Global sampling method based on differential evolution

The invention discloses a global sampling method based on differential evolution, which takes an aeronautical turbine engine as a research object, researches are carried out aiming at the rotor and stator gap variation of a structure, and uncertainty quantization is realized and is used as a main target of the global sampling method. According to the method, the differential evolution algorithm and the Metropolis-Hastings algorithm are combined, meanwhile, the single Markov chain sampling process of a traditional algorithm is improved, and calculation of posterior distribution is carried out by adopting multiple Markov chains. According to the method, a sampling method in the differential evolution algorithm is applied to the sampling process, and the scaling factor and the random walk process of the differential evolution algorithm are improved and are not limited to the assumption of Gaussian distribution. Compared with a traditional random walk Metropolis-Hastings method, the method has the advantages that