运动动力学约束下基于自适应参数的运动规划方法

Stable sparse RRT(SST) is a sampling-based asymptotically optimal motion planning algorithm. Compared with the traditional asymptotically optimal algorithm RRT*, the SST employs random forward propagation to generate new nodes, without solving the two-point boundary value problem(BVP), and can directly plan a feasible trajectory that satisfies the system's kinodynamic constraints. Considering the issues associated with SST's sensitivity to parameters and challenges in adapting to complex and dynamic environments, an improved SST algorithm with adaptive parameters(ASST) is proposed. By utilizing known information such as node collision rate and node density during the planning process, the environmental area and neighborhood information of the node are estimated, and then the node selection radius and node pruning radius are adaptively changed. Simulation experiments have evaluated various types of system dynamics and complex environments, and the experimental results show that the proposed algorithm can reduce the dependence on parameters, improve the success rate and computational efficiency in complex environments, and have strong adaptability to different motion planning problems.