基于贝叶斯因子分析的声纳探测误差模型研究

摘要

The sonar detection error modeling is the core technology in sonar simulation. To model the sonar detection error, after using statistic method to analysis lots of sonar detection data, a method based on Bayesian factor analysis is proposed to find the sonar detection error model. This error model is used to forecast the detection error in a sonar simulation. The result of comparing with conventional Gaussian noise model shows that the proposed model has high performance on the sonar detection data.%声纳的探测误差模型是声纳仿真的核心技术.针对建立某型声纳探测误差模型的问题,在对大量声纳实测数据进行统计分析的基础上,提出采用贝叶斯因子分析的方法,建立声纳的探测误差模型.利用该误差模型预测误差,将预测结果应用于声纳仿真中,并与白噪声模型仿真方法进行比较.仿真结果表明,基于贝叶斯因子分析的声纳误差仿真模型可以更有效的对声纳探测误差进行建模和预测.