基于动态规划的TBD改进算法

Improvement on Track-Before-Detection Algorithm Based on Dynamic Programming

针对在低信噪比条件下基于动态规划的检测前跟踪(TBD)算法在给定的有限帧数情况下可能无法检测到目标的问题,提出了一种基于动态规划的新算法,即在确定目标状态转移时,应用后续观测值对状态转移进行修正.在该算法中,为了减小搜索范围及减少伪航迹,又引入了方向加权的思想,即重点考虑目标存在的可能区域.理论分析和仿真结果表明,新算法的检测性能优于一般的基于动态规划的TBD算法,得到的目标航迹也更加精确.

Aimed at the problem that the target can not be detected using the dynamic programming-based TBD algorithm in the case of the given numbered frames under the circumstance of low signal-to-noise ratio （SNR）,a new dynamic programming-based algorithm was proposed,namely the state transition was modified by using the post observed value when determining the state transition of target. For purpose of reducing the searching range and false track,the idea of weighted direction was introduced in the new algorithm,namely the range in which a target is possibly existing was taken into account emphatically. By theoretical analysis and simulation it shows that the detection performance of the new algorithm is superior to that of the typical dynamic programming-based TBD one,and the track obtained is also more accurate.