To achieve high-precision and highly robust path planning for mobile robots in complex scenarios, this paper proposes a hybrid path planning method that integrates an improved Theta* algorithm with the dynamic window approach (DWA). Firstly, the global path planning Theta* algorithm is optimized by introducing safety distance constraints, an adaptive neighborhood expansion strategy, improved line-of-sight checks, and B-spline curve smoothing, thereby enhancing path safety and smoothness. Secondly, the local path planning DWA algorithm is improved by incorporating a fuzzy logic control-based dynamic parameter matching mechanism, velocity change curvature constraints, and first-order low-pass filter smoothing optimization to suppress velocity fluctuations. Finally, the two algorithms are deeply integrated within the robot operating system(ROS) move_base framework to achieve synergy between global and local planning. Simulation results show that, compared to the traditional Theta* algorithm, the optimized version increases the minimum safety distance by 10.36%, reduces the number of path nodes by 27.27%, and optimizes the path length by 2.43%. Compared to the traditional DWA, the improved version enhances path smoothness by 77.78%, reduces high-frequency variance by 90.91%, and decreases planning time by 3.12%. Real-world experiments further verify that the integrated algorithm effectively maintains safety distance and ensures motion smoothness in environments with both static and dynamic obstacles.