Judging And Forecasting Stability Of Surrounding Rock Of Tunnel Based On Bionic And Intelligent Method

Along with the in-depth development of our country's road traffic, water and electricity construction as well as minerals and energy sources development, especially since the development of the western region, diversified uses of tunnels are more and more long, more and more extended, more and more deep. The geologic conditions and technical problems are more and more complicated. The stability of surrounding rocks of tunnels decide the security, stability and conditions of normally use of tunnels at first hand. Therefore, the result of analysis, judge and forecast to the stability of surrounding rock directly contact success and failure of tunnel constructions. If we can judge and forecast the state of surrounding rocks accurately and in time, we may foresee the instability of tunnels as soon as possible and adopt corresponding measures to mostly avoid the losses which are brought by tunnels losing stability.For above-mentioned purpose, this thesis aim at three typical and difficult problems in forecasting and judging the stability of surrounding rock: we can't reject the false noise of monitoring displacement well which influence forecasting and judging the stability of surrounding rock, partial important preliminary displacement value is lost for the reason of monitoring points can not be installed with the pushing of tunnel face, and we don't adequately use monitoring displacement information of multi-monitor. This thesis bases on widely consulting literatures about the stability of surrounding rock, BP Network, Immune clone Algorithm and Multi-sensor Information Fusion, emphatically research about preliminarily dealing with monitoring displacement information of surrounding rock, deducing onwards and forecasting monitoring displacement value, and displacement information fusion of multi-monitor. This thesis de-noises monitoring displacement information of surrounding rock by applying the signal processing method based on singular value decomposition after analyzing the limitation of traditional de-noising method. Then it aims at the limitation of BP Network, develops it and Immune clone Algorithm, and combines them to put out Immune clone BP Network, and apply it into deducing onwards and forecasting monitoring displacement value of surrounding rock. This thesis put out the information fusion algorithm of Immune clone Algorithm and structural model of information fusion of forecasting and judging the stability of surrounding rock base on the developed Immune clone Algorithm, the theory of Information Fusion and Deformation Rate Ratio Criterion which has been put out recent years. It fuses and analyzes the displacement information of multi-monitor synthetically, and pick out the synthetical displacement information which can reflect the changing of holistic state of surrounding rock. And forecasting and judging the stability of surrounding rock by the methods. Then it certifies, analyzes and summarizes the methods and results throughout the applying of Longba tunnel of Maoergai Hydropower station.The result of instance indicates that the method to de-noise monitoring displacement of surrounding rock in the article has a good effect, the method to deduce onwards and forecast displacement of surrounding rock has High accuracy and the fusing methods about displacement information of multi-monitor have also definite fusion ability. The research achievements of this thesis enrich the analysis, judgment and prediction theory and afford new theory, new methods and effective, feasible approach for the forecasting and judging stability of surrounding rock of tunnel.