Research On Oil Condition Monitoring And Malfunction Diagnosis System Of Mine Truck Engine

As the key assembly of the mining vehicle,the engine provides power for the mining vehicle.Thus,the normal operation and effective maintenance of the engine are important to the safe production of the mine.In the context of the continuous expansion of mine production capacity,the operating conditions of mining truck engines have also become more and more complicated,the tendency of deterioration is prominent,the failure rate are increasing year by year with the service life.Therefore,the accurate assessment and malfunction diagnosis of the conditions of the various parts of the mining vehicle engine,and the improvement of the maintenance quality and efficiency of the mining vehicle engine have always been the subject of painstaking research by many scholars.The mining vehicle maintenance department has introduced lubricant atomic spectroscopy technology to detect various wear elements in mining vehicle engine lubricants so as to better meet the requirements of the development of the industry and improve the specialization and intelligence of mining vehicle engine maintenance.The concentration of the elements reflects the wear status of various parts of the mining truck engine in real time,achieves that the operating status of the mining truck engine can be understood without disassembling,and improves the maintenance efficiency of the mining truck engine.In order to make more effective use of lubricant atomic spectroscopy information,this paper first utilizes wavelet transform to denoise the spectral analysis data of mining vehicle engine lubricants,and then utilize the improved three-line value method to evaluate the wear status of the mining truck engine,and finally use the nuclear extreme learning machine optimized by the grey wolf algorithm to establish the corresponding malfunction diagnosis model to achieve the sort of the mining truck engine malfunction.The main research contents of this thesis are as follows:1.Obtain the spectral analysis monitoring data of a certain mining vehicle engine lubricant from a highway transportation operation department,utilize wavelet transform to reduce the noise,and then extract the steady-state information and singular information of the spectral analysis signal and combine it with the noise standard deviation to construct a three-line value method for determining the wear limit of mining truck engines,and use three indicators based on typical failure rate curves of mechanical equipment and statistics to compare the strong points and weak points of the improved three-line value method and the traditional three-line value method in this thesis.Quantitative analysis is carried out through the improved three-line method,and the working state of the engine is judged according to the concentration of the wear element of the mining truck.Draw the normal line,warning line and dangerous line of the spectral analysis project of this model of mining vehicle engine lubricant,and then observe the newly collected oil spectral analysis data to fall on the chart in the area to infer the wearing condition of the mining truck engine.2.Study the solution algorithms of feedforward neural networks such as BP neural network and extreme learning machine,compare their strong points and weak points,a nuclear extreme learning machine with stable performance and strong generalization ability is selected and combined with the grey wolf algorithm to construct a mining truck engine malfunction diagnosis model finally.Through the simulation test of the GWO-KELM malfunction diagnosis model and the comparison of the results with the simulation test results of the BP neural network model and the extreme learning machine model,the accuracy and rationality of the GWOKELM mining vehicle engine malfunction diagnosis model established in this thesis is verified.3.In order to improve the professionalization and intelligence of wear evaluation and malfunction diagnosis of mining vehicle engine,and improve the work efficiency of field maintenance personnel,a set of mining vehicle engine oil condition monitoring and malfunction diagnosis system has been developed.The system mainly includes engine wear evaluation module,query and print service module,account management module and engine fault diagnosis model module based on GWO-KELM.