Research On Maneuvering Target Tracking Algorithm In Doppler Radar

With the development of target tracking technology,Doppler radar is widely used in maneuvering target tracking.Tracking and defending the maneuvering targets with time-varying motion modes are challenging tasks for radar system,which leads to actually a non-linear state estimation problem in filtering.This paper will focus on the issues of non-linear measurements processing and adaptive maneuvering target tracking methods.By reducing the non-linear effects of Doppler measurement,the tracking accuracy and stability of filter can be improved in maneuvering tracking cases.The main research contents and contributions of this paper include:1.The current statistical model algorithm multi-model filtering algorithm and interactive multi-model algorithm are studied and the tracking performance of the above algorithms are verified,moreover,the recommendations of algorithms are analyzed in different simulation cases.2.For the non-linear problem of Doppler measurement,a sequential filtering algorithm for maneuvering target tracking is proposed,in which the statistics of converted measurement errors are calculated based on the predicted information to avoid the drawbacks that based on the measurement information.By using the LOS method,the unbiased estimation of the non-linear system with the cross-range rate information is proposed.The algorithm can improve the tracking accuracy with a small increase in computational complexity.3.The best linear unbiased estimation is studied,the Kalman filter form of BLUE is deduced and a sequential filter based on BLUE with adaptive pseudo information algorithm is proposed.The performances of pseudo-state and pseudo-measurement are estimated in real-time,and the filter parameters are adjustied adaptively,then the optimal target tracking with Doppler measurement is achieved.This algorithm can overcome the problem of long pre-convergence time when using pseudo-state.The calculation burden is less than that based on the statically fused method.4.For the problem of multiple maneuvering target tracking in non-linear measurement and clutter environment,a multi-model CM-GMPHD algorithm is proposed.When the Doppler measurement is available,two improved algorithms of multi-model SQD-GMPHD and multi-model SQ-GMPHD are proposed.Compared with SQD-GMPHD algorithm,the pseudo measurement is constructed to reduce the non-linearity in the multi-model SQ-GMPHD algorithm,which improve the accuracy of the estimated target number and state.The results show the algorithms proposed above can achieve effective multi-maneuvering target tracking with non-linear complex measurements.The performance of the proposed algorithms are demonstrated via simulations results.Compared with the conventional algorithms,the maneuvering target tracking performance in non-linear measurement and clutter environment is improved,which is useful to the certain practical application in engineering.