基于虚拟势场法的全局收敛视觉路径规划

作为一种局部路径规划策略, 虚拟势场法由于其简单易用, 效果良好而在机器人领域得到了广泛的应用. 但是这种方法的缺点是在路径规划的过程中, 机器人经常会陷入局部极小. 本文提出了一种基于虚拟势场法的视觉路径规划策略, 可以保证图像中的特征点在伺服过程中始终保持在摄像机的视野之内, 并且通过稳定性分析证明了该策略具有全局收敛性. 另外, 本文进一步探讨了视觉伺服中另一个常见问题: 如何在规划过程中得到较好的三维运动轨迹. 最后用仿真结果验证了本文所提出的路径规划策略具有良好的性能.

A PFM-based Global Convergence Visual Servo Path Planner

As a classical local path planning method,potential field method (PFM) is used widely in the robotics field because of its simplicity and elegance.However,a main drawback of this method is the existence of local minima when generating a path for a robot to follow.In this paper,we present a path planning approach for visual servo by using PFM method to keep the features within the camera field of view (FOV).A rigorous analysis is then presented to prove the global stability of the constructed path planning method.Moreover,the problem of how to obtain a better three-dimensional (3D) camera path is also studied extensively. Simulation results are provided to verify the performance of the proposed path planner.