Oil Pollution And Mechanical Wear Conditions Research Base On Gray Theory And Neural Network Theory

A large number of datas indicate that wear can not only affect the reliability ofmachinery and equipment, but also one of the important factors affecting the mechanical life.Abrasive wear due to mechanical friction generated by the device is a reflection of its internalstate of wear is extremely important information carriers, therefore, machinery and equipmentstate detection and fault diagnosis is an important means for the detection of wear particles inthe oil and the oil pollution monitored.The topics to gearbox oil pollution particles for the study, the use of oil detectiontechnology, BP neural network and gray theory, a mathematical model through a lot of data,research the correlation between oil pollution and mechanical wear between. Research by therelationship between oil pollution and mechanical wear of the device to determine areasonable time for repair and replacement of oil, as well as through relationships predictpotential problems that may arise between the device data has been reached to reducedowntime, saving repair, maintenance costs, maximize the benefits achieved.Research objectives of this project are: gearbox oil through research on the effects ofpollution mechanism gearbox wear, using a variety of analytical methods to detect oilpollution in the state and the state of wear of the gearbox, analysis gearbox oil pollutionindicators and correlation between the wear indicators, the gray theory to establish a gearboxpollution-wear the relational model, gearbox-based pollution index, abrasion index ofhistorical data and current data to predict trends gearbox wear. Finally, the measured data tovalidate the confidence level of the model.The main topic of the following major elements:(1) Gearbox failure due to wear different forms and have different wear mechanisms, theuse of technical analysis to produce oil analysis abrasive wear mechanism under differentcharacteristics.(2) Using a variety of means, such as iron oil analysis spectrum, spectrum, PQ and othertechniques to detect pollution status and state of wear of the gearbox, using BP neural networkto establish the correlation between oil pollution and mechanical wear indicators indexbetween models.(3) The correlation between oil pollution and mechanical wear indicators betweenindicators based on gray theory to establish pollution-wear the relational model.(4)According to the established model, the trend prediction gearbox wear. And based on the measured data on the model constantly corrected, improved confidence in the model inorder to evaluate, through modeling and analysis, the measured data is consistent with theforecasts indicate that oil pollution indicators can be used to predict the mechanical wear.Through the use of BP neural network and gray theory to establish the appropriatecorrelation model for oil pollution indicators and mechanical wear state, indicating a closerelationship between pollution and mechanical wear of mechanical engineering in the oil andget the oil in a use particle content, species and prediction methods to determine theprobability of mechanical failure.