透過您的圖書館登入
IP:3.135.221.0
  • 學位論文

共軸式機械手臂伺服控制

Servo Control of Dual-Axis Differential Robot Arm

指導教授 : 陳傳生

摘要


傳統的機械手臂多為使用串聯式,單一關節由一顆馬達來驅動,而各關節成前 後相連的形式,每一關節馬達必須負擔前端所有重量,導致越接近底座的馬達越大。 為增加靈活性及減少體積採用差動機構設計的機械手臂。齒輪差動機構具有俯仰 及旋轉兩個自由度,利用差動機構特性可在兩個自由度上獲得兩倍於單一馬達的 扭力。差動機構需同時控制兩顆直流無刷馬達來達到動作,市面上卻不見同時驅動兩顆直流無刷馬達驅動器。本論文使用dspic33FJ128MC804 微控制器為控制核心,利用Dspace 透過Mathworks 的軟體Matlab、Simmulink 開發控制器,以快速完成共軸式機械手臂的控制器。 控制器的開發過程中,先對開路系統輸入激發訊號,再將量測到訊號透過Matlab 的系統鑑別工具鑑別出各軸的系統模型,再利用State-Space 與RST 控制方法來設計定位控制器。為使機構能平順運動,避免突然的加減速造成機械結構損壞,加入路徑規劃

並列摘要


Traditional robot arm is design by series connection in the system. Each joint is driven by a motor; which causes the drawback of heavy load. Since the total weight of the whole system are the number of motors which connected in direct proportion. Thus,the base motor supports the load of all robot arm will need more torque compares to others. In order to add more flexibility and use less space in design, we offer the concept of dual-axis in differential robot arm. There is two degree of freedom in one joint by using differential driver configuration. Because of the advantage of differential driving mechanism, one degree of freedom can get power from both input motor. Differential drive mechanism needs two BLDC motors with synchronization to achieve desired movement. However, it is unavailable in the market. In the thesis, we use the dspic33fj128mc804 microcontroller as our main control unit, which includes the Dspace,the Matlab and Simulink from Mathworks to quickly realize the controller design in coaxial. In the process of our design, we have to give a multiple-sine signal to the open-loop system first. Then we apply the identification toolbox form Matlab to identify the system mode. Second, we use the concept of State-Space and RST to develop the controller. Lastly, in order to have our mechanism move smoothly, we add the method of path planning to avoid the damage caused by suddenly acceleration or deceleration.

參考文獻


Controller for a Flexible Joint Robot with Strong Joint Coupling,” IEEE
International Conference on Robotics and Automation, pp.10-17 April 2007
[4] Ulrich Hagn,et al.”TheDLR MIRO: a versatile lightweight robot for surgical
Advanced Motion Control, pp88-93 March 2010
[6] L. Ljung, “System Identification Toolbox

延伸閱讀