In order to effectively grasp the mechanism of internal deformation of rock strata, Accurate prediction of rock pressure law in excavation process。Distributed optical fiber is used to monitor the internal deformation of rock mass and support vector machine is used，Optical fiber frequency shift variation as the main characteristic parameter，The phase space of chaotic rock pressure data is constructed，Superparametric optimization of support vector machine regression (SVR) using genetic algorithm (GA)。Developing Model Tests of Similar Materials，Mining in simulated working face ，The concept of frequency shift variation of optical fibers is introduced ，Establishment of GA-SVR Time Series Prediction Model 。The prediction results are compared with traditional regression model (SVR) and BP neural network model (BPNN)。The results show that BPNN is prone to over-fitting, and the traditional SVR model depends on the selection of super-parameters，GA-SVR model is more scientific in super-parameter selection ，Not easy to，The prediction accuracy is higher than the two algorithms mentioned above。fit It has good generalization and provides scientific basis for the quantitative prediction of rock pressure time series。