Target State Estimation Based On Event-triggered Mechanisms And Its Application

With the technical progress of cruise missiles,fighter planes and attack helicopters,detecting and tracking these targets have become the main difficulties in the field of short range air defense.In order to solve the aforementioned problems,developing networked short-range anti-aircraft detection systems based on the idea of “Network centric warfare” has become the future trend.In the networked short-range anti-aircraft detection system,units are connected by communication network,thus different units can share information and track targets cooperatively.Restricted by national conditions,there are still many aerial defense equipments which are based on the idea of “Platform centric warfare”.Developing networked short-range antiaircraft detection systems still needs to utilize these equipments at this stage.However,with the time-triggered communication mechanism,the communication capacity of these equipments may not satisfy with the demand of communication in networked short-range anti-aircraft detection system.Upgrading some hardware components of these equipments is a solution for solving the aforementioned problem,but hardware change needs to be verified by a series of costly tests,such as electromagnetic compatibility test.In conclusion,researching new communication mechanism for decreasing the communication load in networked short-range anti-aircraft detection system,so that the aforementioned equipments can be utilized for constituting the networked short-range anti-aircraft detection system,this is with significant theoretical value and application value.In recent years,event-triggered(ET)communication mechanism and event-triggered target state estimation have gained widely attention and research.With event-triggered mechanism,the current data is transmitted only when the triggered condition is satisfied.Many relevant works have verified that the event-triggered mechanism can remarkably reduce the data transfer amount in sensor networks.Consequently,this thesis devotes to implementing event-triggered mechanisms in the networked short-range anti-aircraft detection system,for reducing the communication load.This thesis studies the following problems: the performance evaluation of target tracking systems with event-triggered communication mechanism,the event-triggered mechanism and the event-triggered state estimation for systems with intermittent measurements,the event-triggered diffusion filtering algorithm for systems with unknown cross-covariance between preliminary estimates,the event-triggered asynchronous fusion estimation for asynchronous sensor networks.The results of this thesis can offer references for implementing event-triggered mechanisms in the networked short-range anti-aircraft detection system.The main contributions of this thesis are summarized as below:1.In order to provide theoretical reference for utilizing event-triggered mechanism in target tracking systems,the estimation performance evaluation for systems with event-triggered mechanism is studied.The ideal Cramér-Rao lower bound(CRLB)based on enumeration method is provided,which quantitatively describes the relation between the threshold of the ET mechanism and the estimation performance.For convenience of application,the statistical CRLB based on information reduction factor method is further given.And relation between the ideal CRLB and the statistical CRLB is studied.Simulation results demonstrate that event-triggered mechanism can significantly reduce the data transfer amount at the cost of only slight estimation performance decline,through setting an appropriate threshold.In other words,it is feasible to implement event-triggered mechanism in target tracking systems.2.In order to reduce the data transfer amount from detectors to fire control computer in systems with intermittent measurements,the ET mechanism for scheduling measurements transmissions and event-triggered state estimation in existence of intermittent measurements are studied.A channel decoupled measurement equation is proposed,which can avoid abandoning valid components of measurements.According to the proposed measurement equation,an ET mechanism based on dual-criterion is prosed to schedule the data transmissions from detectors to fire control computer.With the proposed ET mechanism,the available information about measurements for fire control computer is provided.Based on the prosed ET mechanism,an event-triggered square-root cubature Kalman filter(SCKF)is proposed.According to simulation results,the proposed event-triggered SCKF can significantly reduce the data transfer amount from detectors to fire control computer,compared with the existing SCKF based on time-triggered mechanism.Compared with existing event-triggered filtering algorithms,the proposed event-triggered SCKF can obtain better estimation performance and prevent interference from intermittent measurements.Results of experiment verify that the proposed eventtriggered SCKF can be applied to in actual optical-electronic detection systems.3.For the sake of reducing data transfer amount between fire control computers,the ET mechanism for scheduling preliminary estimates transmissions and the event-triggered diffusion filtering algorithm for systems with unknown cross-covariance are studied.An ET mechanism for scheduling the exchange of preliminary estimates between neighbored fire control computers is proposed.With the proposed ET mechanism,the available information about preliminary estimates for fire control computer is provided.Because obtaining the cross-covariance between preliminary estimates is usually difficult,an event-triggered fusion algorithm based on covariance intersection method is proposed.And the fusion estimates of the proposed method is proved to be consistent.Through combining the proposed event-triggered fusion algorithm with the event-triggered SCKF proposed in Chapter 3,an event-triggered diffusion SCKF is proposed.Simulation results show that the proposed method can remarkably decrease the measurements transfer amount and the preliminary estimates transfer amount in sensor networks.Experimental results show that the proposed method is feasible in actual detection systems.4.In order to reduce the data transfer amount in asynchronous detection system,the ET mechanisms and event-triggered asynchronous fusion estimation for asynchronous sensor networks are studied.Two ET mechanisms for scheduling transmissions of asynchronous measurements and asynchronous preliminary estimates are proposed,respectively.And the available information about measurements and preliminary estimates for the fire control computer is provided.Based on the two ET mechanisms,an event-triggered asynchronous measurements fusion algorithm and an event-triggered asynchronous preliminary estimates fusion algorithm are proposed,respectively.Then,through combining the proposed two algorithms,an eventtriggered asynchronous fusion estimation algorithm is proposed.Simulation results demonstrate that the proposed method is effective in reducing data transfer amount in asynchronous sensor networks.And results of experiment verify that the proposed method can be applied to actual asynchronous detection systems.