以多軸準動態特性匹配為基礎之控制器參數調整法設計

Design of Motion Parameter Tuning Methods based on the Matched Quasi-dynamics of Multiple Motion Axes

沈奕均(Yi-Jun Shen)

指導教授 ： 葉賜旭

國立臺北科技大學/機電學院/機電整合研究所/碩士(2009年)

https://doi.org/10.6841/NTUT.2009.00202

摘要

近年來，由於產業界不斷地提高品質標準，促使工具機必須同時兼顧高精密度、高速度以及高穩健性的加工能力，但往往會受到伺服驅動系統的限制，使得加工品質無法再進一步提升。於是，如何在既有的伺服驅動系統限制條件下，改善工具機的運動控制能力與加工精度，就成為一個重要的課題。
傳統上，提升多軸伺服平台系統精密度的方式是獨立設計各軸控制器，個別地降低各軸的追跡�位置誤差，但此控制概念並不一定能改善輪廓誤差，尤其是高速運動時特別明顯。然而，在調整控制器參數時，為避免運動軸受到機械破壞，或發生不預期的跳機，步階運動命令應盡量避免使用，以能承受的位置命令運動斜率為標準。因此，本文進行多軸準動態特性匹配的參數調整的研究，使同動的運動軸們動態響應能匹配，進而增加輪廓的精度。
目前現有的交流伺服控制器中，較難系統化其運動參數，僅能對控制參數加以調整。因此，本論文使用兩種參數調整方法進行測試與比較。第一種為田口品質設計實驗法，使用少量的實驗方法進行參數的最佳化組合，並使用迴歸分析加以驗證；第二種為Learning Automata自動參數調整法，可在未知的環境下，依照所設定的範圍交互選取來進行參數調整，並且其提供增加參數搜尋速度的收斂參數，使其得更有效率的調整控制參數，故本文在此也使用田口品質設計實驗法加以確定Learning Automata中的收斂參數，使其有較快的目標函數收斂速度。
在經過諸多的實驗測試與驗證後，依據本文所提出的運動控制系統參數調整與分析的研究得知，確實可獲得有效且較佳的控制參數。

關鍵字

運動控制 ； 自動參數調整 ； 準動態匹配 ； 田口方法 ； 迴歸分析

並列摘要

In recent, based on the requirements for industrial applications, CNC machine tools must provide capability of high-precision, high-speed, and high-stability motions for increasing machining quality. However, servomechanisms with driving servo packs usually limit the execution performances of motion control systems, and the machining quality of the applied CNC machine tools is thus limited during machining processes. Therefore, it is important to design motion controllers based on the conditioned servomechanisms to improve motion performances such that machining tools can provide high-quality and high-speed machining results.
Traditionally, CNC machine tools control the servomechanism of each axis independently to reduce tracking and positioning errors such that the motion accuracy of the integrated servomechanisms can be thus improved. However, especially for high-speed motions, the independent servo control could not improve the contouring accuracy of the motion of the integrated servomechanisms because of the mismatched dynamics of servomechanisms. Therefore, in order to improve contouring accuracy in general multi-axis motion systems, the control parameters of the applied servo packs should be tuned to achieve matched dynamic characteristics among all the synchronous motion axes. However, tuning the motion parameters, in order to prevent motion axis from being destroyed by the machine, the motion command should use ramp, not step. In this paper, the tuning methods are considered for achieving the matched quasi-dynamics of the applied servomechanisms.
Among present existing servomotor packs, it is difficult to design motion parameters systematically. However, tuning these motion parameters requires careful determination. Therefore, in this paper, we use two tuning methods for achieving tuning motion parameters. The one is taguchi method, which is use a small number of experiments to achieve motion parameters optimization and it uses the regression analysis to verify. The other is learaing automata method, which operates through interactions with unknown environments using a stochastic trial and error process. It also providers additional convergence information through probability destiny functions. In order to tune the motion parameters more effectively and faster, in this study, we also use taguchi method for confirming learning automata's convergence parameters.
Some simulations and experiments are executed on two axial motion systems to evaluate the proposed tuning method, and the results indicate that the proposed approaches can improve the motion accuracy of the applied servomechanisms.

並列關鍵字

Motion control ； Auto-tuning method ； Matched quasi-dynamics ； Taguchi method ； Regression analysis

參考文獻

1. Bush, R. R., and Mosteller, F., “Stochastic models for learning,” Wiley, NY, USA, 1958.

2. Chuang, H. Y., and Liu, C. H., “Cross-Coupled Adaptive Feedrate Control for Multiaxis Machine Tools,” ASME Transactions on Journal of Dynamic System, Measurement and Control, vol. 113, no. 3, 1991, pp. 451-457.

3. Chuang, H. Y., and Liu, C. H., “A model Reference Adaptive Control Strategy for Improving Contour Accuracy of Multiaxis Machine Tools,” IEEE Transactions on Industry Application, vol. 28, no. 1, 1992, pp. 221-227.

4. Her, M. G., and Weng, F. T., “A Study of the Electrical Discharge Machining of Semi-Conductor BaTiO3,” Journal of Materials Processing Technology, vol. 122, no. 1, 2002, pp. 1-5.

5. Koren, Y., “Cross-Coupled Biaxial Computer for Manufacturing Systems,” ASME Transactions on Journal of Dynamic System, Measurement and Control, vol. 102, no. 4, 1980, pp. 265-272.