Hoisting Load Identification,Fault Diagnosis Of Spindle System And Cyberized Equipment Management On The Multi-rope Mine Hoist

It's extremely important for secure and stable operation of the multi-rope hoist,which is the basic large-scale equipment for the coal mine manufacture enterprises.The multi-rope hoist works under the harsh circumstance of low speed,heavy and variable load.So the characteristics of the vibration signals are non-linearity,non-smoothness and strong noise.This paper based on the sensor data syncretization technique solves the key problems,such as hoisting load identification,fault diagnosis of spindle system and cyberized equipment management of the multi-rope hoist.The study results not only enrich and perfect the condition monitoring and fault diagnosis theories of the low speed and heavy load equipment,but also have important practical meanings on effective operation and scientific management of the equipment.First of all,the hoisting load identification method of the multi-rope hoist based on the vibration signals was advanced.The improved Empirical Mode Decomposition(EMD)was adopted to analyze the horizontal direction vibration signals of the bearing supported to the roller.Furthermore,the Intrinsic Mode Function(IMF),energy moment and energy mutual information as the characteristic parameters were extract effectively.As a result,the quantitative identification of the hoisting load was achieved by the Support Vector Regression(SVR)algorithm.This method is an indirect measure way.The conduct process protected the multi-rope hoist configuration and did not interrupt production process.So the method is more accuracy,safe and economic than the existent methods.On the basis of the relevance analysis between abnormal load and fault,the hoist load relevant fault diagnosis method was studied.The fault characteristics were extracted from energy field and the multi-fault classified machine composited of Support Vector Machines(SVM)were applied to fault mode identification.The effectiveness of this method was also validated by the example.Secondly,the rolling bearing fault diagnosis method based on the Local Mean Decomposition(LMD)and the improved Directed Acyclic Graph Support Vector Machine(DAG-SVM)was proposed.The rolling bearing vibration signals were de-noised and frequency divided by LMD,which in terms of those characteristics such as strong background noise,non-stationary,nonlinear,multicarrier modulation and so on.The energy characteristic,as the result of the decomposition,was effective characterization of the fault condition.And furthermore,the characteristic vector was built.Since the number of unidentified fault type of rolling bearing is large,this dissertation put forward an optimized DAG-SVM method based on the complex network theory.The DAG-SVM multi-classifier was established in use of the average similarity measure between classes.Experimetal results show that this method could be used for the typical fault diagnosis of rolling bearing.The accuracy and efficiency presented high value in fault locations and its dimensions identification.Thirdly,the fault diagnosis method of spindle system was provided based on the complex network clustering algorithm.A generalized clustering division guideline was brought forward in terms of the weak and scattered characteristics of the fault sample data of the spindle system.This method brought the complex network theory into the fault mode identification of the spindle system.The local judgement accuracy was enhanced and in the mean time the calculate complexity was reduced.The mode identification of the typical fault of the spindle system was implemented effectively.In addition,because of no preset number of categories in clustering result,this method provides the effective fault diagnosis for the system lacked of prior knowledge and fault samples.Fourthly,the cyberized hoisting equipment management system based on integrated condition monitoring and fault diagnosis was built.The management mode consisted of fieldbus,enterprise Intranet and long-distance Internet.The system hybrid structure and implementation process which satisfied to the need of the cyberized management were discussed.The overall function model of the system was built and the data information requirements and the process of the each function were analyzed under the circumstances of network convergence.In the end,the system exploitation was realised by the integrated use of the network communication technology,database technology and computer technology.Finally,based on the field operating multi-rope equipment,the multi-censors testing system which reflected the work statement and potential faults was designed.The mostly parameters of the operating equipment and the vibration signals of the spindle system were acquired.The above-mentioned theories and methods were proved by the tests in use of the digital signals dispose technique and fault mode identification method.The cyberized equipment management system developed was tested and analyzed.Thus,the engineering application of the condition monitoring and the fault diagnosis method was completed successfully.