论文部分内容阅读
应用贝叶斯-蒙特卡罗(Bayesian-MCMC)方法将海洋波导参数的先验信息描述为先验概率密度,结合雷达回波资料(电磁波传播损耗),得到待反演海洋波导参数的后验概率密度,用马尔可夫链蒙特卡罗(MCMC)-Gibbs采样器采样后验概率密度分布,并用样本最大似然估计值作为对海洋波导参数分布的估计.数值实验结果表明,该方法对先验信息进行了有效利用,反演精度高于遗传算法的反演精度.该方法较为充分利用先验信息,得到解的概率分布,即解的不确定性分析,这在实际应用中有一定的参考价值.
Bayesian-Monte Carlo method is used to describe the a priori information of the parameters of the marine waveguide as the prior probability density. Combined with the radar echo data (electromagnetic wave propagation loss), the posteriori parameters Probability density, the posterior probability density distribution was sampled by Markov chain Monte Carlo (MCMC) -Gibbs sampler, and the sample maximum likelihood estimation was used as the estimation of the parameter distribution of the ocean waveguide.The numerical experiment results show that this method first The information is effectively used and the inversion precision is higher than that of the genetic algorithm.The method makes full use of the prior information and obtains the probability distribution of the solution, that is, the uncertainty analysis of the solution, which has certain practical application Reference value.