多亮点声纳目标轨迹起始技术

高分辨率声纳目标具有多亮点分布特性,传统轨迹起始算法很难有效对其轨迹起始。针对声纳目标多亮点分布特性,设计了基于蒙特卡洛方法的轨迹起始算法,其根据跟踪门内每个分布亮点的特性,估计目标的概率分布,推出目标位置信息。由于轨迹起始算法考虑了声纳目标的多亮点属性和每个亮点的信息,并在轨迹起始过程中没有将速度、加速度、角度等作为轨迹起始的限制条件,而是在确认阶段,作为限制条件,对不符合逻辑的轨迹进行淘汰,对符合逻辑的进行动静分类。所以此轨迹起始算法克服了传统算法容易将多亮点目标起始为多个轨迹或轨迹起始错误的缺点。用不同方法对多亮点分布声纳目标数据进行了轨迹起始,结果验证了该方法的有效性。

Research on track initiation of multi-highlight sonar target

The high resolution sonar target has the characteristic of multi-highlight. And it is difficult for the general track initiation algorithms to initialize the track of multi-highlight sonar target effectively. A method of track initiation based on Monte Carlo is proposed. The method deduces the location of sonar target from its probability distribution, which is obtained by the information of highlights of sonar target. Since the algorithm takes account of the multi-highlight characteristic of sonar target and the information of every highlight, and only takes account of the target velocity, motion angle on the stage of track acknowledgement when the algorithm has washed out the illogical track and classified the motional and the motionless tracks, the initiation method overcomes the shortcomings of general ones, those maybe initialize a wrong track or more than one track for a multi-highlight sonar target. By processing high resolution sonar data, it is shown that this method is more effective in initializing the track of a multi-highlight sonar target than the ＇Nearest Neighbor＇ algorithm.