In the process of mobile robot path optimization, the traditional A * algorithm has some problems, such as poor search efficiency, lack of safety, many turning points, large turning angles and unable to achieve dynamic obstacle avoidance, etc. To solve these problems, this paper proposes a fusion algorithm which combines the safety A * algorithm with the dynamic window approach (DWA). In the global path planning, the safety estimation is introduced into the evaluation function of the traditional A * algorithm, and the heuristic search neighborhood is expanded and the search direction is simplified. Then, the secondary path optimization is carried out to delete redundant nodes and smooth the path. By using the improved dynamic window evaluation function, the safety A * algorithm and the dynamic window method are integrated to realize the dynamic obstacle avoidance of the robot along the global path. The simulation results show that the improved A * algorithm reduces the path length and the number of corners by 2. 39% and 25% on average compared with the literature algorithm, and verifies its dynamic obstacle avoidance effect in the dynamic complex environment, which can meet the actual needs of robot path planning and has certain application value.