论文部分内容阅读
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。
In order to effectively solve the problem of motion planning of narrow passage in pose space, a stochastic motion planning method based on heuristic function is proposed. The artificial potential field function is established. The heuristic function along the equipotential line of the potential field avoids the local minimum problem. The heuristic function in combination with the stochastic programming method can guide the robot to move quickly toward the target point avoiding the obstacle. The special nature of the artificial potential field in terms of the local maximum and the steepest descent direction further optimizes the algorithm. Experiments on plane robot motion planning show that compared with the original simple stochastic programming method, the stochastic motion planning method based on heuristic function has obvious performance improvement in narrow channel planning.