Radar Multi-Frame Joint Detection And Tracking Technology For Maneuvering Target

With the rapid development of unmanned intelligent devices,radar systems play an irreplaceable role as the "eyes" of perception in both military and civilian fields.The key to a radar system is accurate target detection and tracking.Multi-frame joint detection and tracking is such a technology that collects target energy from multiple original echoes by utilizing target motion features.Since noise exists randomly,the final detection statistics obtain an improvement in signal-to-noise ratio,thus enhancing the detection performance of weak targets.However,existing multi-frame joint detection algorithms are mainly based on the assumption of uniform motion models.For other real-world motion types such as turning and maneuvering,research is still relatively limited,and the theoretical methods,engineering implementations,and practical problems of multi-frame joint detection algorithms based on these motion models need to be further improved.For instance,inaccurate descriptions of target maneuvers,errors in the accumulation process of value functions due to state space discretization,and the explosion of dimensions caused by increased state dimensions due to target maneuvers are still challenging issues.This paper addresses the above issues in the multi-frame joint detection algorithm for maneuvering targets through the following main work:1.A multi-model interactive approach for motion modeling of maneuvering targets was proposed to address the inaccurate description issue of target maneuvers.This method improved the accuracy of target state prediction and enhanced the robustness of estimating maneuvering targets compared to traditional methods.2.To solve the error accumulation problem caused by discretizing the state space,a greedy algorithm based on Kalman filtering was proposed for target state estimation.This algorithm effectively enhanced the ability to accumulate target energy over multiple frames.3.To prevent the dimension explosion problem caused by increasing the state space dimension,a greedy-based fast value function accumulation algorithm was proposed on the basis of the greedy algorithm.This algorithm solved the measurement search problem during the process of accumulating target energy and significantly improved the computational speed of the multiple-frame accumulation algorithm.4.A maneuvering target tracking strategy based on maneuver detection was proposed.Compared with the multi-model interactive approach,this strategy achieved effective target energy accumulation under lower state space dimensions.It addressed the dimension explosion problem from the perspective of maneuver modeling and significantly improved the real-time performance of the multiple-frame joint detection and tracking algorithm for maneuvering targets.The multi-frame joint detection and tracking algorithm mentioned above has been verified through relevant simulations.The simulation results confirm its effectiveness.