Field Environment Path Planning Based on Improved A* Algorithm and Dynamic-Window Method

A field environment can include an uneven ground surface, along with complex and diverse surface properties, making path planning difficult. In this study, a combination of global and local approaches was used for path planning, which considered the type of ground material, variation of the ground slope, and kinematic constraints of a vehicle. A 3D bidirectional A* algorithm for field environments was developed for global planning by designing a slope cost function, distance cost function, node expansion condition, and bidirectional search strategy, which effectively improved the planning speed. A global path distance function and curvature phase velocity evaluation function were designed for local planning by considering the kinematic constraints of a vehicle, along with dynamic obstacle avoidance, which effectively improved the probability of successful path finding, path safety, and traceability via the dynamic window method.