基于遗传算法的OTHR机动弱目标检测

天波超视距(OTH)雷达系统中,为了获得较高的多普勒分辨率,通常会采用长的相干积累时间,但对于机动目标,长相干积累时间会导致回波的多普勒展宽,不利于检测。对于弱目标,由于其能量低,容易被强目标掩盖,加大了检测难度,针对这一问题,提出一种基于目标运动参数估计的OTHR机动弱目标检测方法。利用遗传算法优越的参数估计性能这一特点,采用遗传算法估计各目标的运动参数,并引入"clean"算法的思想,在时域上逐个减去强目标,以消除强目标的掩盖效应。又考虑到遗传算法的运算量较大,进一步提出采用时频分析算法估计各参数范围,减小遗传算法的运算量。仿真结果表明,与已有算法相比,文中算法具有更高的参数估计精度和弱目标检测性能。

Maneuvering Weak Target Detection Based on Genetic Algorithm for OTH Radar

In order to improve resolution of Doppler spectrum, the coherent integration time （CIT） of over-the-horizon（OTH） radar system is usually very long. But for maneuvering targets, long CIT will lead to Doppler spread, which is a disadvantage to maneuvering detection. And for weak target, due to low energy, it is easy to be covered by stronger targets. Aiming at this problem, a new genetic algorithm（GA） based method to detect maneuvering weak targets in the OTHR sys- tems is proposed. By taking the advantage of the good parameter estimation performance of genetic algorithm to estimate the targets＇ parameter, and introducing the idea of ＇ clean＇ to subtract the stronger target from the echo, the proposed algorithm has good weak target detection performance. Considering of the computation problem of GA, this paper further puts forward time-frequency dis- tribution techniques to estimate the range of parameters of targets firstly, this can reduce the compu- tational load of GA. Then, GA is used to estimate the parameters accurately. The simulation result shows that, compared to the existing algorithm, the proposed algorithm has higher estimation accu- racy of parameter and better performance of weak target detection.