To address the challenges of traversing numerous nodes, path redundancy, unsmooth trajectories, lack of global orientation, and insufficient obstacle avoidance safety in intelligent vehicle path planning, a new strategy is proposed that integrates the improved A* algorithm and the DWA algorithm. First, for the A* algorithm, the neighbourhood search strategy is optimised to reduce the computational burden; then the cost function is optimised and the heuristic function is improved, which effectively enhances the efficiency of path planning in complex environments; at the same time, a three-step redundant node rejection process is implemented to reduce the path transitions and improve the path smoothness. For the dynamic window method, the local path planning is fine-tuned by extending the velocity sampling range and improving the evaluation function, which significantly improves the dynamic obstacle avoidance capability. Simulation results show that the improved A* algorithm reduces the number of search nodes by 17.2% and 16.1%, the search time by 26.4% and 28.9%, the path planning time by 33% and 26%, the path length by 4.2% and 2.8%, and the number of turns by 38% and 53% compared to the traditional A* algorithm in both environments. The combination of the improved DWA algorithm achieves an efficient synergy between global and local path planning, providing a safer and more efficient path planning capability for intelligent vehicles.