针对在未知环境无人车受约束控制条件下的动态路径平滑规划问题,提出了一种基于支持向量机(SVM)的D*改进算法。该算法通过栅格法对环境建模,建立车辆受约束动态方程,在动态规划路径选取之后使用SVM算法对无人车转向位置处进行局部路径平滑优化。实现无人车在真实环境中流畅稳定的运动。实验结果表明:该算法能够在未知环境中规划出平滑的动态路径,具有较好的可靠性和稳定性。
For the problem of the dynamic path planning and path smoothing of automated vehicle in unknown environment with constraints, an improved algorithm of D* based on support vector machine (SVM) is presented. In this algorithm the environ- mental modeling is based on the grid method and then establish the constrained dynamic equation of the vehicle. After the dynamic planning path is selected, the SVM algorithm is used to smooth optimization the local path of automated vehicle at the steering position. By this algorithm automated vehicle can smooth and stable movement in real environment. Experimental results show that this algorithm can dynamically plan a smooth path in unknown environment, which has sufficient reliability and stability.