Research On Networked Control Systems With Variable Sampling Intervals

Networked control systems (NCSs) are some feedback control systems where thefeedback loops are closed by some communication networks. NCS offers manyadvantages such as less wiring, increased system’s flexibility, easy installation andmaintenance, etc. But for imperfect communication and limited band constraints, there aresome disadvantages such as time-delay, packet dropouts, variable sampling/transmission,quantization errors, which will degrade the performance of the systems significantly.Therefore, study on networked control systems such as stability analysis, controller design,fault detection and fault tolerant is very important in both theory and applications.Compared with the rich results on time-delay, packet dropouts and quantization errors,there are few results on variable sampling intervals in networked control systems.Moreover, very limited papers take time varying sampling intervals into account for theresearch of fault detection and fault tolerant control. Therefore, based on a newly proposedmodel, the dissertation investigates the sufficient condition for existence controller to keepasymptotically stable, robust controller design, fault detection and fault tolerant control forNCS with variable sampling intervals and time-delay. The main work and contributionsare as followings.We give a review of the previous work on networked control systems, including theresearch topics, the modeling approaches, and the achieved results about the variablesampling intervals. Meanwhile, some problems which need further study are pointed out.A new model is proposed which simultaneously describes variable and boundedsampling intervals and time-delay. A discrete-time model with time-varying parameterslying inside a convex polytopic system framework whose vertices are determined throughthe Real Jordan form approximation is built. Using the convexity of the discrete-timepolytopic uncertain systems, i.e., the result of a finite number of the vertices will beapplicable to the entire polytopic system and the new model becomes the basis of laterdesign and analysis.Based on Lyapunov stability theory and linear matrix inequalities (LMIs) method,parameter-dependent and parameter-independent states feedback controller, guaranteed cost states feedback controller and dynamic output feedback controller are designedrespectively. The sufficient conditions of existence the above controllers are given in theform of LMIs. Considering the disturbance and noise, the mixed robust H2/H∞performance of the discrete-time polytopic uncertain system is proposed out, which isused to improve the system transient performance and to ensure the robustness of thesystem simultaneously. The robust states feedback and dynamic output feedbackcontrollers are designed.An optimal H∞robust fault detection filter is built. For all unknown inputs, uncertainparameters, the error between the residual signal and the fault signal is made as small aspossible. The sufficient conditions for existence of the desired robust fault detection filterare established in terms of LMIs. Based on H∞index performance for minimizing thedisturbance and the corresponding the finite frequency H－index performance formaximizing the worst-case fault sensitivity, a fault detection observer is designed forresidual generation. With the aid of Bounded Real lemma and the generalizedKalman-Yakubovich-Popov (GKYP) lemma, the existence condition of such a mixedH∞/H－fault detection observer is given in terms of LMIs.Considering the possible actuator failures, an observer-based integrated robust faultestimation and fault tolerant controller are studied. First, an augumented fault estimationobserver containing an exponential stability performance index and an H∞performanceindex are proposed not only to guarantee the convergence speed of fault estimation butalso restrict the uncertainties to be as small as possible. Second, two robust fault tolerantcontrollers based on states feedback and dynamic output feedback are designed to ensurethe robust stability of the closed-loop system in the presence of disturbance and faults. Forthe optimal robust H∞fault tolerant control, the parmeters of the controllers are obtainedby solving some convex optimization problem with LMIs constraints.At the end of each subsection, some numerical simulations which demonstrate theeffectiveness of our results are given. Meanwhile the comparison and analysis of thecontrol performance between the various controllers, fault detectors and fault tolerantcontrollers are given.