The Study On Adaptive Fuzzy Control Of Nonlinear Systems With Input And Output Constraint Characteristics

From the perspective of Control Engineering, the ultimate control performance of a system is not only relevant to the controlled object, but is also affected in the control loop by the performance of a variety of such physical devices as actuators and sensors as well as by the performance of communication channels. Due to system modeling errors, the influence of the work environment and other factors, the controlled objects are usually characterized by essential nonlinearity and uncertainty. Actuators and sensors, more often than not, share non-smooth and nonlinear constraint characteristics, which cause the deterioration of system performance and even the system instability when control signals and output signals go through the actuators and the sensors. In addition, the communication channel is limited by network bandwidth, which makes it inevitable for the control signal to be quantized prior to transmission. Subsequently, quantization errors will cause great negative impact on the control performance of systems.In view of this, the thesis takes backstepping as framework and fuzzy logic system as function approximator to study the adaptive control problem of nonlinear systems featuring input and output constraints. And this thesis consists of six chapters. Based an the introduction of the background and the existing studies on nonlinear systems with input and output constraints in Chapter 1, the main research contents are made up of the other five chapters.Chapter 2 presents a direct adaptive fuzzy output feedback scheme for a class of nonlinear systems with unmodeled dynamics and dynamic disturbances. In the control design, a linear state observer is introduced to estimate the state of the system, a fuzzy logic system (FLS) is adopted to approximate the unknown virtual control signal, and a backstepping technique is utilized to design an adaptive fuzzy controller. Then the generalized small-gain approach is applied to prove the input-state practical stability (ISpS) of the system. The proposed control scheme here, compared with the existing papers, has relaxed the forced assumption on the dynamic disturbances to the system, and has reduced to three parameters the number of on-line adjusting adaptive parameters by the estimation of the vector norm of fuzzy logic systems, which speeds upthe efficiency of online adaptive control algorithm.Taking into full consideration the uncertainty and perturbation characteristics of the actuator dead zone under complex environment, Chapter 3 designs a novel fuzzy dead zone model to investigate the tracking control of nonlinear systems with uncertain dead zone input. By applying the fuzzy set theory and adopting an integrated control idea, an adaptive control scheme, first of all, is proposed for a class of nonlinear strict-feedback systems with fuzzy dead zone and immeasurable states. The proposed scheme guarantees that all the signals in the closed-loop systems are bounded, and the desired tracking performance is achieved. Then, the tracking control problem is probed into for a class of nonlinear systems with fuzzy dead zone and unmodeled dynamics. By utilizing an auxiliary signal to control unmodeled dynamics, and combing the integrated control idea and the dynamic surface control technique, an innovative adaptive controller is constructed.Considering that the actuator hysteresis direction is easy to jump, Chapter 4 proposes a new Bouc-wen hysteresis model of variable direction. On the basis of this hysteresis model, the adaptive tracking control of a class of pure-feedback stochastic nonlinear systems with hysteresis input is researched. Based on Lemma 5.1, by introducing an auxiliary virtual controller and applying the properties of Nussbaum function to stochastic pure-feedback nonlinear systems, the major difficulty arising from the variable direction of the hysteresis input is overcome. And by combining the backstepping, a novel adaptive fuzzy control scheme is established. Compared with the existing literature on the hysteresis input, the system this chapter takes into consideration is more generic so as to expand the application of the hysteresis input.To counterbalance the negative impact of the output nonlinearity in the transmission on the system performance, Chapter 5 investigates a tracking control problem of a class of strict-feedback nonlinear systems with unknown dead-zone output. On the one hand, the existing works on the output nonlinearity focus on the stabilization of linear systems or non-linear systems with matching conditions, in which the approaches can not be applied to the tracking control problem of such complex nonlinear systems as strict feedback nonlinear ones. On the other hand, the state variables in practical systems aredifficult to obtain, which results in the failure of the previous backstepping design schemes containing some or all the state variables to control such systems. By establishing relations between the nonlinear function of the state and the system output as well as introducing a Nussbaum function and an auxiliary virtual controller, a novel controller design scheme is worked out, which succeeds in dealing with the tracking problem of such complex systems.Taking account of the wide application of the quantized feedback control in the digital control and networked control systems, Chapter 6 studies the performance control problem of stochastic nonlinear systems with quantized input constraints. First, by taking advantage of the sector bounded property of the hysteretic quantizer, a new nonlinear decomposition strategy of the quantized output is put forward, which surmounts the difficulty in defining the bounds of disturbance terms in traditional linear decomposition. Then by applying this decomposition strategy, a new adaptive fuzzy scheme is presented, which solves the tracking control problem of the stochastic strict feed-back nonlinear systems with quantized input. This scheme compensates the quantization error by use of online learning mechanism, which makes unnecessary the forced assumptions satisfied by systems and the parameters of quantizers so as to guarantee the tracking performance of the systems under the limited communication frequency. After that, the negative impact of the unmodeled dynamics on the quantitative feedback nonlinear systems is taken into full consideration and a stabilization problem of a class of stochastic nonlinear systems with unmodeled dynamics and quantized input is addressed. By utilizing the backstepping technique and employing the stochastic small-gain theorem, a direct adaptive fuzzy output feedback control scheme is developed, which guarantees that the closed-loop system is input-to-state practically stable (ISpS) in probability.