Abstract: The geotechnical mechanical parameters are the main basis for geotechnical engineering design and construction.In order to obtain the geotechnical parameter information that is more in line with the actual situation, a displacement back analysis method based on the support vector machines, faster and more accurate parameter determination method, is proposed.Based on the grey wolf algorithm, a swarm intelligence algorithm developed in nature bionics, and combined with the crossover mutation operation of the differential evolutionary algorithm, the grey wolf algorithm is evolved, and the hybrid grey wolf optimization algorithm is obtained, which is used to reverse the geotechnical mechanical parameters after optimization of the support vector machine parameters.Based on the subway tunnel project in Kunming, the Midas/GTS finite element is used to establish a three-dimensional simulation model to simulate the tunnel construction process.Based on the measured surface deformation of the top section of a tunnel, the support vector machine model optimized by the hybrid differential evolution grey wolf algorithm is used to invert the mechanical parameters of rock mass.The settlements calculated by the back analysis method and the traditional Peck formula method are compared with the meas surface ones.The results show that the calculated surface settlements based on the back analysis are basically consistent with the measured ones, and the calculated values by the other two methods differ greatly from the measured ones.The above analysis shows that the proposed method is reasonable and feasible, the reverse performance parameters are reasonable, and the calculated results are more accurate and in line with the reality.The proposed method can provide a new way for the acquisition and verification of the parameters of rock and soil mechanics.