| As an important logistics equipment,bridge cranes are widely used in manufacturing and logistics industry,which can realize efficient production and fast logistics and provide solid support for industrial development.With the development of intelligent manufacturing and digital industry,the traditional bridge crane gradually advances from digitalization to intelligence.This paper takes the bridge crane as the research object and carries out the research of bridge crane predictive maintenance system,which refers to the realization of the whole life cycle management and optimization of the bridge crane through digital technology and simulation model,combines the real bridge crane with the virtual digital model,monitors and predicts the operating condition of the equipment by collecting and analyzing the real-time data,and at the same time,simulates and predicts the operation process and effect,so as to realize the Digital twin technology to carry out research on key technologies such as bridge crane virtual-real interaction,fault diagnosis,life prediction,etc.,to realize the visualization and analysis of the real data of bridge cranes in the case of virtual-real linkage.The main research contents of this paper are as follows:(1)Build the conceptual model framework of the bridge crane predictive maintenance system,divide it into system layer,functional layer,information layer,and physical layer,and use modularization to divide the data interaction framework into data acquisition module,data processing module,data communication module,and system testing module.Unity3 D and Visual Studio are utilized to develop and construct the basic framework of the bridge crane predictive maintenance system.(2)Realize the fault diagnosis and analysis of crane hoist motor based on one-dimensional convolutional neural network framework.Collect the crane hoist motor fault data during operation as the basic data set,get the fault prediction results through neural network training,get the basic fault classification,and embed it to be able to display the fault type and result analysis in the predictive maintenance system in real time.(3)Perform the life prediction analysis of the main girder of the bridge crane.Use sensors and Io T gate to collect data on the main girder strain and collect data set and open-source data set together to form a training data set.The life prediction of the main girder in the bridge crane predictive maintenance system is realized by using three algorithms: BP,Tent-BP-SSA,and LSTM,comparing the three algorithms to predict and evaluate the life prediction of the main girder of the bridge crane,and selecting the optimal analysis result as the result of prediction through the error analysis.(4)An experimental platform for bridge cranes is built,on which the predictive maintenance framework is constructed.Relevant experiments are conducted to realize the virtual-reality synchronization between the physical prototype and the digital twin model.The life prediction of main girder,fault diagnosis of motor and stress-strain visualization are integrated into the platform,and the predictive maintenance system is driven by remote operation,and the related error analysis of the predictive maintenance system is carried out to verify the reliability and accuracy of the system. |
