小腦模型控制器於超音波馬達定位控制

Cerebellar Model Articulation Controller for the Position Control of Ultrasonic Motor

陳榮志(Rong-Jyh Chen)

指導教授 ： 洪欽銘

國立臺灣師範大學/科技學院/應用電子科技研究所/碩士(2006年)

摘要

本論文研製以小腦模型為理論基礎之控制器(Cerebellar Model Articulation Controller, CMAC)，於直線型超音波馬達(Linear Ultrasonic Motor, LUSM)定位控制，控制精度在10μm的要求，並使用8位元微控制器加以實現。LUSM是利用壓電陶瓷的逆壓電效應產生超音波震動，透過摩擦來轉換動能，極適合運用於需要精密定位的儀器或設備中，但LUSM本身具有時變與非線性的輸出特性，其輸出性能會隨運轉時間與工作溫度有所差異。

CMAC本身具有良好學習與模仿能力的特性，所以使用CMAC來控制LUSM，可以藉由簡單的學習法則，來達到學習LUSM數學模型，為了求得LUSM的數學模型，首先使用PIC18F452微控制器製作Rule-based控制器來找出LUSM的數學模型，再將此數學模型經由PC作離線學習後求得各記憶體的權值，最後使用PIC18F452微控制器來作為整個CMAC的核心，撰寫組合語言程式並輸入CMAC學習樣本， 並証明CMAC可以有效控制LUSM定位問題。

關鍵字

小腦模型控制器 ； 超音波馬達 ； PIC18F452

並列摘要

A procedure is controller based on the theorem of the Cerebellar Model Articulation (CMAC). It can apply to control the position control of Linear Ultrasonic Motor (LUSM) and use an 8-bit microcontroller to achieve under 10μm precision of displacement. LUSM produces the shake of ultrasonic wave by the inverse piezoelectricity effect of the piezoelectric ceramic and changes rubbing to the kinetic energy. It is suitable for applying of accurate instrument or equipment. But LUSM has some defects which has time-variant and non-linear output characteristics. The output performance of LUSM can be changed with working time and temperature.

The specialties of CMAC are good abilities of learning and imitation, so it is suitable to control LUSM. CMAC uses some simple learning rules to learn the mathematical model of LUSM. First, in order to get the mathematical model of LUSM, we use a microcontroller of PIC18F452 to make a rule-based controller to find them. Then, using PC calculates weight values of memory of LUSM from the mathematical model of LUSM. At least, it uses a microcontroller of PIC18F452 as the core of CMAC to proof that can achieve to control efficiently the position control of LUSM.

並列關鍵字

CMAC ； USM ； PIC18F452

參考文獻

[4]沈孟樵，“線性超音波馬達共振驅動電路與智慧型控制”，私立中原大學電機工程學系碩士論文，2002。

[5]Faa-Jeng Lin, Rong-Jong Wai, and Rou-Yong Duan, “Neural-Network Controller for Parallel-Resonant Ultrasonic Motor Drive,” IEEE Trans. On Control Systems Technology, vol.7, no.4, 1999.

[6]Faa-Jeng Lin, Rong-Jong Wai, and Chun-Ming Hong, “Identification and Control of Rotary Traveling-Wave Type Ultrasonic Motor Using Neural Networks,” IEEE Trans. On Control Systems Technology, vol.9, pp.4, July. 2001.

[9]Faa-Jeng Lin, Rong-Jong Wai, and Rou-Yong Duan, “Fuzzy Neural Networks for Identification and Control of Ultrasonic Motor Drive with LLCC Resonant Technique,” IEEE Trans. on Industrial Electronics, vol.46, pp.5, Oct. 1999.

[10]Ya-Fu Peng and Chih-Min Lin, “RCMAC-Based Adaptive Control for Uncertain Nonlinear Systems,” IEEE Trans. on Systems, vol.37, pp.651-666, Jume. 2007.