摘要
全局最优量测数据关联算法是无源定位技术中的关键问题。针对三站无源定位系统全局最优数据关联的三维(3-D)分配问题,提出一种新的直接求解算法——启发式消元算法。该算法通过直接对代价矩阵的元素进行启发式消元,得到分配问题的最优解。给出算法相关原理的分析和证明以及算法的具体步骤。并将其应用于三部光电被动传感器多目标无源定位系统两级数据关联算法的第二级关联过程中,对启发式消元算法和求解3-D分配问题的近似最优求解算法——拉格朗日松弛算法的算法复杂性进行了分析对比。理论分析和仿真实验均表明,该算法较适合与两级关联算法联合使用。与拉格朗日松弛算法相比,该算法计算量小,便于实现,是一种新的、有效的全局最优数据关联算法。
The global optimal data correlation algorithm is the key problem for the multi-passive-sensor location system. Aiming at the three-dimension （3-D） assignment problem of data correlation for three-passive-sensor location system, this paper presents a new direct solution method-heuristic elimination algorithm. This method obtains the optimal solution by eliminating the elements in the correlation cost. The analysis and certification of the relative theory are given, and the concrete process is also described in detail. Furthermore, the new method is used in the second correlation process of the 2-stage correlation algorithm for the 3-passive-sensor multi-target location system. The complexities of both the new algorithm and the Lagrangian relaxation algorithm, which is a saboptimaI solution method to solve the 3-D assignment problem, are analyzed and compared. Theory analysis and simulation result show that the heuristic algorithm is more suitable to the 2-stage correlation algorithm. Compared with the Lagrangian relaxation algorithm, this method has less amount of calculation and is easier for us to achieve. So it is a new and valid data correlation algorithm of passive-sensor location system.
出处
《系统工程》
CSCD
北大核心
2006年第4期118-122,共5页
Systems Engineering
基金
全国优秀博士论文作者专项基金资助项目(200036200237)
关键词
分配问题
启发式消元算法
拉格朗日松弛算法
数据关联
Assignment Problem
Heuristic Elimination Algorithm
Lagrangian Relaxation Algnrithm
Data Correlation
