Rock Mesoscopic Parameter Determination Method Based on BP Algorithm
[Purposes]It was very common to use PFC to simulate the rock failure mechanism,and the ra-tionality of the mesoscale parameters used in the particle flow model directly affected the numerical simulation results.Therefore,aiming at the problems such as the variety,complex influence and difficult analysis of program PFC meso-scale parameters,an intelligent prediction method based on BP algorithm was proposed to accurately predict model meso-scale parametric on the basis of accurate basic macro pa-rameters.[Methods]Sufficient data samples are obtained by PFC numerical simulation test.A neural network prediction model is established by combining the data samples and BP algorithm,and the pre-dicted values of PFC microparameters were obtained according to the input of actual meso-scale param-eters.[Findings]By comparing and analyzing the actual values and inverse values of the meso-scale pa-rameters of three different rock specimens,it is found that the relative errors between them are small.[Conclusions]The results show that the method is reasonable and feasible,and can be used to quickly determine various mesoscopic parameters of PFC.
NETL
NSTL
万方数据
