Path planning for autonomous mobile robots in dynamic constrained environments

This manuscript considers an integrated path planning algorithm for autonomous mobile robots operating in dynamical constrained environments, which combines an enhanced A-star algorithm with an enhanced Dynamic Window Approach (DWA). In the global planning stage, the A∗ algorithm is refined through several key strategies. A redundant space mechanism is incorporated to mitigate the risk of generated paths passing too near to obstacles. Moreover, a dynamic adaptive heuristic function is developed to accelerate the node search process. The algorithm's performance is further improved by integrating a direction selection strategy alongside a method for eliminating redundant path points. For close-range obstacle avoidance and trajectory modification, enhancements are applied to DWA. The introduction of temporary target points and a deviation evaluation function effectively resolves the issue of the robot becoming trapped in local minima. Meanwhile, the weight coefficients within the comprehensive trajectory evaluation function are adaptively adjusted, leading to substantial gains in critical indicators such as total path length, computational time, and average travel speed. Simulation results verify that the proposed integrated algorithm empowers the robots to generate both feasible and reliable trajectories in complex dynamic environments, demonstrating a marked advancement in overall performance compared to baseline methods.