A Robust Servo Control System Design for a Parallel Robot

• Main Authors: Yuan-Hung Tai, 戴源宏
• Other Authors: Chao-Shu Liu
• Format: Others
• Language: zh-TW
• Published: 2013
• Online Access: http://ndltd.ncl.edu.tw/handle/23882180778264786676

碩士 === 國立高雄應用科技大學 === 機械與精密工程研究所 === 101 === This Master Thesis is mainly to study and develop a servo control system for a parallel robot. Parallel robots are often applied in the field of fast moving objects, such as plugging loads with high-speed positioning. A parallel robot has a closed loop kinematic structure which is different from an open structure of a serial robot. For the limitation of the closed loop kinematic structure, a parallel robot will have a highly coupled dynamics between the end-effector and actuators. When one of the actuators moves in the coupling mechanism, it will disturb the other actuators simultaneously and then induce the results of inaccurate positioning. Therefore, this thesis will develop and implement a smoothly robust controller for a parallel robot, so that the parallel robot can be controlled stably and has excellent performances. Manifold Deformation Design Scheme (MDDS) is a kind of servo control scheme with the characteristics of smooth robustness. The scheme is a mapping technology to perform the system dynamics in the state space, and then to analyze the dynamic behavior and design the controller further. By the way of designing the desired manifold to be the reference of the system dynamic behavior in the state space, and calculating and compensating for the difference between the current state and the desired state, then the control efforts will force the behavior of the controlled system to approach the desired manifold. In fact, the mechanical system is always affected by gravity, friction, inertia of physical quantities and external disturbances, which must block the controlled system to reach the target state accurately. Manifold Deformation Design Scheme has the capability of prediction and estimation of the system dynamics so as to predict and estimate the system state in the next sampling interval. When the system state is different from the estimation state, the controller will compensate for the control efforts directly to overcome the state errors induced by the disturbance, and to keep the system dynamic on the desired manifold. In the experiments, the main controller is combined by a FPGA and a PC, and to realize the motion controls on the MIRDC Delta Robot. The FPGA is used to be the motion kernel, and the schemes of Manifold Deformation Design Scheme and PID control scheme are implemented by the PC in C-language. After various tests and validation, Manifold Deformation Design Scheme has excellent control performances obviously compared with PID control scheme.