摘 要 本研究針對TSB07301C無刷交流伺服馬達搭配滾珠螺桿進行定位控制,並以光學尺偵測作為回授定位控制。在傳統上求PID參數多以試誤法、這需要經驗值等,或透過轉移函數由Routh-Hurwitz求出系統穩定時之PID參數,這些方法不但耗時且非為最佳。因PID參數好壞將影響到平台定位的精準度及系統性能響應,故本論文利用基因演算法搜尋PID參數之最佳解於單軸平台定位控制。 本系統以100mm為定位距離,並設定轉速500000Pulse/sec於2.008sec達到反覆定位精度誤差小於0.002mm 的定位目標。應用基因演算法,經過基因揀選複製、交配與突變等程序尋找最佳控制器參數,根據需要訂定一個性能指標來表示系統性能的優劣,如此便可藉著基因演算法來尋找最佳解的解決方法以改善原PID 控制器對閉迴路系統的性能表現。 經由基因演算法求得K P:16.0001、Ki:0.0000064 以及K d:38.1001並藉由MATLAB/Simulink軟體進行模擬所設計控制系統達到高精度的定位控制。並由實驗驗證系統結果顯示,利用基因演算法所設計之控制器,能有效達到系統在實際控制上有更好的時間反應及穩定性於定位控制上。
Abstract The aim of this study was to determine the positioning control of TSB0730IC brushless AC servo motor with ball screw, and the optical detection of position control feedback. In the traditional technique, try and error method was used to find out the PID parameters, and this required experiences. PID parameters could also be obtained by using the transfer function according to the Routh-Hurwitz stability of the PID system parameters. Both methods were not the best choice and also time-consuming, and the reasons for this were PID parameters would affect the accuracy of the positioning platform and system performance. This paper showed how the genetic algorithms was utilized in searching for the optimal PID parameters in the single-axis positioning control platform. The system for positioning was set at a distance of 100mm and speed at 500000 Pulse / sec that reached a positioning accuracy of 0.002mm within the standard error. By applying genetic algorithms and working through genetic selectioin and replication, the process of mating and evolution would come out with the most suitable controller parameters, based on the need to set a index to indicate the advantages and disadvantages of the perfomance of a system, therefore genetic algorithms could be used to find the optimal solution for improving the PID controller of a closed-loop system performance. Using genetic algorithm a series of value was obtained: KP: 16.0001, Ki: 0.0000064 and Kd: 38.1001, then MATLAB / Simulink software was performed to simulate the control system designed to achieve high-precision positioning control. The results displayed by the experimental verification showing that the use of genetic algorithm designed controller can effectively achieve system control in real time for better response and stability in the positioning control.