基于先验信息的多假设模型中断航迹关联算法

针对经典的中断航迹关联算法在机动目标环境下航迹预测准确性差、关联效果恶化严重的问题，提出基于先验信息的多假设运动模型中断航迹关联算法。所提算法充分利用目标属性、目标运动特征、使用场景等先验信息，基于多假设思想，建立多种可能的目标运动模型并实施航迹预测，基于位置和速度信息的模糊相关函数描述预测航迹与新起始航迹的模糊匹配关系，最后基于多项式拟合原理连接满足关联关系的新、老航迹。经仿真验证，在中断区间目标发生机动运动的条件下，所提算法的关联效果相对于经典的中断航迹关联算法有显著提升。所提算法对于复杂环境具有较强的适应能力，经50次蒙特卡罗仿真，在中断时间小于18个滤波周期条件下，机动目标的平均正确关联率达到90%以上，机动环境的全局关联正确率达到85%以上。

Track segment association algorithm based on multiple hypothesismodels with priori information

As tracks forecasting and associating accuracy of the traditional track segment association algorithms deteriorates seriously in maneuvering targets environment, a new algorithm based on multiple hypothesis motion models with priori information is proposed. The algorithm firstly builds multiple hypothesis motion models for tracks forecasting according to the priori information, for instance target property, target motion features, scenario condition, then describes the matching relations between forecasted old tracks and new tracks according to fuzzy correlation function on location and velocity information. Finally, the associated track segments on the basis of polynomial fitting connected. Simulation results showed that in the maneuvering targets scenario, the proposed algorithm remarkably outperformed the traditional track segment association algorithm. The proposed algorithm is suitable for complicated environment, after 50 times Monte Carlo simulation, when the break interval is less than 18, the average correct association rate of the maneuvering targets is more than 90%, and the global correct association rate is more than 85%.