FIR State Estimation For Systems With Time Delay And Data Loss

The state of the system has important reference significance for the detection,diagnosis,and control of engineering systems,and is related to the overall safety and economy of the system.The phenomenon of time delay and data packet loss widely exists in engineering system processes,such as in power electronic systems,communication systems,microcomputer power systems,etc.Due to external physical reasons or internal characteristics,there are various types of delays;In target tracking,measurement failures and data packet loss will lead to delays in high-precision operations;When data needs to be post-processed,there is usually a delay between obtaining measured values and providing filters.In discrete time processes,data loss caused by sampling can also occur.Overall,time delay and data packet loss are the two main factors leading to system instability and poor performance.Therefore,based on finite impulse response and maximum likelihood theory,this article investigates several state estimation problems under complex delay and data packet loss scenarios.The main research work is summarized as follows:(1)Aiming at the state estimation problem under the time-delay and data packet loss system with time stamp,based on the finite impulse response theory,a maximum likelihood finite impulse response algorithm(Maximum Likelihood Finite Impulse Response,ML FIR)is developed.By reconstructing the original state space model,transforming the delay and missing measurement models,and then combining the maximum likelihood theory,the maximum likelihood FIR processing method is derived by processing and computing in two forms of slow batch processing and fast iteration.Finally,it is verified by the experimental simulation of the harmonic model and the drift random resonator model,which proves that the method has superior performance for dealing with the state estimation problem of the system with time stamp delay and data packet loss.(2)Aiming at the case of Bernoulli distribution time delay and data packet loss,based on the maximum likelihood rule,the slow-rate batch form and fast-rate iterative form of the finite impulse response filter are designed respectively.Firstly,the model under the condition of delay and data packet loss is expressed as a probability linear function obeying the Bernoulli distribution,and then the batch processing form of the maximum likelihood finite impulse response algorithm is obtained through maximum likelihood processing.In order to optimize the computational complexity,Then a fast iterative form algorithm is derived.In order to verify the effectiveness of the method,under the same delay and data packet loss conditions,the experiment part compared other filtering methods,and compared and analyzed from the perspectives of estimation error,root mean square error and uncertainty,and finally obtained the algorithm It has better estimation performance and higher robustness.(3)A maximum likelihood finite impulse response filter is designed based on the fir principle and maximum likelihood theory for the state estimation problem in the case of random time delay and data packet loss.First,aiming at the randomness of time delay and data packet loss,a state-space equation covering all situations is established,and based on the classification idea and maximum likelihood discussion,the problem of processing the model is resolved.By extending the state-space equations,a maximum likelihood finite impulse response algorithm for systems with random time delays and packet loss is developed,and a slow-rate batch form and a fast-rate iterative form are derived.It is verified by experiments that this method has good estimation performance and good anti-interference ability in solving the state estimation problem of random time delay and data packet loss.