水下弱目标跟踪的深度学习方法研究  

Research on Deep Learning Method of Underwater Weak Target Tracking

作　　者：杨家铭 潘悦 王强 曹怀刚[1] 高荪培 YANG Jiaming;PAN Yue;WANG Qiang;CAO Huaigang;GAO Sunpei(System Demonstration Research Center,Hangzhou Applied Acoustics Research Institute,Hangzhou 310023,Zhejiang,China;School of Artificial Intelligence,Peking University,Beijing 100871,China)

机构地区：[1]杭州应用声学研究所体系论证研究中心,浙江杭州310023 [2]北京大学智能学院,北京100871

出　　处：《兵工学报》2024年第2期385-394,共10页Acta Armamentarii

基　　金：浙江省“万人计划”科技创新领军人才项目(2019R52044)。

摘　　要：针对水下弱目标跟踪受干扰和噪声影响容易出现量测丢失或偏差,导致传统Kalman滤波方法跟踪误差显著增加甚至出现发散的问题,为此提出一种基于神经网络的目标跟踪方法,利用深度神经网络解决不同运动模式下目标方位跟踪的问题。水下目标跟踪的神经网路模型可通过运动模型生成大量量测数据进行充分训练,有效解决水声目标数据少、标记样本不足的问题;在量测不连续条件下,提出一种新的损失函数用于增强目标跟踪模型的稳健性;对未学习的仿真数据及实测海试数据进行测试。研究结果表明:构建的卷积神经网络(Convolutional Neural Network,CNN)模型适用于3种不同运动模式下的目标,能在平台静止和运动两种情况下稳定跟踪目标;CNN模型较传统Kalman滤波方法跟踪误差分别降低了7.75°和1.41°,验证了该模型的稳健性和可推广性。The tracking error of the traditional Kalman method is significantly increased or even divergent due to the influence of interference and noise.A neural network-based target tracking method is proposed.The proposed method uses the deep neural network to solve the nonlinear change of target azimuth with time in different motion modes.The neural network model of underwater target tracking can generate a large number of measurement data through the motion model for full training,which effectively solves the problems of insufficient underwater acoustic target data and insufficient labeled samples.A new loss function is proposed to enhance the robustness of target tracking model under the condition of measurement discontinuity.The unlearned simulation data and measured sea trial data were tested.The results show that the convolutional neural network(CNN)is applicable for a target in 3 different motion modes,and can stably track targets when the platform is stationary or moving.Compared with the traditional Kalman filtering method,the tracking error of the neural network model is reduced by 7.75°and 1.41°,respectively,for the unlearned simulation data and the measured sea trial data,which verifies the robustness and scalability of the model.

关 键 词：水下目标跟踪 深度学习 弱目标跟踪 纯方位跟踪 

分 类 号：TB566[交通运输工程—水声工程]