Bayes数据融合方法及其在水下目标检测中的数学仿真验证

对基于Bayes判决法的模式分类方法进行了理论分析，将多变量Bayes模式分类方法应用于决策信息融合过程，证明了用多传感器测得的多维信息进行决策融合性能优于用单传感器测得的一维信息。将基于Bayes判决法的数据融合方法用于水下目标检测，能大大提高系统的目标检测能力，理论分析和数学仿真均证明了该方法的有效性。

Underwater Target Detection Using Bayes Data Fusion Method

Aiming at improving navigation precision of autonomous underwater vehicle, we use multi variable Bayes method to fuse the sonar information of both the target and the remote sonar station. Eqs.(9) and (10) give the miss alarm probability (P m) and the false alarm probability (P f) respectively. The theorem at the end of Section 1 guarantees that the miss alarm probability and the false alarm probability after fusion are smaller than those before fusion (eq.14). Fig.2 gives simulation results of detection probability for three different cases: (1) target sonar information alone (P d1 ) ; (2) sonar station information alone ( P d2 ); (3) fusion of sonar information of both target and sonar station (P d) . They show preliminarily but clearly that our data fusion method improves navigation precision fairly conspicuously.