平行式工具機之基因法參數優化與模糊法扭力控制

Abstract

本文針對3RPS工具機，建立計算力矩控制架構，並與傳統位置控制比較兩者在不同載重中對於連續路徑之追蹤行為。針對軌跡在高速且大曲率時所造成輪廓精度降低的問題，本文利用模糊邏輯控制建立進給速度、曲率及輪廓誤差的關係式，以替代複雜且不精準的數學推導，並結合基因演算法以系統化的方式建立系統控制參數，經由模擬和實驗結果證明，在軌跡為大曲率變化下可經由減速動作達到提升空間軌跡輪廓精度的效果。為了能夠直接量測運動平台在空間中所行走的軌跡路徑，本論文也提出架設循圓量測儀(Ballbar)的方法，藉由循圓量測儀量測實際的平台位置姿態，來比較本文所發展的軌跡修正方法，最後以3RPS平行式機構上驗證其可行性。

In this study the computed-torque control has been developed on 3RPS parallel machine tool, and compared the continuous trajectory tracking in different load with traditional position control. In order to improve the low precision of tracking in high feedrate and large curvature, the relationship of feedrate, curvature and contour error has been established with fuzzy logic control to replace the complicated and inaccurate mathematics model, and also combined with genetic algorithm to build up control parameters. The simulation and experimental results show that this new control strategy improves the performance of contouring. To measure the spatial position of the platform directly, a method is presented using ballbar. The different control methods developed in this paper can be compared by using ballbar, and test in the 3RPS parallel machine tool.