基于模糊理论的软岩黏弹性模型识别及参数反分析

VISCOELASTIC MODEL IDENTIFICATION AND BACK ANALYSIS OF PARAMETERS OF SOFT ROCK BASED ON FUZZY THEORY

对岩石黏弹性模型进行总结,将复杂的黏弹性模型方程转化为统一的线性方程,基于实测数据运用回归分析法和模糊数学方法,通过对线性方程参数的反分析得到统一模型的参数,运用模糊关联度分析法对具体岩石黏弹性模型进行识别,进而根据统一模型参数可反求出黏弹性模型方程的参数。研究表明,对黏弹性本构模型采用线性化处理,用统一的线性方程表示复杂的本构方程简化模型的识别和参数的获取过程。最后利用一矿山软岩流变试验结果对其黏弹性模型进行识别和模型参数进行反分析。计算结果表明,样品Ⅰ-1,Ⅰ-2黏弹性模型服从麦钦特体,样品Ⅱ-1,Ⅱ-2黏弹性模型服从H-K体,通过计算出黏弹性模型参数,得出软岩黏弹性模型的回归方程,所得结果与实际情况比较一致。

The determination of parameters in viscoelastic constitutive equations of soft rocks is very important to analyze the time-dependent behaviors of embankments, foundations and mines in soft rocks. The complex viscoelastic model equations can be transformed into unified linear equations, and the parameters of the model can be got by back analysis method. But the general back analysis method doesn＇t take the randomicity and fuzziness of the parameters into account, therefore, it does not be in accordance with the fact. In the general back analysis method, the membership function is introduced. The methods of fuzzy back analysis and viscoelastic model identification with fuzzy correlation degree are proposed. The method can identify the models such as Kelvin model, H/M model, H-K model, M-K model and Mechant model, and can also get their parameters. On the basis of the creep experiments of two groups of salt rock from a mine factory, the method is used. The results show that the samples of group Ⅰ are Mechant bodies and the samples of group Ⅱ are H-K bodies. The parameters of the two models got by back analysis method coincide with the test results very well.