informed decisions. The goal is to identify a collision-free path that satisfies key criteria such as shortest distance and smoothness. This methodology not only accelerates convergence but also maintains a balance between exploration and exploitation. The effectiveness of the DSI-PSO algorithm is tested using thirteen distinct unimodal and multimodal benchmark functions, serving as rigorous test cases. Additionally, the algorithm's realworld applicability is evaluated through a smart vehicle simulation, assessing its ability to identify safe and efficient paths while minimizing overall path length. The results demonstrate the superiority of the DSI-PSO algorithm over conventional PSO approaches, with significantly enhanced convergence rates and robust optimization performance.