Fault Prediction Of TBM Advance

As a kind of large engineering machinery, shield machine is widely used in urban subway construction and peculiarly prone to failure, whose working conditions affected by many of the natural environment. Therefore, it’s of great significance to research on fault prediction technology of shield machine, but it’s difficult to meet the requirements by using traditional fault prediction technology. With the advent of artificial intelligence failure prediction technology and the very good prediction effect in the practical application of engineering, so it becomes practical to using intelligence failure prediction technology to the failure prediction of shield machine. Through the analysis of the expert system theory knowledge and combining the technical advantages of the fuzzy logic theory and neural network knowledge, this article mainly has carried on the preliminary discussion to the fault prediction technology of the shield machine propulsion system, and finished several aspects of the work as follows:(1)The failure knowledge base of the shield machine propulsion system is established. The mechanism of the shield machine propulsion system fault is analyzed, and this kind of failure is divided into the line fault knowledge and the deep fault knowledge. Also, the failure symptom parameters related to the shield machine propulsion system is selected, at the same time, the fault knowledge base is designed and processed effectively by introducing database technology.(2)The algorithm of failure prediction reasoning machine in shield machine propulsion system is studied. For the complexity and uncertainty of the shield machine propulsion system fault, the fuzzy logic theory and neural network knowledge are introduced into its failure prediction reasoning machine, and have carried on the design and simulation respectively. After analyzing the advantages, disadvantages and accuracy of them, the fuzzy neural network is used in the design of the fault prediction reasoning machine, and the fuzzy neural network fault prediction algorithm is simulated, also the simulation results have higher accuracy compared with the methods mentioned above, which has proved its effectiveness and accuracy in the fault prediction of shield machine propulsion system.(3)The failure prediction software of the shield machine propulsion system is designed. Combining with the software design principles, this article has chosen the Visual C++6.0as the software design platform of the expert system, completed the data exchange between VC and WinCC by using OPC technolog,and realized VC call for the neural network fault prediction algorithm and the fuzzy neural network fault prediction algorithm written in MATLAB using the COM component technology, at the same time using the interface and modular design approach in he process of software development, to make it is more convenient to expand the function module of system software and more in keeping with the the design requirements of the whole system software.