Robust State Estimation For Linear Uncertain Time-delay Systems With Packet Los

With the increasing complexity of practical engineering,the research of traditional control theory faces some problems and challenges that need to be solved urgently.In the process of modeling,many problems such as the aging of physical components will make the mathematical model describing the controlled object show uncertainty.Nowadays,the control system is a combination of network,communication and control.The communication network introduces many uncertain factors such as time delay and packet loss,which leads to poor estimation performance and even instability of the system,making the traditional state estimation theory difficult to apply.Therefore,it has become an important topic in the field of control theory to investigate the robust state estimation of systems with data packet loss and time delay.In this thesis,for a class of uncertain linear discrete systems with deterministic external input,random measurement data packet loss and state delay,a robust state estimator based on the sensitivity penalty of innovation process to model errors is derived and extended to multi-sensor networked control systems.The specific research content is embodied in the following three aspects:1.The problem of robust state estimation for a class of uncertain linear discrete systems with deterministic external input,d-order state delay and random measurement data packet loss is investigated.Firstly,based on the state augmented method and the sensitivity penalty of innovation process to model errors,a robust state estimator with similar computational complexity to Kalman filter and recursive implementation is derived.Secondly,it is proved that the pseudo covariance matrix of the estimator’s estimation error converges to the corresponding stationary distribution with probability 1 under some weak assumptions and some controllable and observable matrices.Finally,a numerical simulation case proves that the estimation accuracy of this estimator is more robust than the Kalman filter.2.The robust fusion estimation problem of multi-sensor networked control system composed of multiple remote sensor nodes and a fusion node under the condition of limited communication rate is investigated.Each sensor has uncertain parameters,deterministic external input and d-order state delay.Firstly,each sensor calculates its own local estimation value based on the robust state estimation method that the sensitivity penalty of innovation process to model errors.Then,based on the data-driven transmission strategy,the local estimation value is transmitted to the fusion node to save energy consumption of sensors and reduce network traffic congestion.An analytical robust fusion estimator is proposed at the fusion node,and the condition that its estimation error is uniformly bounded is discussed.Finally,the effectiveness of the robust fusion estimator is verified by a numerical simulation case.3.Based on the research results of the previous two items,the robust fusion estimation problem of multi-sensor networked control system composed of multiple remote sensor nodes and a fusion node under the condition of limited communication rate is investigated.Each sensor has uncertain parameters,deterministic external input,d-order state delay and random measurement data packet loss.Based on the sensitivity penalty of innovation process to model errors and the data-driven transmission strategy,an analytical robust fusion estimator is derived.The numerical simulation case shows that the robust fusion estimator has good fusion estimation performance.