Stability, Filtering And Control Of Distributed Parameter Systems Based On Sensor/Actuator

Distributed parameter systems have broad application background because of their space-time evolution characteristics and can more accurately describe the actual system.In view of this, study on the distributed parameter systems is of profound theoretical and practical signi?cance. Based on Lyapunov stability theory, combining with It?o differential formula, by using Riccati equation, differential operator feature, event-driven control, model reduced-order, adaptive control as the main techniques, and sensor nodes gather information, actuator nodes perform control tasks in distributed parameter systems. This dissertation focuses on stability, ?ltering, optimal control and fault tolerant control problems of distributed parameter systems based on sensors and actuators. The main contributions can be generalized as follows:1. Problems for non-collocated moving sensors and moving actuators are unitized for improving stability performance in distributed parameter systems and stochastic distributed parameter systems are studied. Each moving sensor/actuator trajectory is obtained respectively when the system is faster to stable. Through simulations, it can be found that the bene?t of such sensors and actuators in enhancing distributed parameter systems stability performance.2. Schemes for non-collocated sensors and actuators are utilized for ?ltering in distributed parameter systems and stochastic distributed parameter systems are discussed. A class of distributed consensus ?lters with coupling terms result in the state errors forming stable evolution systems and converge to distributed parameter systems and stochastic distributed parameter systems respectively. Then the preferable state estimations are derived.3. On the basis of event-driven control, the optimal switching of actuators in distributed parameter systems is developed. By designing an appropriate state observer and event-driven condition, the switching time interval of un?xed length can be obtained.The event-driven control loop and minimum inter-event time are bounded. Then, based on linear quadratic optimal control, the actuators optimal switching policy is obtained.Finally, the simulation results illustrate the bene?t of the proposed method.4. An architecture for fault detection and fault-tolerant control for distributed parameter systems with only one actuator and part of sensors be activated is studied. For the ?nite-dimensional reduced order model, a class of detection observers with corresponding time-varying thresholds that ensures actuator faults are detected. Under the condition of unknown actuator faults, a class of adaptive observers that estimates actuator faults is designed. An switching policy is used to arrange switching from faulty actuator to healthy ones. And sensor fault detection and fault-tolerant control is considered.Then, actuator and sensor fault detection and fault-tolerant control implementation in the in?nite-dimensional system are discussed on the basis of singular perturbation formulation.5. For multiple actuators are activated, the model-based fault detection and faulttolerant control problems of distributed parameter systems are studied. A set of observers are designed to detect the occurrence of fault in each actuator separately, which determine the positions where the faults have occurred. After the faults occur, adaptive diagnostic observers are then constructed to diagnose corresponding actuator faults. Based on the online approximation of fault parameter, automated control recon?guration architecture is developed to accommodate actuator faults and to preserve close-loop system stability and not affect other actuators’ operation.