Filtering For State-Constrained Systems Based On Encoding-Decoding Scheme

With the development of the control theory and method of state-constrained system,the research on state-constrained system has attracted more attention.State-constrained system is a kind of system with constraints in control inputs,outputs and system states.Due to the limitation of security and physical conditions,state constrained system is widely used in practical engineering.How to improve the robustness and dynamic performance of the system as much as possible under the premise of satisfying state constraint conditions is of great significance in theoretical research and practical applications.In recent years,with the emergence of new control theories and methods,filtering problem for state-constrained system has gradually become a research hotspot.In addition,in the practical application of engineering,due to the limitations of actual conditions,it is impossible to directly obtained part or all state information of the system,so that the desired optimization,estimation,prediction and other purposes cannot be achieved.Therefore,it is particularly important to estimate the actual state information as accurately as possible based on the measurement output of the system.Furthermore,with the continuous development of computer and digital communication technology,encoding-decoding technology has a wide range of engineering application prospects because of its significant advantages such as error control,strong anti-interference and easy encryption.In view of this,we will deeply study the filtering problem of state-constrained systems with encoding-decoding technology.In this dissertation,a new state-constrained system model is established by deeply analyzing the influence of filter characteristics,noise and other factors.Then,based on encoding-decoding scheme,a class of state estimation problem of constrained nonlinear system is studied,which alleviates the inaccurate estimation of designed recursive filter due to the state-constrained conditions of the system.Subsequently,based on the results of previous studies,a type of nonlinear systems filtering problems with state-constraints were studied by using Lyapunov stability analysis and matrix theory.Moreover,with the introduction of encoding-decoding scheme,which reduces data redundancy,saves network communication resources,improves data transmission efficiency and security.Specifically,this article works as follows:1.A binary encoding scheme is adopted for discrete nonlinear system with constraint,the signal measurements received by the sensor are encoded into a group of binary bits,and then transmitted to a remote filter through binary symmetric channel(BSC).Firstly,the statistical characteristics are quantitatively analyzed for the random bit errors.Secondly,an upper bound is derived for the constrained filtering error covariance.The desired filter gain matrix is further designed by utilizing completing-the-square method.Then,the boundedness of the proposed constrained filtering schemes are analyzed in the presence of nonlinearities and binary encoding schemes.Finally,by constructing Lyapunov function,a sufficient condition is obtained to satisfy the bounded error of the system constraint estimation,which effectively reduces the conservatism of the system.2.For discrete nonlinear systems,we consider constraints of the system,the state estimation problem of constrained nonlinear system based on multi-description encoding scheme is studied.First,the multi-description encoding mechanism is introduced to encode the measured signal into two equally important descriptions,and two independent random variables are introduced to control packet loss rate in the channel,which greatly improves the efficiency and success of data transmission.Then,by using Lyapunov stability and matrix theory,a nonlinear system filter which satisfies the state-constraint is designed.