| URN | etd-0828109-155012 | Statistics | This thesis had been viewed 1009 times. Download 9 times. |
| Author | Yu- Sheng Jheng | ||
| Author's Email Address | No Public. | ||
| Department | Institute of Mechatronic Engineering | ||
| Year | 2008 | Semester | 2 |
| Degree | Master | Type of Document | Master's Thesis |
| Language | zh-TW.Big5 Chinese | Page Count | 109 |
| Title | Mechanical Design and Control of a 7 DOF Humanoid Robot Arm | ||
| Keyword | |||
| Abstract | The main purpose of this thesis focused on the mechanical design and motion control of a 7 DOF humanoid robot arm subject to parameter variations and external disturbances. In mechanical design, the Inventor computer software is used to achieve the design of robot mechanism, and the Adams computer software is used to perform system dynamic simulation. Under imitating basic movements of human arm, the required torque and maximum velocity of every joint motor of the robot arm can be obtained via dynamic simulations. Thus, we can finish the exertion of proper motor specifications and the reduction ratio of driving elements. In the control problem, because the dynamic model of a humanoid robot arm is highly nonlinear and complicated, a traditional controller is difficult to control it. In this thesis, an independent joint sliding mode controller is proposed to control the robot arm. The proposed controller possesses not only the sliding mode control like robustness but also the independent joint control like simplicity. Moreover, in order to overcome the difficulty needing the closed-form model of robot arm in the general dynamic simulations, a simple and easily implemented algorithm is also proposed in this thesis. The algorithm can finish very efficient dynamic simulation, does not need the closed-form arm model, so long as use the well-known recursive computed torque technique. Finally, the proposed controller is applied to the humanoid robot arm through simulations to evaluate the control performance,. The simulation results confirm that the proposed controller can achieve better control performance over the conventional independent joint PID in terms of high payload variation. This implies that the controller is robust to high payload variation and can be applied to a variety of high-DOF humanoid robot arms with unknown but bounded parameter uncertainties and external disturbances. |
||
| Advisor Committee | |||
| Files | indicate in-campus access at one year and off-campus not accessible | ||
| Date of Defense | 2009-07-29 | Date of Submission | 2009-08-28 |