摘要
基于假设检验的方法,研究了目标检测与定位中的传感器管理方法,给出了一种最大后验概率法,讨论了另外两种基于贝叶斯推理的最大正确检测概率法和最小代价函数法。通过仿真改进了最大正确检测概率法,对三种不同方法的检测正确率和平均采样次数进行了比较,分析了各种方法的适用场合。
Based on hypothesis testing, methods of sensor management used in target detection and localization are studied. The detector used in this problem can operate in 揻ocused mode?and 揵road search mode? The former mode offers higher detection and localization accuracy but less coverage area than the latter. It is supposed that a signal source is to be detected and localized with a sequence of tests, each may use different mode. The goal of sensor management is to build an object function for selecting proper mode in the sequence of tests in order to improve the detection performance. In this contribution, A method of maximum a posteriori probability is presented, and methods of maximum correct detecting probability and minimum cost function based on Bayesian Reasoning are discussed. The performance of the three methods are analyzed and compared. The method of maximum correct detecting probability is modified through the simulating process.
出处
《系统仿真学报》
CAS
CSCD
2004年第12期2805-2808,共4页
Journal of System Simulation
关键词
多传感器管理
目标检测
假设检验
贝叶斯推理
代价函数
multisensor management
target detection
hypothesis testing
Bayesian reasoning
cost function
