Parameter Identification of 3D Elastic-Plastic Model for Tunnel Engineering Based on Improved Genetic Algorithm

Abstract

In order to overcome the deficiencies of the existing intelligent displacement back analysis methods, the authors propose a parameter identification method for the 3D elastic-plastic model of tunnel engineering based on an improved genetic algorithm. An improved SVR algorithm is introduced, which solves this problem by transforming the multidimensional output variable regression into a multilayer standard one-dimensional output variable regression; combined with the decimal-coded genetic algorithm, an improved GA-SVR algorithm is formed, and genetic algorithm is used to search for optimal SVR model parameters, in order to establish the optimal nonlinear mapping relationship between the parameters to be identified and the displacement. The genetic algorithm is used to carry out the optimal identification of the parameters to be identified. In order to compare the effect of this improved GA-SVR algorithm, the genetic algorithm is combined with the BP neural network to form the GA-BP algorithm and compile the corresponding calculation program. The two algorithms are applied to the intelligent identification of the parameters of the 3D elastic-plastic model of the same tunnel engineering. The experimental results show that the maximum relative error of the 6 parameters to be identified is 18.51%, only 2 are over 15% and 14 are within 10%, and the average relative error of the 6 parameters is only 10.32%, the minimum is 2.63%; for the GA-BP algorithm, the maximum relative error of identification is 22.73%, 4 are more than 15% and 9 are within 10%, and the average relative error of 6 parameters’ identification is up to 13.91%, and the minimum is 4.64%. The improved GA-SVR algorithm can achieve higher identification accuracy and better calculation efficiency than the GA-BP algorithm and can be used in the identification of similar geotechnical engineering parameters.

Thermal Parameters Inversion Method for Concrete Dam Based on Optimal Temperature Measuring Point Selecting

“…In addition, the quality of the inversion results is also related to the optimization method. A variety of optimization algorithms, such as artifcial bee colony (ABC) [34], PSO [35], gray wolf optimization (GWO) [36], GA [37], whale optimization algorithm (WOA) [38], and bat algorithm (BA) [39], applied in the model parameter identifcation have shown good adaptability. Compared with other methods, PSO has the characteristics of simple implementation, wide application, and strong global search capability.…”