基于区分度增量和CPCRLB的传感器管理算法  

Sensor management algorithm based on discrimination gain and CPCRLB

作　　者：徐瑞阳[1] 冯新喜[1] 孔云波[1] 鹿传国 

机构地区：[1]空军工程大学信息与导航学院 [2]95806部队

出　　处：《传感器与微系统》2015年第9期138-140,144,共4页Transducer and Microsystem Technologies

基　　金：国家博士后基金资助项目(2014M562585)

摘　　要：针对传感器管理中存在的能量和资源有限问题,提出了一种基于区分度增量和条件后验克拉美—罗界(CPCRLB)的传感器管理算法,首先推导了CPCRLB在无源传感器粒子滤波情况下计算公式,然后采用条件后验CPCRLB作为矩阵遗传的适应值对传感器分配方案进行选择,同时计算每个目标的区分度增量,合理分配传感器资源。仿真结果表明:该算法能较好地实现传感器的分配,并降低了整个跟踪过程中的能耗。To solve the energy limitation problem in sensor management,an algorithm based on discrimination gain and CPCRLB is proposed. Deduce calculation formulas to calculate CPCRLB for negative sensor with particle filtering,adapts CPCRLB as the fitness of matrix genetic algorithm. The discrimination gain of each target is calculated to allocate the sensor resources properly. Simulation result shows that the proposed algorithm can achieve a good performance in sensor assignment and reduces energy consumption in target tracking.

关 键 词：传感器管理 多传感器多目标跟踪 区分度增量 条件后验克拉美—罗界 矩阵遗传算法 

分 类 号：TP274.2[自动化与计算机技术—检测技术与自动化装置]