The Improvement Of Genetic Algorithm And Its Application Of The Deformation Analysis Joint Inversion

In order to solve the parameter inversion problem which is in the fields of Deformation Monitoring, Geodynamics and Geotechnical Engineering, this paper tries to use Genetic Algorithm and the Finite Element Method to joint inversion. At the same time, in view of the traditional Genetic Algorithm has some defects of “premature convergence” and low local searching efficiency, the Genetic Algorithm is improved by using simulated annealing. Generally, Genetic Algorithm has the advantages of strong robustness, implicit parallelism, global optimization, and the Finite Element Method has accuracy and efficiency when it analyses mechanics problem. So combining them can inverse parameters effectively. The main work is as follows:(1) With the analysis on the deficiency of traditional Genetic Algorithm, the improvement of the crossover and mutational operator has been proposed by using Simulated Annealing Algorithm(SA), which can make the implicit parallelism and global optimization of Genetic Algorithm stronger.(2) The structural parameter of some simulation example is inversed by using genetic algorithm and FEM. The result shows that the joint inversion method can inverse elastic modulus, Poisson's ratio, boundary load and so on, and the improved joint inversion method has higher efficiency and stronger convergence ability.(3)Based on the observations of the measuring point displacement, the structure stress is analyzed by using forward method and inversion method. The results show that the defect of the large stress concentration near the measuring point, which is caused by adding the observations of the measuring point to the model directly when we use the forward method, can be overcome by using the new inversion method from this paper. Besides, the result of inversion method is closer to the theoretical value than the forward method.(4) The physical parameters of Xiangjiaba Dam are inversed by using the method presented in this paper. At the same time, a model of predicting displacements of the dam is established. The calculated value of the displacement can match the observed value well.So the model can provide some effect in the safety monitoring of Xiangjiaba Dam.