Research On Prognostics And Health Management System Of Shiled Machine Based On Virtual Instrument

The shield machine plays a vital role in China's infrastructure construction and urban municipal construction.In view of the problems such as low efficiency of data management,difficulty in finding faults,failure to repair faults,and low utilization of fault data,the shield machine is in use and the faults of the shield machine are deteriorated,and the engineering is shut down for a long time.Shield machine fault prediction and health management system based on virtual instrument.This paper first determines the monitoring object of the shield machine fault prediction and health management system,and analyzes the possible failure causes of these monitoring objects and the corresponding fault diagnosis methods or trend prediction methods.Then the architecture of the system is designed.The basic state data acquisition platform of shield machine based on NI OPC and the vibration data acquisition platform based on NI DAQ are built by LabVIEW,and the local storage of these two parts of data is realized.The real-time display and alarm are realized for the basic state data collected by the NI OPC to the lower-position PLC,and the intelligent fault identification method for fuzzy diagnosis is established for this part of the basic state data,and the fuzzy diagnosis is improved by the automatic update of the fuzzy matrix.Correct rate.The fuzzy diagnosis scheme of shield main motor and the fuzzy diagnosis scheme of shield propulsion system are established.The ARIMA model was established for the collected basic state data to complete the health trend prediction.The ARIMA model was established for some state data of the Nanjing Yangtze River Tunnel during the 1-172 ring construction process and the validity of the model was verified.Standard management of vibration intensity and vibration level measurement are realized for vibration data collected by NI DAQ.And this part of the data through the LabVIEW software to establish a wavelet-neural network diagnosis program to complete intelligent fault identification.The collected vibration signal with fault information is introduced into the system,and the vibration signal is decomposed using a three-layer wavelet packet to obtain signals in eight frequency bands,and the signal energy is obtained for the eight frequency band signals respectively,and the obtained energy is obtained.Make up a feature vector.The eigenvectors of multiple faults are obtained and input into an improved BP neural network for neural network training.Finally,the neural network training results are used to identify the fault signals collected.Finally,the fault prediction and health management system is improved,including online monitoring module,fault diagnosis module,fault prediction module,various information management modules,data import and export modules,and member management modules.The management system realizes real-time monitoring of state data of the shield machine,fault analysis,equipment health trend prediction,data storage and other functions,so that the shield machine changes from passive maintenance to active maintenance,improving the utilization rate of the shield machine data and reducing the major The probability of failure,the construction cost is reduced,and the construction safety is guaranteed.