高效能向量控制系統之小腦模型控制器設計

Design of Cerebellar Model Articulation Controllers for High Power Efficiency Vector Control System

李谷元(Ku-Yuan Li)

指導教授 ： 王順源

國立臺北科技大學/機電學院/電機工程系所/碩士(2005年)

https://doi.org/10.6841/NTUT.2005.00100

摘要

本論文設計一小腦模型PI控制器，並植入感應馬達向量控制系統中作速度控制及參數估測。此小腦模型PI控制器具有快速學習、結構簡單、線上訓練和非線性之學習能力，再加上投影演算法作為學習架構，即可單獨作線上控制，並且依權重可信度分配誤差，可加快其收斂的速度。
本文之感應馬達向量控制系統結合直接轉子磁場導向與適應性控制法則，提出一適應性磁通估測器，並利用小腦模型PI控制器作轉速與轉子電阻估測，完成無轉速量測器控制與避免轉子磁通值漂移。在適應性磁通估測器方面，本研究使用虛擬降階型磁通估測器，相對於全階型磁通估測器來說，其不需要隨著轉速命令調整置根比例常數，並具有高度穩定性。另外加入高效能控制，可在感應馬達運轉於穩態時提高其效率。
最後經由實驗證明，此新型小腦模型控制器加入感應馬達向量控制系統中，運轉在2%到100%之額定轉速，負載為8Nm時轉速皆能快速響應及保持良好的強健性。另外，在加入高效能控制下，向量控制系統確實能夠達到提高效率之目的。

關鍵字

小腦模型控制器 ； 參考模型適應性系統 ； 適應性磁通估測器

並列摘要

In this thesis, a cerebellar model articulation PI controller (CMAPIC) is proposed and embedded into the induction motor vector control system to control speed and estimate parameters. The advantages of the CMAPIC include rapid learning convergence, simple structure, on-line training, and non-linear learning ability. With the aid of projection algorithm, the CMAC is applicable for on-line control. Moreover, by introducing the distribution errors with the credibility of weights, credit assigned CMAC (CA-CMAC) can accelerate the convergence rate.
The induction motor drives considered in this thesis adopts the direct field orientation control (DFOC) and adaptive control law to propose the adaptive flux observer. To establish the sensorless control and prevent the drift of rotor flux, the CMAPIC is used to control the speed and estimate the rotor resistance. As to the adaptive observer, the pseudo-reduced-order flux observer (APRO) is used. Comparing to the adaptive full-order flux observer (AFO), the locations of assigned poles is not necessarily adjusted as the speed command varies. Also, the stability is improved. By applying the high power efficiency control, the efficiency is increased during the steady-state operation of induction motor.
From the experimental results, in the operation conditions: the range of speed is set from 2% to 100% of rated speed, and the load condition is 8Nm, it is seen that the speed response of induction motor vector control system equipped with the novel CMAPIC is fast and the robustness is retained In addition, the experiment results demonstrate that the power loss is reduced by using the high power efficiency control algorithm.

並列關鍵字

Cerebellar Model Articulation Controller ； Model Reference Adaptive System ； Adaptive Flux Observer

參考文獻

[1] W. S. Mischo, “A CMAC-Type Neural Memory for Control Applications,” Proceedings of Fifth International Conference on Microelectronics for Neural Networks, San Diego, CA, 1996, pp. 161-167.

[2] Y. Wong and A. Slideris, “Learning Convergence in the Cerebellar Model Articulation Controller,” IEEE Transactions on Neural Networks, Vol. 3, No. 1, 1992, pp. 115-122.

[3] F. C. Chen and C. H. Chang, “Practical Stability Issues in CMAC Neural Network Control Systems,” IEEE Transactions on Control Systems Technology, Vol. 4, No. 1, 1996, pp. 86-91.

[4] W. T. Miller, III, “Real Time Application of Neural Networks for Sensor-Based Control of Robots with Vision,” IEEE Transactions on Systems, Man and Cybernetics, Vol.19, No. 4, 1989, pp. 825-831.

[5] W. T. Miller, III, R. Hewes, F. H. Glanz and L. G.. Kraft, III, “Real-Time Dynamic Control of an Industrial Manipulator Using a Neural-Network-Based Learning Controller,” IEEE Transactions on Robotics and Automation, Vol. 6, No. 1, 1990, PP. 1-9.