適應性遞迴小腦模型控制器於切換式磁阻馬達驅動系統之設計

Design of Adaptive Recurrent Cerebellar Model Articulation Controller for Switched Reluctance Motor Drive System

林桂禾(Guei-Ho Lin)

指導教授 ： 王順源

國立臺北科技大學/電資學院/電機工程系研究所/碩士(2014年)

摘要

因切換式磁阻馬達驅動系統為非線性系統，所以控制器設計具有困難度。本研究結合小腦模型控制器與高斯基底函數，來設計適應性遞迴小腦模型控制器(adaptive recurrent cerebellar model articulation controller, ARCMAC)，本研究新穎的觀點是採用遞迴神經網路之架構，使原本靜態的小腦模型控制器轉換具有動態的概念。本控制器內含遞迴小腦模型控制器和補償控制器，ARCMAC用來作為切換式磁阻馬達的速度控制器，補償控制器設計為補償理想速度控制器和RCMAC之間的近似誤差。本研究以Lyapunov定理推導出RCMAC之權重與遞迴權重更新法則以確保控制器於系統中的穩定度。因此本研究將ARCMAC之控制技術植入切換式磁阻馬達驅動系統，此控制器可克服系統非線性的能力，使得切換式磁阻馬達系統速度響應具有高性能的表現。
經由實驗結果證明，在馬達負載轉矩為1 Nm，轉速為低、中、高轉速與變轉速下，穩態之轉速誤差皆能維持在 2 rpm內，且具有良好的電磁轉矩響應。另外，並將其控制結果與無遞迴小腦模型控制器之控制結果以方均根誤差做個比較，證明此控制策略確實具有效提升系統動態響應與良好的抗外部擾動強健性。

關鍵字

適應性遞迴小腦模型控制器 ； Lyapunov定理 ； 切換式磁阻馬達

並列摘要

Due to the switched reluctance motor (SRM) drive system is nonlinear, designing a controller for this system is difficult. In this thesis, an adaptive recurrent cerebellar model articulation controller (ARCMAC) was developed for use in a direct torque control system for an SRM. A CMAC with Gaussian basis functions was used to design the ARCMAC; this method represents a novel approach to adapting a recurrent neural network structure and converting the static CMAC into a dynamic controller.
The proposed controller consists of the ARCMAC and a compensated controller. The ARCMAC was applied as a speed controller for the SRM, and the compensated controller was designed to compensate for the approximation error between an ideal speed controller and the recurrent CMAC (RCMAC).
The Lyapunov theorem was applied to derive the weight of the RCMAC and the renewal rule of recurrent weight was applied to ensure the stability of the controller in the system. The SRM drive system is integrated with ARCMAC so that it can provide a good capability to deal with nonlinear characteristics and obtain a better speed response.
The experimental results revealed that the steady state error was maintained within 2 rpm, excellent electromagnetic torque response was achieved when the system operated at low, medium, and high speeds with the load torque of the motor was 1 Nm. In addition, the root mean square error of the ARCMAC control results was compared with that of an adaptive nonrecursive CMAC. It showed that the proposed controlling strategy effectively enhances the dynamic response of an SRM system and strengthens its ability to resist external disturbances.

並列關鍵字

adaptive recurrent cerebellar model articulation controller ； Lyapunov theorem ； switched reluctance motor

參考文獻

[1] R. Krishnan, Switched reluctance motor drives modeling simulation analysis design and applications, New York: CRC Press, 2001, pp. 1-167.

[2] T. J. E. Miller, "Electronic control of switched reluctance machines," IEEE Transactions on Power Electronics, vol. 16, no. 3, 2002, pp. 222-223.

[3] N. H. Fuengwarodsakul, M. Menne, R. B. Inderka and R. W. D. Doncker, "High-dynamic four-quadrant switched reluctance drive based on DITC," IEEE Transactions on Industry Applications, vol. 41, no. 5, 2005, pp. 1232-1242.

[4] J. Liang, D. H. Lee and J. W. Ahn, "Direct instantaneous torque control of switched reluctance machines using 4-level converters," IET Electric Power Applications, vol. 3, no. 4, 2009, pp. 313-323.

[5] S. K. Sahoo, S. Dasgupta, S. K. Panda and J. X. Xu, "A Lyapunov function-based robust direct torque controller for a switched reluctance motor drive system," IEEE Transactions on Power Electronics, vol. 27, no. 2, 2012, pp. 555-564