| Summary: | 博士 === 國立臺灣科技大學 === 機械工程系 === 107 === This dissertation proposes the adaptive finite time intelligent tracking control strategies for parallel manipulators in the case of complex system uncertainties and external disturbances. First, an adaptive chattering free neural network controller is developed and discussed. The proposed controller is a combination of a feedforward neural networks with online learning, a nonlinear sliding mode technique, and an adaptive technique. The adaptive neural networks are adopted to handle system uncertainties and disturbances. Second, in order to resolve the finite-time convergence existing in the use of the adaptive chattering-free neural network controller, a finite-time adaptive fuzzy tracking controller is proposed. The proposed approach is based on fuzzy logic systems, nonsingular fast terminal sliding mode control and adaptive sliding compensation from nonlinearities in output feedback. The adaptive chattering free neural network and finite-time adaptive fuzzy tracking controllers can have much better tracking control performance. However, in the proposed controllers, the intelligent learning algorithms may take time, but for computation, those are only parameter updates. Therefore, we propose a global finite time active disturbance rejection control scheme which combines an active disturbance rejection control and a global finite time control. This proposed scheme not only can convert fast to the globally finite-time stable equilibrium but also can have superior tracking control performance. In the global finite time active disturbance rejection control approach, the extended state observer is employed to handle the estimation of complex uncertainties and external disturbances. Hence, it may increase the computational burden of the systems. To address this issue, an adaptive nonsingular fast terminal sliding mode control strategy is proposed to have finite time and high-speed trajectory tracking for parallel manipulators with unknown bounded uncertainties and external disturbances. The proposed approach is a hybrid scheme of the online non-negative adaptive mechanism, tracking differentiator, and nonsingular fast terminal sliding mode control. The proposed controller has several advantages such as simple structure, easy implementation, rapid response, chattering-free, high precision, robustness, singularity avoidance, and finite time convergence. Finally, this dissertation gives some examples to validate the effectiveness of the proposed methods.
|
|---|
