通过训练自适应神经网络构造水雷识别分类决策器,突破了水雷以往常采用的人为确定阈值法对样本数量及特征量维数、相关性等的限制。采用神经网络的高容错性解决目标浅水频谱结构的不稳定性带来的识别结果的不稳定性问题,有效解决了“锚雷声抗扫”难题。 By training self-adaptive neural network to construct mine identification and classification decision maker, this paper breaks through the limit of sample quantity, dimension and correlation of man-made determination thresholding used by mines. By adopting the high fault tolerance of neural network to solve the problem of instability of recognition result brought by the instability of the target shallow water spectral structure, the problem of “anti-sweep anchor anti-sweep” is effectively solved.