基于Bayesian-MCMC估计的隐身飞机RCS模型优化

发布时间：2018-01-18 04:31

  本文关键词：基于Bayesian-MCMC估计的隐身飞机RCS模型优化　出处：《北京航空航天大学学报》2016年04期 　论文类型：期刊论文

  更多相关文章： 隐身 雷达散射截面(RCS) 起伏模型 贝叶斯-蒙特卡罗 拟合优度检验

【摘要】：对隐身飞机的雷达散射截面(RCS)统计建模时,传统方法通过直接计算RCS样本的统计特征估计模型参数,可能会产生较大的拟合误差。本文提出采用贝叶斯-蒙特卡罗(Bayesian-MCMC)方法提高起伏模型的参数估计精度,从而减小模型的拟合误差。首先将卡方分布模型和对数正态分布模型进行贝叶斯推导,得到其特征参数的后验估计表达式。然后采用MCMC算法构造后验分布的马尔可夫链,从而计算特征参数的估计值。最后通过比较2种方法的拟合曲线及其误差可知,本文方法适用于2种起伏模型,模型参数的估计误差比收敛误差门限值低1~2个数量级,2种分布模型的拟合精度均提高50%以上。
[Abstract]:In the statistical modeling of radar cross section (RCS) of stealthy aircraft, the traditional method directly calculates the statistical characteristics of the RCS samples to estimate the model parameters. The Bayesian Monte Carlo Bayesian-MCMC method is proposed to improve the parameter estimation accuracy of the undulating model. In order to reduce the fitting error of the model. Firstly, the chi-square distribution model and the logarithmic normal distribution model are derived by Bayesian method. The posteriori estimation expression of the characteristic parameters is obtained, and then the Markov chain of the posteriori distribution is constructed by using the MCMC algorithm. Finally, by comparing the fitting curves and their errors of the two methods, we can see that this method is suitable for two kinds of undulating models. The estimation error of the model parameters is 1 ~ 2 orders of magnitude lower than the convergence error threshold. The fitting accuracy of the two distribution models is improved by more than 50%.
【作者单位】： 电子科技大学电子工程学院;中国电子科学研究院;
【基金】：国家“863”计划(2012AA01A308) 国家“973”计划(613206)~~
【分类号】：TN953;V218
【正文快照】： 网络出版地址:www.cnki.net/kcms/detail/11.2625.V.20150917.1650.009.html引用格式:代小霞,曹晨,冯圆.基于Bayesian-MCMC估计的隐身飞机RCS模型优化[J].北京航空航天大学学报,2016,42(4):851-857.DAI X X,CAO C,FENG Y.Optimization on stealth aircraft RCS models using B，

本文编号：1439479