Study On Grey Identification And Prediction Method Of Rockmass Mechanics Parameters

Rockmass mechanics is a branch of mechanics which studies the deformation and failure law of rock mass under various force fields and its practical application.Its applications involve mining,civil engineering,water conservancy and hydropower,petroleum,underground engineering,military engineering and many other engineering fields.At present,identification and prediction of rockmass mechanics parameters is still an important research subject in the field of rock mechanics.The mechanical parameters of rockmass can be roughly divided into three categories:rheological parameters,deformation parameters and strength parameters,among which the accurate performance of rheological parameter identification can explain the aging deformation and failure characteristics of rockmass reasonably;the accurate prediction of deformation parameters and strength parameters has a very important influence on the stability,safety and economy of rockmass engineering.In the practice of rockmass engineering,there are many engineering accidents caused by insufficient understanding of the mechanical properties of rock masses.How to estimate rockmass mechanical parameters reasonably is related to the feasibility of rockmass engineering projects directly,and has very important theoretical and practical significance.At present,the commonly used rockmass mechanical parameter identification and prediction methods do not consider the three characteristics of rockmass mechanics at the same time: availability and integrity of rock samples are limited by destructive testing indoors or outdoors with little experimental data;in different physical environments,the deformation and failure of rockmass under various stress states lead to uncertainties in many mechanical properties;the rockmass is cut by multiple sets of structural planes,which has obvious discreteness.These three characteristics are typical grey system characteristics.Therefore,it is reasonable and feasible to use grey system theory to discuss the identification and prediction of rockmass mechanical parameters.The main work of the thesis is as follows:According to the grey system characteristics of rheological parameter identification of generalized Kelvin model,such as less information and uncertainty of experimental data,a new grey identification method for rheological parameters in generalized Kelvin model is proposed based on the principle of grey difference information.The new method uses the grey accumulation operator to establish the grey-generalized Kelvin model,and uses the accumulation method to obtain the relationship between the grey parameters and the mechanical parameters.The rheological parameters of the generalized Kelvin model were identified by the new method using the stress-strain-time relationship curves obtained from uniaxial compression and salt rock creep mechanics experiments with different concentrations of permeate.The four evaluation indexes and three traditional parameter identification methods are used to compare and analyze,and it is found that the rheological parameters identified by the new method can correctly reflect the creep of salt rock at various stages,and can better solve the problem of rheological parameter identification of salt rock.According to the grey characteristics of Burgers model rheological parameter identification and the difficulty of selecting initial parameter values,a new grey identification method for rheological parameters in Burgers model is proposed based on grey whitening differential equations.The new method combines the creep equation of the Burgers model with the GM(1,1)model to establish the grey-Bergers model,and introduces the grey evolution algorithm to optimize the background value,and uses the least square method to establish the relationship between mechanical parameters and grey parameters.Combined with uniaxial,triaxial compression and graded loading rock creep tests commonly used in engineering,and according to the measured displacement-time loading curve,a new method is used to identify the parameters of the Burgers rheological model.Compared with the five commonly used parameter identification methods,it is found that the rheological parameters identified by the new method are more in line with the mechanical characteristics reflected by the Burgers model,and can better describe the viscoelastic creep characteristics of rock mass materials.Aiming at the discreteness of discontinuous surfaces such as a large number of faults and joints developed in the rock mass and the prediction of rock mass deformation parameters with little information characteristics,combined with the multiplicity of rock mass deformation modulus and the equal interval grey convolution model,an equi-gap grey prediction model for rock mass deformation parameters is established.The new model utilizes the strong correlation between deformation modulus and longitudinal wave velocity,combined with the modeling idea of grey multivariate prediction model,and discusses the parameter estimation and analytical expressions of the new model.The validity of the new model is verified by a numerical case of deformation modulus fitting.According to the results of field in-situ test and acoustic wave test,the prediction of the engineering rock deformation modulus of the dam foundation of a hydropower station in Tibet and Jin’anqiao hydropower station in Yunnan Province is discussed.The results are consistent with the law that the rock mass longitudinal wave velocity increases with the increase of the tunnel depth,which can provide reliable data information for the stability analysis of rock mass engineering.Aiming at the prediction problem of strength parameters with unequal interval characteristics of parameters such as uniaxial compressive strength and Brinell hardness,based on the multivariate characteristics of strength parameters,the equidistant grey convolution model is extended to the case of unequal intervals,and a non-equigap grey prediction model for rock mass strength parameters is established.The validity of the new model is verified by an example of tensile strength prediction,and according to the in-situ direct shear test of rock mass,the prediction problem of shear strength parameters cohesion and internal friction angle of the dolomitic limestone rock mass in the dam foundation of a reservoir in Shanxi Province is discussed.The prediction result has high precision and can estimate the development of the plastic zone of the rock mass in the state of unequal interval of uniaxial compressive strength,which has important engineering practical value.