论文部分内容阅读
生物气溶胶施放源项参数反演是生物气溶胶袭击危害评估的反问题,对危害评估及应急响应具有重要指导意义。本文基于贝叶斯推理方法,利用生物传感器检测数据和正向大气扩散模型,构造似然函数,采用结合Metropolis-Hasting算法的马尔可夫链蒙特卡洛(Markov chain Monte Carlo,MCMC)抽样,对施放源位置、高度、施放剂量进行反演。统计分析表明,反演结果和初始源项参数设置吻合非常好,证明了方法的有效性。
Inversion of bio-aerosol application source parameters is the inverse problem of biohazard assessment of aerosol attack, which has important guiding significance for hazard assessment and emergency response. Based on the Bayesian inference method, the biosensor detection data and the forward atmospheric diffusion model were used to construct the likelihood function. Markov chain Monte Carlo (MCMC) sampling combined with the Metropolis-Hasting algorithm The source location, altitude, dose to be inverted. The statistical analysis shows that the inversion results agree very well with the initial source parameters, which proves the validity of the method.