Research on multi-objective path planning of a robot based on artificial potential field method

Abstract

Aiming at the shortcomings of traditional artificial potential field method, which easily falls into zero potential field in the case of complex obstacles and can't guarantee the optimal path, a multi-objective path planning based on artificial potential field method is proposed. The method takes the path length and path security as the optimisation targets, and the obstacle repulsion coefficient as the decision variable. The objective function is calculated by establishing the artificial potential field performance model, and the NSGA-II algorithm is used for multi-objective optimisation to obtain the optimal path. The flexibility of decision variables prevents the path from falling into zero potential field points, and multi-objective optimisation ensures path length and security optimisation. The simulation results prove the feasibility and superiority of the method.