Path planning for WMRs based on multi-strategy potential field ant colony optimization and elastic dynamic window approach

To improve the path planning capability of wheeled mobile robots (WMRs) in complex environments, this paper proposes a path planning method for WMR based on multi-strategy potential field ant colony optimization (MFACO) and an elastic dynamic window approach (EDWA). In the global planning layer, an MFACO algorithm containing four improvement strategies is developed to address the problems of slow convergence, high blind exploration and easy to fall into local optimization in the ant colony algorithm (ACO). Specifically, first, a heuristic function incorporating directionality and turn penalties is introduced to guide ant movement and reduce the times of turns. Secondly, a pheromone updating method based on multi-objective optimization is presented to reduce the number of algorithm iterations. In addition, pheromone initialization and deadlock resolution strategies are constructed to effectively eliminate local optimum traps. Finally, an adaptive pheromone evaporation factor and path optimization are designed to improve the robustness of the algorithm as well as to remove the redundant path nodes. In the local planning layer, to address the problem that the fixed forward simulation time in the dynamic window approach (DWA) is difficult to generate suitable paths in any environment, an EDWA algorithm is proposed, which can dynamically adjust the forward simulation time according to the real-time environmental risk levels and obstacle density to improve the environmental adaptability of the algorithm. Subsequently, a series of comparative simulation and real experiments are carried out to verify the practical application value of the proposed path planning method, which show that the path planning method proposed has obvious advantages in terms of the number of turns, convergence speed, safety distance and environmental adaptability.