Research On The Key Technology Of Multiple Maneuvering Targets Tracking For Airborne Radar Under Complex Environment

The multi-target tracking refers to the process of real-time estimating the state of unknown number of targets based on sensors,such as radar,sonar,infrared sensor,etc.The multi-target tracking plays a key role in military field,including airborne early-warning radar,Missile guidance and Antimissile defense systems,underwater target tracking system and so on.As the result of the enhancement of the target maneuverability in modern warfare,the requirements of the tracking system are gradually improved,so that the problem of multi-maneuvering target tracking has become an important research hotspot in the field of multi-target tracking.With the development of sensor technology and the emergence of new aircraft,the multi-target tracking environment becomes more and more complicated.Dense clutter,measurement errors,time-varying noise,deceptive interference and other uncertainties might significantly affect the performance of multi-maneuvering target tracking.Therefore,how to predict and track the state of multiple maneuvering targets in the complex environment has become an urgent issue.This paper begins with the multiple target tracking task for airborne radar.The key technology of multiple maneuvering target tracking in complex environment is then studied.It covers the multiple targets track initiation techniques in dense cluttered environments,estimation of the number and state of targets in sea clutter background,angle estimation with the uncertain statistical characteristics of measurement errors and the data association problem of multiple maneuvering targets with false targets jamming.Consider the practicability,the multi-maneuvering target simulation system for airborne radar is designed to vertify the proposed strategy.The main conributions in this paper are listed as follows.1.Aiming at the problem of long computation time and high false track initiation probability for the track initiation of multiple maneuvering targets in dense cluttered environments,a data stream clustering track initiation algorithm based on Interactive Multiple Model Core Estimation is proposed.The proposed algorithm describes the problem of multi-target track initiation in dense cluttered environments to the clustering problem of correlation points,and uses Denstream algorithm to cluster the measurements.In order to reduce the computation time of Denstream data stream algorithm,the Interactive Multiple Model method is introduced to estimate the state of the core of the potential micro cluster,and the measurement is determined whether it fall into the ellipse gate of the estimated core.At last,the Renyi Entropy is introduced to adaptively extract thepotential track.The simulation was carried out for the track initiation of multiple maneuvering targets,the results show the proposed algorithm has the accurate initial position and good real-time performance.2.Regarding to the problem of low track initiation accuracy in the track initiation of multiple maneuvering targets,a track initiation algorithm based on the Maximum Entropy Relief feature weighted Denstream clustering is proposed.The objective function of Relief feature weighting algorithm is modified by the sample force coefficient weighting method,the maximum entropy constraint method and the fuzzy difference degree technique.The weights of the characteristic vectors of data set is calculated using Lagrange method.The weights obtained are used in the Denstream clustering algorithm.Simulation results show that the proposed algorithm can achieve higher track initiation accuracy and lower false track initiation probability.3.Since the Airborne Radar is difficult to accurately track multiple sea maneuvering targets in sea clutter background,a Central Difference Kalman Cardinalized Probability Hypothesis Density filter based on maximum likelihood background parameter estimation(BE-CDKF-CPHD)is proposed.According to the actual process of sea clutter with unknown parameters,the Maximum Likelihood(ML)method is used for estimating the parameters of sea clutter,and calculating the detection probability and false alarm probability.Because the maneuvering characteristics of target and the target number changes over time,the advantage of the Central Difference Kalman filtering algorithm for nonlinear system filtering and the feature of estimate the target number of Cardinalized Probability Hypothesis Density filter is combined,a Central Difference Kalman Cardinalized Probability Hypothesis Density filter is proposed.Amplitude likelihood function is combined with the likelihood function of clutter and target position of the probability hypothesis density filter to estimate the mean and variance of posterior multi-target states.Simulation results show that the proposed method reduces the impact of sea clutter and nonlinear model for multiple maneuvering target tracking systems,and improves the estimation accuracy of the number and state of targets.4.In the maneuvering target angle tracking process,the measurement error and time-varying noise might lead to the filter divergence.Aiming at this problem,a Strong Tracking Cubature Kalman filter(STCKF)based on Variational Bayesian Inference is proposed.By using of the fictitious noise compensating technique,the angle tracking problem with the measurement error and time-varying noise has been transformed into the nonlinear filtering problem with unknown angle measurement noise statistics.To overcome the shortcoming that the state covariance matrix may not be solvable,the fading factor isintroduced into the prediction error covariance matrix,a Strong Tracking Cubature Kalman filter is presented.At last,the mean and variance of fictitious noise and system state are computed recursively by the Variational Bayesian Inference and the Strong Tracking Cubature Kalman filter,respectively.Simulation results show that the proposed method can accuracy estimate the target angle and enhances the robustness of filtering algorithm.5.Aiming at the tracking failure problem in presence of false target for the Airborne Radar tracking Air to Air multiple maneuvering targets,a multiple maneuvering target tracking algorithm based on Adaptive Spectral Segmentation is proposed.The problem of multiple target tracking with false target is described as the multi path search problem in the network cost flow framework,then the path optimization problem is transformed into an integer programming problem and solved by A* search algorithm.Considering the time-cost of using the A* search algorithm to search mass data is long and the storage space is large,a Adaptive Nystr?m Spectral Segmentation is proposed.Firstly,Nystr?m approximation theory is used to sample the measurement data and the number of clusters is adaptively selected by Matrix perturbation theory;then the measured data are clustered,and the subclasses are searched for A* search algorithm;at last trajectory set is obtained by Track Mosaic and Rauch-Tung-Striebel(RTS)smoother.Simulation results show that the proposed algorithm is effective in reducing computational time while guarantee the accuracy of track extraction.