Research On Multi-Frame Track-Before-Detect Methods In Airborne Radar Systems

The airborne radar target detection and tracking technology can automatically obtain the targets of interest from a wide range of surveillance,and accurately estimate their motion state.It achieves the conversion from raw baseband data to intelligence information,and is a hot technology in the research of radar signal processing.The traditional airborne radar detect-before-track methods exist information loss after the threshold detection,which can be challenging to detect stealthy targets,low-flying targets,and small or weak targets in complex backgrounds of ground,sea,and urban clutters.The multi-frame track-before-detect(TBD)technology integrates the energy from the original echo data of multiple scans and utilizes threshold judgment to detect possible target tracks.This approach can preserve more comprehensive target information and has the capability to detect weak targets in strong cluttered backgrounds.It represents a significant direction for the development of target detection and tracking technology in radar systems.Early multi-frame TBD methods were primarily applied in fields such as groundbased radar and underwater sonars.However,due to the characteristics of airborne radar itself,there are still many issues that need to be resolved,such as motion-induced measurement mismatch,ambiguities of range-Doppler measurements for pulse Doppler radar,target maneuvering,and proximity interference of the formation target.To address these problems,this dissertation conducts theoretical analysis,methodological research,and simulation-based verification.The primary innovations of this work are outlined below:1.An adaptive multi-frame TBD algorithm for airborne radar was proposed.By establishing a nonlinear discrete grid space model under the condition of a dynamic platform,a transition cost based adaptive multi-frame measurement mapping is derived.This solves the problem of integrated energy loss caused by the nonlinear conversion between polar measurements and target motion states.The proposed algorithm effectively improved the probability of target detection and algorithm robustness.2.Two multi-PRF multi-frame TBD algorithms were proposed: one resolved the contradiction between the accurate model assumption of multi-frame TBD and the ambiguous measurements by introducing range and Doppler ambiguity factors,which achieved high probability detection and solved the ambiguity of range and velocity; The another algorithm completed single-PRF multi-frame TBD by establishing a target evolution model in the ambiguous state space,and then used a covariance cross-fusion algorithm for multiPRF fusion and ambiguity resolution.The proposed algorithm solved the problem of high data throughput in multi-PRF and multi-frame joint processing,and effectively improved computational efficiency and target tracking accuracy.3.A multi-frame TBD algorithm based on maneuvering feature adaptive estimation was proposed.By establishing a high-order state estimation model directed by historical measurements,the problem of energy loss caused by unknown target maneuvering is solved.The proposed method effectively improved the detection and tracking performance of multi-frame TBD under target maneuvering.4.A multi-frame TBD algorithm for the formation target was proposed.By establishing a multi-frame detection statistic composed of a path cost function and an echo likelihood function,the problem of detection and tracking performance loss of multi-frame TBD algorithm when multiple targets are close is solved.The proposed method avoided tracks swap or mispairing of estimated results,and effectively improved the tracking accuracy of the formation target.5.A fast multi-frame TBD algorithm based on greedy accumulation was proposed,along with its practical data processing flow.By introducing a successive-interferencecancellation strategy,the problem of strong target energy peaks masking weak targets during the multi-frame integration was solved.While the proposed algorithm reduced the computation complexity of the multi-frame TBD algorithm in airborne radar.The methods proposed in this dissertation have been validated using both simulation and real measurement data.The results show that these methods can effectively achieve high probability detection and high accuracy tracking of weak targets in complex backgrounds for airborne radars.