基于隐Markov模型的纯方位轨迹检测方法

针对方位历程图中微弱目标信号的轨迹检测问题,分别在纯噪声和目标存在条件下推导了方位历程图中测量值的分布特性,提出了一种基于隐Markov模型的目标信号轨迹检测方法;同时,针对轨迹的起点与中止点的自主判断问题,基于序列检测提出了两种检验算法,充分利用了轨迹点的测量值分布特性和方位的连续性来提升检测性能。相比于传统的能量检测,轨迹点的检测性能提升约3dB,降低了所估计轨迹的均方根误差,同时保持了更低的虚警概率。湖试数据验证了该算法在单目标条件下的有效性。

Bearing only trajectory detector based on hidden Markov model

For the detection of the weak bearing line in the bearings vs. time record in passive sonar system, the distribution of the measurements is derived under the noise only and target present conditions respectively. Then a bearing only trajectory detector based on the hidden Markov model (HMM) is given. Meanwhile, two methods based on sequential detection are proposed to automatically decide the start point and the end point of the trajectory. These two methods fully use the distribution characteristic of the measurements and the continuity of the trajectory and performe much better than the conventional energy detection. The detection performance is improved by about 3 dB, the estimating precision of the trajectory is improved and the probability of the false alarm is low. The real experiment data verify the effectiveness of the detector when there is a single target.