A multi-information intelligent identification method of rock mechanics parameter and application in underground engineering
Jiaxin Zhang, Xiaohong Guo, Xiaohua Ke, Wei Liang · 2009
To solve the high nonlinear mapping relationship between mechanical parameter and rock performance in underground engineering and to improve the uniqueness of optimal result in parameter recognition, a new multi-information intelligent identification method of rock mechanics parameter is proposed. By coupling neural network and genetic algorithm as an evolutionary algorithm, the global and nonlinear optimization search of mechanical parameters in the method is realized. And by building the associated fitness function to absorb multi-information, the method can improve the uniqueness of searched objective as far as possible. Application in Wuyang tunnel, the key tunnel in Xia-Rong expressway engineering, indicates that the intelligent method is very reliable and efficient in identification of rock mechanics parameter and in assistant design of underground engineering.