手輪馬達電動輪椅力矩控制暨參數識別

Torque Control Strategy and Parameter Identification for a Powered Wheelchair Driven by Rim Motors

林怡劭(Yi-Shao Lin)

指導教授 ： 陽毅平

國立臺灣大學/工學院/機械工程學研究所/碩士(2017年)

https://doi.org/10.6342/NTU201703701

摘要

本文研究目的在於開發與整合手輪馬達電動輪椅與控制系統策略。文中先介紹手輪馬達電動輪椅之整車架構，整車特色在於動力輪技術整合，包含手輪馬達、馬達驅動器、機械式煞車和鋰鐵電池，故一般輪椅車架只需加裝兩個動力輪即可成為電動輪椅。再經由磁場導向控制應用於整車控制系統中，透過上控制器和下控制器的互相訊號傳輸，將整個架構完整建出，其中下控制器之控制策略為本文重點之一，主要為電流的PI控制器，藉由其參數的調整，使電流回授能快速追上參考值，進而讓電動輪椅在行駛途中增加整車的穩定性，提高運行之舒適度。
另外，為了使控制器參數能有效達到馬達控制，因此需要配合手輪馬達的參數識別來做調整，手輪馬達的參數主要包含機械參數和電氣參數，分別需要透過不同的方法來量測。
另一方面，利用三顆數位霍爾感測器做估測磁場導向所需要的轉子角度也是本文著重焦點之一，考量到估測角度出現誤差可能會使馬達產生振動進而使輪椅使用者感覺到不舒適，因此希望透過角度修正的方法來避免此情形發生。
而手輪馬達電動輪椅之控制硬體架構中主要分為上控制器和下控制器，上控制器為Arduino Due，主要負責整車的人機介面操作、藍牙連結、下達電流命令與實行雙輪耦合；下控制器為TI驅動板，主要負責實行磁場導向控制、力矩控制、霍爾感測器角度估測和煞車控制，最後本文之控制策略將會透過模擬軟體做分析以及在整車上做實驗驗證，進而比較其差異性並將其結果與討論於本文末呈現，最後得出力矩控制能有效使手輪馬達電動輪椅之穩定性、緩啟動和舒適度提升。

關鍵字

手輪馬達 ； 電動輪椅 ； 力矩控制 ； 參數識別 ； 數位霍爾感測器角度估測

並列摘要

The main objective of this study is to develop and integrate the control system of the electric wheelchair driven by rim motors. The integration of the powered wheel technology including the rim motor, mechanical brake, motor driver and Li-Fe battery. As a result, for general wheelchair rack, only two powered wheels need to be installed to become the electric wheelchair. Next, the field-oriented control is applied into the whole control system to construct the complete structure with upper and lower controller.
The control strategy of the lower controller is one of the most important part in this study. For the purpose of improving the stability and the fast response of the electric wheelchair, tuning the parameters of the proportional and integral controller is needed. Besides, it is necessary to do motor parameter identification to make the motor control more effective. Motor parameters are divided into mechanical and electrical parameters with different measurements. Further, the rotor position estimation with three low resolution hall-effect position sensors is also one of the most important part in this study. To prevent the user from feeling uncomfortable from the motor vibration due to position estimation error, the correction method is implemented.
The electric wheelchair hardware is formed by upper and lower controller. The upper controller is Arduino Due which is mainly responsible for human machine interface operation, Bluetooth connection, the cross-coupling control and command of the current order to the lower controller. In addition, the lower controller is TI motor driver which is responsible for field-oriented control, torque control, low resolution hall-effect position estimation and brake control. Lastly, simulation and experimental results are shown to verify the operation of the proposed algorithms and to compare the difference between them.

並列關鍵字

rim motor ； electric wheelchair ； torque control ； parameter identification ； low resolution hall-effect position estimation

參考文獻

[58] 董紹安, 手輪馬達電動輪椅之控制系統整合策略, 碩士論文, 國立台灣大學, 台北, 2016.

[59] 陳柏叡, 運用粒子群最佳化法於手輪馬達式電動輪椅用交互耦合轉速比控制器設計, 碩士論文, 國立台灣大學, 台北, 2013.

[61] 施華宇, 手輪馬達電動輪椅之霍爾感測器失效控制策略, 碩士論文, 國立台灣大學, 台北, 2017.

[62] 丁奕元, 基於霍爾感測器之改良型轉子角度估算法應用於內藏式永磁同步馬達之驅動控制, 碩士論文, 國立台灣大學, 台北, 2011.

[5] M. F. Moussa, Y. Gaber and M. El Attar, "Vector control drive of permanent magnet motor without a shaft encoder," 2008 12th International Middle-East Power System Conference, Aswan, 2008, pp. 249-254.