高机动雷达目标快速长时间混合积累算法

Fast Long-time Hybrid Integration for Highly Maneuvering Radar Target Detection

  • 摘要: 长时间积累是提高雷达微弱目标检测性能的有效方法,但高速高机动目标的跨距离单元和跨多普勒单元现象严重限制了积累性能。为此,本文提出了一种长时间混合积累的新方法,将积累时间划分为若干个子孔径,在子孔径内同一个距离单元相参积累,在子孔径间实现高效的跨距离单元非相参积累。该方法中采用启发式搜索策略,有效地解决了遍历搜索中运动模型阶数增长带来的运算量爆炸问题。同时,其可通过高阶包络补偿有效降低信噪比损失。数值实验表明,新方法较现有混合积累方法可显著改善检测性能和降低运算量。

     

    Abstract: Long-time integration is an effective method to improve weak radar target detection performance. However, the problems of across range unit (ARU) and across Doppler unit (ADU) of high-speed and maneuvering targets severely limits the integration performance. A new long-time hybrid integration approach is proposed in this paper. It divides the coherent process interval (CPI) into several sub apertures, and implements coherent integration in the same rang unit for each sub aperture. Subsequently, it implements non-coherent integration among several sub apertures with compensation of ADU effect. Furthermore, it applies a heuristics search strategy to reduce the huge computational complexity with increase of the motion order. Besides, it can decrease the signal-to-noise ratio losses efficiently by compensating high-order envelope shift of moving targets. Finally, some numerical experiments are provided to illustrate the detection performance improvement as well as the computational burden reduction of the proposed method.

     

/

返回文章
返回