具有振動平台基座機械手臂之力量卅位置控制

本研究首先利用Lagrange能量的觀念，分析機械手臂結構間幾何關係與運動狀態，因而建立具有振動平台基座機械手臂之動態方程式，透過Matlab的視覺化建模環境(Simulink)中S-函式求得此動態方程式之多階非線性項，進一步的將機械手臂動態系統利用Jacobian方式將接點座標轉為笛卡爾座標系由此作位置控制，系統中也利用Pseudo-inverse解決非一般型之反矩陣運算。本研究利用圖形化的建模方式來下達控制命令，來模擬機械手臂尾端定位之控制，進而運用在即時控制上。欲知完整的建立機械手臂動態模組實屬不易，由於機械手臂運動過程中，常需要面對機械手臂之馬達與外界的干擾等，造成參數變化的問題，然而這些數據往往不易取得，更提高了機械手臂控制上的難度，因此將機構設計於振動平台上，且在平台上架設振動感測器，利用振動感測器偵測平台上的振動量，且進一步利用振動產生器產生力量去抑制振動。力量控制方面則利用阻抗控制觀念，規劃機械手臂與環境間接觸力間之響應，有鑑於此，本論文首先利用工業界最為常用的PID控制法則，並利用灰關聯來調整PID控制器於機械手臂尾端定位之最佳控制參數值；另外，本研究亦利用PID控制器最佳化參數之軌跡誤差、誤差微分與輸出關係透過ANFIS(adaptive network fuzzy inference system)訓練觀察是否將誤差再降低，此部份研究PID與ANFIS兩種控制策略，應用於機械手臂於振動平台上之力量卅位置之控制，研究結論可應用於相關機械手臂系統之力量卅位置複合控制設計之用。

關鍵字

機械手臂/振動平台/灰關聯/Jacobian/Pseudo-inverse/PID/ANFIS

並列摘要

Complex robotic systems have recently garnered increased attention in many fields. An example of such a system is a dexterous manipulator mounted on a flexible base. In literature, such systems are known as macro-micro systems, characterized by the number of control actuators than state variables. The dynamic equation of the robot manipulators mounted on oscillatory bases by using Lagrange formulation is proposed in this research. The equation for robot system motion in the joint space can then be represented as Cartesian space coordinates based on the Jacobian matrix. Under the assumption that the manipulator is kinematically redundant, the pseudo-inverse matrix was introduced in the system. Therefore, this work proposes pseudo-inverse Jacobian feedback control laws and applies grey relational analysis for tuning outer-loop PID control parameters of Cartesian computed- torque control law for robotic manipulators mounted on oscillatory bases. Experiments are conducted to confirm the efficiency of the proposed methodology. As for the perspective, this study primary uses PID control rule, which commonly used in industry, and adjusts the optimum control parameter value of localization of the mechanical arm end of the PID controller by using the ash connection. Moreover, this research also uses error path of the optimum parameter, the erroneous differential and the output relations of the PID controller by ANFIS (adaptive network fuzzy inference system) to observe whether the inaccuracy have been reduced or not. This part is about putting these two kinds of control schemes, PID and the ANFIS, into controlling the force/position of the mechanical arm on the oscillatory bases. The conclusion of research may apply in the design of compound control of the force/position that related to mechanical arm system.