基于扩展模型的高分辨雷达目标检测前跟踪

Target Detection and Tracking of High Resolution Radar Based on Extended Model

在雷达目标检测优化中,由于高分辨雷达会随目标距离、方向的变化而变化,具有很大的随机性,需要进行实时判断,导致传统的基于目标运动连续性原理的高分辨雷达目标检测前跟踪算法,适应性不强,造成跟踪发散。为此提出一种采用目标扩展模型的高分辨雷达目标检测前跟踪算法,给出目标的动力学模型及高分辨雷达的观测方程。通过空间分布模型描述扩展目标,采用空间分布概率形式检测目标散射中心的可能位置,给出目标上点源的空间扩展分布函数。通过目标初始存在概率及存在状态转移矩阵对下一时刻粒子的存在状态进行预测。依靠目标的初始分布进行采样获取新生粒子的状态预测,完成对粒子权重的计算,对目标进行检测和状态估计,从而实现目标跟踪。仿真结果表明,所提算法具有很高的跟踪性能。

A target detection and tracking algorithm of high resolution radar is put forward based on a target expanded model. A dynamic model of target and an observation equation of high resolution radar are given. The extension target is described by a spatial distribution model, a possible position of the target scattering center is obtained via the space distribution probability, and a spatial extension distribution function of point source on the target is given. The existing state of next particles is predicted through the initial target existing probability and the existing state transition matrix, the state of new particle is predicted according to the sampling of the initial distribution of the target, the particle weights are calculated, and the target is detected and the state is estimated, so as to realize the target tracking. Simulation results show that the proposed algorithm has high tracking performance.