Research On The Method Of Detecting Boulder Stratum With Shield Tunneling Parameters

With the popularization of shield method,the problems encountered in construction are increasing gradually.In Guangzhou,the construction ground consists of boulders.When shield excavates across the boulder stratum,boulders with different sizes and strength easily cause damage to the cutter-head and cutters.Therefore,the detection of boulder stratum has become an urgent problem.Taking Hengqin tunnel construction as background,combining with the project geological condition,formation mechanism and distribution of boulders are analyzed.Statistical analysis of data obtained in the construction of shield is made to know the distribution of the tunneling parameters in boulder stratum.Through systemically analyzing their components,total thrust and cutter torque calculation model in boulder stratum is built.Based on the special energy of shield,modulated special energy(SM)method is put forward to detect boulder stratum,and their coefficient is determined by using the comprehensive evaluation method of probability.According to the SM,the recognition matrix of boulder stratum is established.Additionally,the method is tested to more accurately and stably detect boulder stratum by using the machine's operating data.Based on BP neural network method whereby the same detection rules are used.The input is tunneling parameters,and the output is the value of SM and detection of boulder stratum.Compared with the value of recognition matrix and neural network for boulder stratum,result shows promise as the BP neural network for detecting and analyzing boulder stratum.Based on the GUI tool of MATLAB,the software is developed to detect the tunneling stratum of shield.In the process of shield tunneling,the software is able to display the change of data and the result of stratum detection in real-time.