基于能量泄露目标采样模型的运动声纳空时自适应检测

Space-time Adaptive Detection (STAD) is an adaptive processing technology which combines space and time processing to achieve the goal of target detection. The STAD integrates the reverberation suppression and target detection, and has several advantages, such as high detection probability, simple structure of implementation, and flexible design procedure etc. As a result, it has been one of popular research topics in underwater acoustic field. However, all traditional STAD methods are based on an ideal target sampling model, namely, no spillover of the target energy to adjacent range cells. This is not a practical hypothesis, which would make motional sonar suffer a spillover loss with more than 3 dB. In this project, for the first time we deal with the STAD problem for motional sonar assuming a target sampling model with energy spillover. Our scheme would make full use of the spillover; not only decrease spillover loss, but also achieve accurate estimations of target distance and Doppler. This research greatly changes the implication of motional sonar STAD, namely, the STAD become a method of measuring target parameters instead of a traditional method only for target detection. For motional sonar, the new method can greatly improve its detection capability to a weak and low speed signal. Moreover, it can provide new thought and theoretical ground for detection and trace of motional sonar. These research results also have a reference to bearing-only radar STAD.

空时自适应检测（STAD）是以空时联合为框架、以目标检测为目的的自适应处理技术，它实现了混响抑制与目标检测的一体化，具有检测概率高、结构简单、设计灵活等特点，目前已成为水声领域的前沿研究课题之一。但是传统STAD均采用理想的目标采样模型，即假设目标回波能量不会泄漏到任何邻近的距离单元，这样一个过于理想的假设，将使运动声纳在实际应用中遭受3dB以上的泄露损失。本课题首创性地开展基于能量泄露目标采样模型的STAD研究，将泄露的目标能量创新地加以利用，不仅有效地减小泄露损失，还将充分利用泄露能量所提供的信息，实现对目标距离和多普勒的精确估计。这项研究将深刻改变运动声纳STAD的内涵，使其从传统的仅用于检测的方法，发展成可以对目标参量进行测量的方法。新方法能显著提高运动声纳对低速弱信号的探测能力，为运动声纳的检测和跟踪手段提供新思路和理论依据。相关研究成果对机载雷达STAD也具有一定的借鉴作用。

空时自适应检测（STAD）是以空时联合为框架、以目标检测为目的的自适应处理技术，它实现了混响抑制与目标检测的一体化，具有检测概率高、结构简单、设计灵活等特点，目前已成为水声领域的一个前沿研究课题。但是传统STAD均采用理想的目标采样模型，即假设目标回波能量不会泄漏到任何邻近的距离单元，这样一个过于理想的假设，使运动声纳在实际应用中遭受3dB以上的泄露损失。为解决该问题，本课题首创性地开展基于能量泄露目标采样模型的STAD研究，提出新的运动声纳目标采样模型，实现对目标泄露信息的充分挖掘；在此基础上，提出一整套适用于运动声纳的稳健STAD方法。新方法显著提高了运动声纳对低速弱信号的探测能力，为运动声纳的检测和跟踪手段提供新思路和理论依据。该项研究还丰富了运动声纳STAD的内涵，使其从传统的仅用于检测的方法，发展成可以对目标参量进行测量的方法。相关研究成果对机载雷达STAD也具有很大的借鉴意义。