The Output Feedback Stabilization And Tracking For A Class Of Switched Nonlinear Systems

As an important kind of hybrid systems,switched systems describe the multimodal switching and have been widely used in practical engineering systems.Under the background of network communication,complex systems with the characteristics,such as sampling data,quantization signal,random noise and so on,are one of the current hot research tropics.In this dissertation,based on the sampled-data control and quantized control strategies,the output feedback stabilization and output tracking problems for several kinds of switched nonlinear systems with discrete feedback are solved by using the method of adding a power integrator,dynamic gain technology,homogeneous system theory,Lyapunov stability theory and stochastic system theory.The main contents are concluded as follows:1.The global practical output tracking problem is studied for a class of high-order nonlinear systems with different switching powers.The system power and nonlinear terms are both related to the switching signal.For this case,a series of constant gains related to the switching signal are introduced to ensure the feasibility of designing the output feedback controller via a common Lyapunov function.The control input is quantized by a logarithmic quantizer,which is dealt with the sector-bounded method.By adding a power integrator technique and the homogeneous domination theory,an improved design method of homogeneous output feedback controllers is proposed.It is proved that by adjusting the quantization parameters,all states of the closed-loop system are bounded,and the tracking error converges into an arbitrary small neighborhood of the origin within a finite time.2.The problem of the sampling stabilization problem is studied for a class of high-order nonlinear systems with different switching powers.Based on the continuous-time controller of the nominal part and a reduced-order observer,the sampling output feedback controllers are designed.By virtue of using the homogeneous dominant technique and the Gronwall-Bellman inequality,the relationship among the sampling states,the current states and the error between them is established.Combined with Lyapunov stability theory,it is proved that the closed-loop system is globally stable in finite time with small enough sampling period.3.The global practical output tracking problem for a class of switched nonlinear systems is studied.Considering that the system input and output are quantized by the uniform quantizer,which results to the discontinuity of the right-hand side function of the closed-loop system.The Filippov solution and differential inclusion theory are utilized to analyze the closed-loop system.Based on the quantized information,an observer is constructed to estimate the unknown states of the system.The output feedback controller is designed with the sampling information.It is proved that by choosing the small enough sampling period and regulating the quantization parameters such that the tracking error can converge to an arbitrary small neighborhood of the origin within a finite time.4.Consider a class of switched stochastic nonlinear systems with unknown homogeneous growth coefficients.Based on the dynamic gain technique and adding a power integrator technique,a design strategy for the adaptive output feedback controller is proposed.The sectorbounded method is used to deal with the control input which is quantized by a logical quantizer.By applying stopping time,Borel-Cantelli lemma and non-negative semimartingale convergence theorem,it is proved that all signals of the closed-loop system are globally almost bounded.By means of the generalized stochastic Barb?lat's lemma,it is further proved that all states of the closed-loop system almost surely asymptotically converge into the origin.5.Based on a reduced-order observer,an improved design method of a sampled-data output feedback controller is proposed for a class of p-norm switched stochastic nonlinear systems.By using Lyapunov functional method and stochastic analysis techniques,such as It?'s equidistance and Chebyshev's inequality,it is proved that the closed-loop system has a global solution and is globally asymptotically stable in probability with a sufficiently small sampling period.Finally,a three-order numerical example and a single link manipulator system with two switching modes are used to verify the effectiveness of the proposed control method.