Research On The Early Fault Identification And Monitoring Method For Transmission Bench Test

Transmission is the key equipment of automobile transmission system.Because the quality of which determines the performance of automobile transmission.Before delivery,transmissions need ex-factory tests of property,which provides reference databases for further product optimization.In the bench tests,the tested transmission may appear fault phenomenon,then other parts of transmission may lose efficacy,cause widespread damage or even break on the test-bed without taking early actions to stop the test-bed running.Therefore,the on-line monitoring system is the significant part for the test-bed experiment.The classification of early fault can help us accurately find out the fault source,and offering important information for backing repair in factory and performance tuning.The favorable monitor method can stop the test in time when early fault happens.It is convenient for experimenter to analyze the fault that protecting the characteristic of the early faults of transmission.This thesis analyzes the characteristics about the faults of transmission.Using vibration quantity as main identification signal for the early failure of transmission on the basis of studying fault recognition methods.Then,the order analysis,one of the frequency-domain analysis,is used to extract the early fault features of the transmission.Order analysis,which is based on re-sampling methods,can extract clearly characteristics of gear faults.Since the characteristics of bearing early fault is weak,this article try to find a envelope-order analysis method which combines the envelope analysis and order analysis,to extract fault feature in the high frequency modulation signal,and on basis of that builds a kurtosis filter which can filter the jamming signal involved in high frequency resonance period.This paper studies transmission threshold setting method in on-line condition monitoring system,it put forward a interval dividing method based on the rotational speed torque and temperature.Once work into a difference condition,it generates a self-learning reference.It used the reference line calculated by order spectrum and monitoring characteristics threshold based on the beta analysis for self learning.The monitoring system achieves its automatic alarm by jointing monitoring characteristics.This article designed the on-line condition monitoring system of transmission test-bed by the research of early fault identification and monitoring.It builds this monitoring system hardware platform by using PXI acquisition device of NI Company andvibration acceleration sensor of Danish B&K Company,and set up the software system based on the c# language.Finally,the performance of the on-line condition monitoring system is verified.