In order to solve the problem of optimal path planning of Mecanum wheels automated guided vehicle(AGV) in a narrow and complex working environment

an improved A* algorithm based on the kinematic model of Mecanum wheels chassis was proposed. Firstly

the Mecanum wheels AGV was equivalent to a two-dimensional minimum bounding rectangle

and its omni-directional movement characteristics were used to design the path search strategy. Secondly

in order to improve the safety of the planned path

an extended model obstacle avoidance matrix was constructed according to the characteristics of the model. Finally

two-dimensional Gaussian kernel function was introduced to adaptively adjust the weight coefficients of the actual cost function and heuristic estimation cost function of the algorithm

so as to balance the global and fast search. The simulation results demonstrate that the improved algorithm is better than the common algorithm in search time and safety performance

and improves the ability of Mecanum wheels AGV to pass through narrow space or turn dead corner

and enhances the path search efficiency.