- 學位論文
多軸工具機智能化控制之研究
A study on intelligent control of multi-axis machine tools
摘要
CNC工具機之參數設定應隨著慣性及摩擦效應而改變，但是現有的機台功能僅允許設定單一一組參數來對應不同的切削應用，本研究首先探討慣性及摩擦效應對機台精度控制的影響，並研究自動調校參數之方法，其將可在加工條件改變時，依據當時加工狀態，於線上即時調整控制參數，使機台隨時可得到良好的加工精度。 在探討慣性及摩擦效應對機台精度控制的影響方面，本研究以模擬方式分析加工工件之重量與上述兩種效應的關係，再以Heidenhain KGM Grid Encoder於實機測試模擬的正確性。為了達到線上自動調整參數之功能，研究中以實驗分析法建立工件重量與伺服控制參數之特性曲線，以作為一重量調整參數之依據。使系統達到完全自動化且無需中斷製程上下工件，研究中亦需開發一套線上重量即時鑑識系統，可於換刀程序間偵測工件的現時重量。最後，為讓所發展的系統可與CNC控制器整合，利用Fanuc控制器之Macro Executor功能，將上述功能撰寫成G指令並植入CNC控制器中，達到線上自動調整控制參數之目的。研究中也進行了實驗驗證，確認所發展的系統確實可提升機台的加工精度。
並列摘要
To obtain optimal control performance, the control parameters of a CNC machine should be changed based on different inertia and friction effects of the controlled system. However, current commercial CNC controllers can only use one set of control parameters for all applications which was initially set when machine was calibrated. In this study, the influence of inertia and friction effects to the accuracy of servo control was first investigated. Subsequently, the auto-tuning method that can on-line tune the control parameters according to change of machining condition (workpiece weight) was developed to ensure the machine can always maintain good machining accuracy Simulation was made to analyze the correlation between the two effects and the workpiece weight. Experiments with Heidenhain KGM Grid Encoder were conducted to verify the simulation results. The characteristics curve that describes the correlation between control parameters and workpiece weight was built with empirical method. The curve was used to determine the parameters when the workpiece weight was known. A weight auto-monitoring system was developed so that the updated weight of workpiece can be identified on-line for the parameter tuning. Finally, G-codes for the proposed parameter auto-tuning function and weight monitoring function were created and embedded in the CNC controller by using Macro Executor of the Fanuc controller. The G-codes enable that the parameter tuning function can be executed on-line. Experimental results have shown that the developed system can effectively improve the accuracy of machine tool.