利用位能場模型規劃雙機械手臂之移動路徑

Abstract

本論文針對三維空間中的雙機械手臂路徑規劃問題提出一演算法。該演算法在工作空間中，利用三度空間的廣義位能場模型計算障礙物對手臂所產生之推斥力，以保證手臂前進時所經過的路徑不會與障礙物或是兩個手臂互相碰撞。由於我們的方法完全是利用工作空間中的資訊，不需透過大量計算將相關資訊轉到組態空間作處理，因此規劃速度相當快速。另外，本論文採用「主僕式架構」來處理兩個機械手臂之間的協調問題，讓僕臂在規劃路徑的過程中視主臂為障礙物並閃避之，而主臂則完全忽視僕臂之存在。電腦模擬顯示，透過交錯式的規劃方式，本論文所提出之演算法在大部分的情況之下都能讓兩個手臂規劃完成各自的路徑，所規劃出來的路徑也能夠平滑且連續。

An algorithm for path planning of two manipulators is presented in this thesis. The proposed algorithm utilizes a generalized potential model to evaluate repulsion between manipulators and obstacles in workspace, so collision avoidance of the planned path can be guaranteed. Without converting information in workspace into C-space, which is computationally expensive, the proposed path planning approach is very efficient. In this thesis, the “master-slave” architecture is adopted to deal with the coordination of two manipulators, while the master manipulator is regarded as an obstacle by the slave manipulator, the existence of the latter is completely ignored by the former. Simulation results show that by alternately planning configurations of the two manipulators, safe and spatially smooth paths can be obtained successfully almost always.