Research on rock mass strength parameter perception based on multi-feature fusion of vibration response while drilling

To accurately and reliably perceive the strength parameters of rock mass, a perception method based on the combination of multi-feature fusion of vibration response while drilling and BP neural network is proposed. Firstly, the feasibility of using vibration response while drilling to reverse rock mass strength parameters is discussed by the Ls-Dyna. Afterwards, the concept of the window function is introduced, and the hybrid domain features of different sequences are extracted. The experimental results show that the fluctuation degree of the vibration signal is positively correlated with the strength of rock mass. Also, the kernel principal component analysis is used to extract the principal component with a cumulative contribution ratio greater than 85% to construct the fusion feature vector. Finally, the fusion feature vector was input into the BP neural network optimized by the Bat algorithm to identify the rock mass strength. The results show that the perception accuracy is 93.75%. Compared with other input features, the accuracy of perception through multi-feature fusion is at least 3.125% higher.