Abstract:Taking the multi-point displacement data of the main powerhouse surrounding rock as example,through multi-point measuring displacement data from different sections of surrounding rocks in underground powerhouse of Xiaowan Hydropower Station,the chaos dynamic features of displacement data of underground powerhouse are studied by applying the power spectrum analysis method,and G-P method is improved to gain correlation dimension D;then the maximum Lyapunov exponent( ) is calculated by reconstruction phase space with adaptable parameters m,t that are obtained through synthetic analyzing G-P method,C-C method and multiple autocorrelation method. Results show that all the measuring multi-point displacement data in roof and of depth 30 m in abutment don¢t have chaos dynamic feature,but data from depth 2,7,15 m in abutment,the wall and those pre-measuring multi-point displacement data gained from constructing branch and drainage caves have chaos dynamic feature. The power spectrum graph has been gained to reflect chaos dynamic feature of displacement data. Correlation dimension D and the maximum Lyapunov exponents are greater than zero. For the displacement data,which have chaos characteristics,many methods can be used during excavation on the basis of the chaotic time series,which can provide a new practical method to forecast the surrounding rock displacement of underground powerhouse system.