Multi-Point Dynamic Path Planning for Logistics Robots Using Hybrid Ant Colony Optimization Algorithm

Aiming at the defects of traditional algorithms in multi-point dynamic path planning for logistics robots, such as slow convergence, local optimum and poor dynamic obstacle avoidance, this paper establishes a topology-grid hybrid environment model and proposes a hybrid ant colony algorithm named PHDACA. It optimizes the probability transition function, integrates 2-opt local search, hierarchical pheromone update and adaptive parameter adjustment. A multi-objective model considering path length, operation time, turning times and motion smoothness is also constructed. Combined with improved A* and dynamic window approach (DWA) algorithms, a hybrid planning scheme integrating global optimization and local obstacle avoidance is formed. Simulations on traveling salesman problem library (TSPLIB) datasets and real logistics scenarios verify that the proposed algorithm outperforms traditional and other improved algorithms in accuracy and stability. It effectively reduces redundant nodes and turns, and strengthens dynamic obstacle avoidance capability, possessing practical engineering value.