Application of Chaotic Phase Space Reconstruction into Nonlinear Time Series Prediction in Deep Rock Mass

Deep rock mass is well known as a highly complex nonlinear dynamic system, the present deformation forecasting methods still remains low precise. Thanks to chaos theory, these problems can be well solved through reconstructing chaotic phase space. In the paper, deformation behavior and its property such as insane, sensitive dependence on initial conditions were analyzed. Then a new forecasting method based on chaotic phase space reconstruction for nonlinear time series were brought out and applied in the displacement forecasting of surrounding rocks in Tongyu tunnel. Compared with traditional models, this model was proved not only accurate but of less workload, whatpsilas more, easy to handle. Through which, the vital information can be arrived to find the rational time of re-supporting surrounding rocks, and to take measures to prevent or minimize disasters incused by surrounding rock failure.