Research Of Stabilization Of Quantized Feedback Control Systems Based On Event Trigger

In recent years, with the development of network communication technology and control theory, networked control systems are getting more attention and developping at high speed. This type of systems has great advantages when compared to traditional control systems, closed feedback control systems can be set up in a larger region without linklines of point-to-point form any more. The long-distance transmission of signal depends on digital communication network, this change improves the execution efficiency and flexibility. However, in networked control systems, some measured amount and controlled quantity need to be quantified and coding to be transmitted through digital communication channel, this behavior leads to quantization error, we mainly study this type of closed control system with quantization error, it is called quantized feedback control systems. As the scale of networked control systems become larger and larger, its characteristics get more complicated, the consumption becomes more, people hope to improve its performance, decrease the consumption and optimize the system, for example reducing the number of bits of coding transmission or lowering the frequency of sampling and updating the controller or finding more efficient dynamic quantization strategy and so on, so about studying, analysising and designing quantized feedback control systems, providing new method and strategies is of important meaning and value both in theory and practice.The research and exploring of quantized feedback control system of this article is Divided into three parts, first a kind of discrete time linear time-invariant output feedback system with bounded system noise and output noise is studied, then it is extended and apply a dynamic quantized strategy based on space division to the system, and make the system achieve practically stable, while others usually consider the case of full-state feedback system without noise, in this problem this part manages two kinds of error, the observational error and quantization error, and make the state become bounded when it goes to steady state.In the second part the optimizing of the strategy of sampling and updating the system are considered, usual quantized feedback control is based on continuous sampling or fixed perodic sampling, in fact these method may lead to unnecessary sampling and updating, and causes some waste. To this, this paper brings event-triggered sampling machanism to improve this problem. Event-triggered sampling is with a view to the actual requirement of a system, its form is usually that a kind of error related to the states of the system changes as the time goes and when it exceeds a threshold the system samples once and updates the controller, in the rest of time the controller maintains its state under the ZOH. this type usually has a longer average period than perodic sampling. This paper brings event-triggered sampling to a kind of Continuous time invariant full-state feedback control system with bounded noise, and give the condition that guaranteeing the system achieveing L2stability and the constraint of the quantization density of logarithmic quantizer, at last we compare with another paper in this field, with the across comparison it itsverificated that our method has a lower average frequency of sampling by simulation.The third part of this paper shows how to bring the event-triggered sampling to output feedback quantized control system, in this part bases on the event-triggered control of output feedback system without quantizer, and is extended to the system with logarithmic quantizer and bounded noise, by analysising the Lyapunov function, we provide the condition to stable the system, this paper gives the event-triggered condition by testing the existance of feasible solution of LMI, in the end we verify the conclusion by simulation.