本論文主要提出兩種機械手臂必要的軌跡規劃方法：(1) 關節空間的軌跡規劃以及(2) 工作空間的軌跡規劃。在關節空間的軌跡規劃上，本論文提出一套具有啟動與結束速度限制的多軸梯形速度曲線之同步與串聯方法，當機械手臂各軸馬達具有不同的最大速度與最大加速度之限制下，所提出的方法可以使各軸馬達達到速度與加速度的同步，減少末端點軌跡的移動範圍，降低末端點軌跡的不確定性。在多個路徑點的串聯上，則提出一套串聯梯形速度曲線的方法，使機械手臂可以不必在每個路徑點下停止，提升機械手臂通過連續路徑點的效率；在工作空間的軌跡規劃上，本論文提出一套在三維空間中基於參數式曲線的軌跡規劃方法，可使用三次NURBS曲線對輸入的路徑點進行軌跡插補，並依據路徑點的弦長將曲線重新參數化，使產生的軌跡具備位置、速度與加速度的連續性。最後再利用軌跡的切線與曲率強度進行分段式速度規劃，此方式可讓機械手臂有效率的在限制的移動速度與加速度下完成軌跡，並可避免泰勒展開插補方法所造成的近似誤差。在實驗結果方面，本論文利用自行開發的六軸機械手臂實現這兩種軌跡規劃方法，在模擬與實驗結果中皆可以看出，本論文提出的方法確實可以有效提升機械手臂於任務中的效率與精準度。 In this dissertation, two trajectory planning methods in the joint space and the work space are proposed for robot manipulators. In the joint space trajectory planning, a multi-axis synchronized and series connection method is proposed based on trapezoidal velocity profiles with an initial velocity and a finial velocity. The proposed method can let multi-axis synchronous work even each motor has different limit of a maximum velocity and a maximum acceleration and it can improve unpredictable situation of the end-effector when the robot manipulator is in the moving status. Moreover, the proposed method can promote the efficient when the robot manipulator needs to pass through continuous via-point. In the work space trajectory planning, a three dimensional trajectory planning method based on a parametric curve is proposed. Three-degree NURBS curve is applied to interpolate a set of via-point and it can reparameterize by a chord length. According to the tangent and curvature strength, the proposed method can generate a trajectory which is continuities in the velocity and acceleration. Finally, in order to avoid the approximation error by using the Taylor series interpolation method and promote the performance under limited velocity and acceleration, a segmented velocity planning method is also implemented. In the experimental part, the proposed trajectory planning methods are implemented in a six-DOF robot manipulator which is own designed and development. From the experimental results, the proposed methods can effectively improve the efficiency and accuracy of the trajectory planning for the robot manipulator.
