Adaptive sand cat swarm optimization algorithm for unmanned ground vehicle path planning
摘要
Path planning for unmanned ground vehicle (UGV) is crucial for its autonomous operation and has gained increasing attention due to its wide range of applications. Existing UGV path planning methods suffer from poor optimization quality. To address this issue, this paper proposes a path planning method based on the Adaptive Sand Cat Swarm Optimization Algorithm (ASCSO). First, an adaptive step-size perturbation strategy is introduced to mitigate population homogenization and enhance exploration capability in early iterations, while improving convergence accuracy and solution quality in later iterations. This strategy assigns weights to individuals and employs a roulette wheel mechanism to enhance reference selection diversity. Meanwhile, the perturbation probability is dynamically adjusted, enabling selected individuals to execute partial-dimensional perturbation toward reference individuals using an adaptive Brownian motion step size. Second, a trajectory point repair strategy based on an empirical penalty mechanism is proposed. By adaptively selecting repair directions, this strategy improves the population’s search efficiency for feasible paths and further enhances solution quality by guiding the search toward high-quality feasible regions. Finally, various map models with differing complexities are constructed for experimental validation.Experiments demonstrate that the proposed ASCSO outperforms other comparison methods, with the minimum cost reduction rate of