In order to solve the problems of traditional A* algorithm

including low computational efficiency

excessive turning points and poor obstacle avoidance performance

a robot path planning algorithm based on the combination of improved A* algorithm and dynamic window approach was proposed. Aiming at the problem of low search efficiency of the traditional A* algorithm

a bidirectional search mechanism was introduced to reduce search nodes and improve evaluation function

improve the operation efficiency of the algorithm. Aiming at the problem of numerous redundant points

a method to judge the obstacle points of the connection combined with the key point selection strategy was proposed

which could effectively eliminate the redundant nodes on the path and optimize the global path. Finally

starting from the starting point

the key points on the optimized global path were taken as local targets

and the dynamic window approach was used for path planning in segments to ensure that the robot could avoid obstacles in real time

and finally reach the target point safely. The simulation results showed that the proposed fusion algorithm could effectively plan the path of mobile robots in complex dynamic environments.