基于贝叶斯网络的水下目标识别

Underwater Target Recognition Based on Bayesian Network

针对水下目标的识别问题,研究了基于贝叶斯网络的水下目标识别方法。文中介绍了贝叶斯网络基础,水下目标识别主要因素,建立了水下目标识别贝叶斯网络模型,并给出了条件概率表和推理求解过程。利用贝叶斯网络进行水下目标识别能很好地对多种证据进行综合评判,而且能很好地表达专家知识,特别适合水下复杂环境目标识别。仿真试验分析证明了论文方法的有效性。

To solve the problem of underwater target recognition,a method based on Bayesian Network is proposed in this pa⁃per.The basic BN theory and the factors associated with underwater target recognition are analyzed.The BN model of underwater tar⁃get recognition,the conditional probability table and the process of probability inference are established.BN is an effective model with expert experience knowledge for multiple evidence application problem,which is useful for underwater target recognition.The simulation experiment result verifies the effective of the proposed method.