结合启发式函数的随机运动规划方法

为有效解决位姿空间中存在狭窄通道的运动规划问题,提出一种结合启发式函数的随机运动规划方法。建立了人工势场函数,沿势场等势线方向的启发式函数避免了局部极小值问题。启发式函数与随机规划方法结合,能够引导机器人避过障碍物快速朝目标点运动。人工势场在局部极大值和最速下降方向方面的特殊性质进一步优化了算法。平面内机器人运动规划的实验表明,与原有单纯随机规划方法相比,这种结合启发式函数的随机运动规划方法在狭窄通道规划问题上性能有明显提高。

Randomized sampling-based motion planning algorithm combined with heuristics

A randomized sampling-based motion planning algorithm was combined with heuristics to solve difficult motion planning problems where the configuration space contains narrow passage.The local minimum problem is avoided since the heuristic function follows the contours of potential field.The planner combines the heuristic function and randomized sampling to guide the robot along a collision-free trajectory.The algorithm is optimized using the local maximum value of the potential and the steepest descent direction.Experiments show that the algorithm much more effectively solves planning problems with narrow passages.