Risk Management Of The Whole Process Of Highway Tunnelling Use NATM

Tunnel is an important part of highway network, It risks must be controlled during the highway tunnel construction. It’s contents high technical and money costs, combine with high security risks, determined that the tunnel projects, which risk factors must be analyzed in detail during each stages and each process. Through timely and effective economic or technical means, not only make sure the investment effectively, but also make sure the quality and safety during tunnelling. Through continuous development and improvement, now, NATM has become one of the main method of the tunnelling in domestic. Although there are a lot of professors and scholars focused on the NATM risk management studies from different angle of view, but lots of them are generally concentrated on the risk evaluation about the entire length of the tunnel, ignor the truth that risk factors are nonlinear changed because of the complicated geological structure. this paper take the ZY tunnel in Fujian as example, evaluated the whole length and whole process of the tunnel’s risk in a "Sub to Total" way based on the AHP-Fuzzy-BP neural network evaluation model. The main research contents are as follows:(1) Learning from the research literatures and practical tunneling experience of NATM. I finally use Causal Analysis Diagram(fishbone diagram) to identify the risk factors during the process of tunnel construction.(2) Establish the risk evaluation index system by summarizing and classificating of the fishbone diagrams, and using analytic hierarchy process(AHP) to calculate the weights of each index.(3) Using the subjective ideal point method from the fuzzy theory, establish an reference of the index value. Through linear interpolation method to get sufficient data as the BP neural network input training vectors, and use the corresponding vectors calculated from the AHP-Fuzzy comprehension system as the output training data. Training and checking the BP neural network so that we just established an AHP-Fuzzy-BP neural network model for NATM risk evaluation.(4) Taking Fujian ZY Tunnel as example, divide the tunnel into several "cell blocks" according to different geological formations or other subjective and objective conditions, use the well-established AHP-Fuzzy-BP evaluation model, input the indicators vector of corresponding cell block as BP neural network test vectors. So,we can get the risk assessment results of each "cell block". At last, prove the validity and applicability of this risk assessment model by contrast with the tunnel risk analysis.(5) Analyze the results of the evaluation, list the response measures and control methods of the whole NATM process, provides some guidance and reference for future NATM tunneling risk management.