The Research On Oil Monitoring And Fault Diagnosis Of Diesel Locomotive

With the rapid development of railway transportation industry, higher requirements are put forward for train maintenance, fault diagnosis and failure prediction. As a important force of railway transportation, diesel locomotive is playing a more and more significant role, a continuous monitoring and scientific maintenance is important and essential. Because of its simple operation, easily diagnosis, higher efficiency and accurate diagnosis result, oil monitoring technology has been widely used in the diesel locomotive fault diagnosis field. However, because every kinds of locomotive is different in structure, manufacture processes and raw material, the related parameters that is used for fault diagnosis haven't has a uniform standard, this brings great inconvenience to the oil monitoring staff and locomotive maintenance staff, meanwhile, it also affects the accuracy of fault diagnosis.Based on analyzing the advantages and disadvantages of all kinds of oil monitoring technology, this paper applies every oil monitoring technology's advantages comprehensively. In order to improve the accuracy of fault diagnosis and fault prediction, it mainly does the following research:(1) For a certain type of diesel locomotive, based on a lot of historical monitoring data, this paper calculated the linear regression equation and standard judgment equation about diesel locomotive running mileage and every related parameters by using principle of linear regression and MATLAB software, then makes it as the following monitoring data's criteria, we can finish diesel locomotive failure diagnosis by means of the abnormal mornitored data.(2) Based on statistical methods of probability, this paper finished establishing a model about oil monitoring and fault diagnosis and found the concentration limit value and gradient limit value of the main monitored elements, we can judge the monitored data by means of it. Moreover, it can analyze the abnormal data synthetically and diagnose the failure of diesel locomotive.(3) By classifying the common failures of diesel locomotive and coding for them, this paper established a fault diagnosis model with artificial neural network technology and finished the neural network building, training and simulation, accordingly accomplished the diesel locomotive fault diagnosis.(4) By comparing and arguing the above three methods, this paper analyzed their advantages and disadvantages, therefore found the best diagnosis method and regarded it as the theoretical basis of developing the oil monitoring and fault diagnosis system.(5) In order to meet the practical needs of locomotive depot, this paper developed a oil monitoring and fault diagnosis system with Visual Basic and Access database. This system facilitates the oil monitoring staffs scientific management and application for the oil monitoring data. Moreover, it can provide automatic diagnosis and overcome the traditional methods'disadvantages that failure diagnosis depends on the staff's experience.