Adaptive Sliding Mode Control For Nonlinear Systems Based On Intelligent Methods And Application Research

Generally, exact mathematical models of controlled plants are needed for many control design methods. However, unknown uncertainties, perturbations, disturbance and nonlinearities may always exist in practical processes and many classical control design methods can not be applied in practice. In this dissertation, nonlinear systems with unknown dynamic information are mainly considered and their sliding mode control (SMC) syntheses are completed by adopting the approximation techniques of fuzzy logic system (FLS) and wavelet network (WN). Afterwards, the proposed intelligent adaptive sliding mode control (IASMC) design methods are applied to the experiments of an inverted pendulum system in order to verify their validity and performance.Firstly, adaptive wavelet sliding mode control (AWSMC) synthesis method based on the approximation of wavelet network (WN) is presented for a class of non-affine nonlinear systems whose dynamics are unknown. In the case that the nonlinear dynamic lies in L~2 space and satisfies several assumptions, the equivalent control of SMC is constructed according to the properties of the or-thonormal basis of WN in L~2 space. The adaptive laws of the parameters in WN are designed to self-tune online and thus the adaptive equivalent control is implemented. The hitting control is added to ensure that the reaching condition of the sliding mode can be satisfied. The system stability and robustness are proved by Lyapunov stability theory. Whereafter, the AWSMC control synthesis problem for a class of nonlinear systems without Brunovsky canonical form is considered, whose dynamic is unknown and the state information is immeasurable. Under the circumstance that the system can be feedback linearized by I/S(Input/State) or I/O(Input/Output), adaptive observer (AO) based on WN is firstly designed to obtain the unknown system state. The observation error is proved to be asymptotically stable by using Lyapunov stability theory. Then the sliding mode and the corresponding SMC synthesis processes are given on the space of the observation state.Secondly, two direct fuzzy adaptive sliding mode control (DFASMC) design methods are presented for general affine nonlinear systems with unknown dynamics. The former is mainly to resolve the chattering problem in SMC system and adaptive fuzzy logic system (AFLS) is used to approximate the general SMC, instead of directly using the general SMC. In this way, the high frequency signals are filtered by the low-pass characteristic of AFLS, hence the final control signal is smooth. In the latter one-step design method based on AFLS is given to overcome the difficulties of modeling when the nonlinear dymamic is unknown. The parameters of the controller and sliding mode are both determined, and simultaneously the reachability of the sliding mode and the stability of the closed-loop system are both guaranteed. Numerical examples are proposed to validate their feasibility and performance.Finally, an experiment equipment of inverted pendulum is introduced. The proposed AWSMC and the other two DFASMC design methods based on intelligent methods are all applied to this inverted pendulum system to verify their practical performance. Above all a general SMC control design based on the linearization model near the equilibrium place is implemented and the experimental results are proposed. Then the AWSMC and the two DFASMC design methods are applied to the experiment system respectively and many real operation results are reported. The comparison and analysis of these results validate that the performance of the proposed three IASMC design methods are all better than that of the traditional SMC, and the desired effects are all obtained.