Research On Electric Vehicle Reducer Life Evaluation Method Based On Durability Test Data

With the rapid development of China’s new energy vehicle industry in recent years,the performance and safety issues of vehicles have become more and more prominent.As one of the key core components of electric vehicles,the reducer’s mistake may affect normal driving or cause serious accidents.According to industry standards,enterprise need to do endurance testing to verify whether the life of the reducer meets the design requirements under the given conditions.The purpose of this behavior is to ensure the reliability and safety of the reducer.However,due to the long service life of reducers,batch endurance testing often takes thousands of hours,resulting in high testing costs and long lead times.In order to effectively utilize the monitoring signals to evaluate the endurance life of the reducer,this paper proposes a life evaluation method based on a joint data-driven support vector regression and nonlinear Wiener process.The methods use the test data of the first specimen to establish the model and test termination conditions.From this the life of the remaining specimens can be evaluated accurately in advance,which effectively reducing the endurance testing time.The main research work and conclusions of this paper are as follows:(1)This paper describes the current development status of electric vehicle reducers and the significance of the research.The current status of domestic and international research on the life assessment methods of reducers based on physical,data-driven and fusion models is reviewed.The existing problems of reducer life assessment are summarized,and the research ideas are given in this paper.(2)The main failure forms of the gearboxes and the causes of failure were determined using fault tree analysis.The contact stresses of the gears at all levels were calculated,and the life of the gears was quantitatively analyzed by combining the S-N curve of the material and Miner damage theory.On this basis,the sensor arrangement scheme and test condition design were carried out,and the validity of the test results was verified.The full-life vibration signals of the gearbox were collected through the bench test.(3)To address the problem that a single feature cannot effectively characterize the degradation process of the reducer,which leads to inaccurate life assessment,multiple indicators extracted by the multidimensional scale transformation algorithm are used for feature fusion.Thus,a comprehensive performance degradation index is constructed to better describe the degradation trend of the reducer.In order to obtain real-time degradation data to provide to the online remaining life prediction model,the feature indicator set with degradation indicators is used as the data set.The performance degradation modeling of the reducer is implemented by a genetic algorithm optimized support vector regression model.The validity and superiority of the proposed model were verified by experimental data.(4)The uncertainty of the degradation process caused by environmental conditions,individual differences,and other factors was considered,based on linear and nonlinear Wiener process models.The unknown parameters were calculated by using the great likelihood estimation through the a priori degradation data.Then the online remaining life prediction model was established by using real-time monitoring data to update the parameters in real time using Bayesian method.Through the goodness test and accuracy comparison analysis,it is determined that the nonlinear Wiener process model with random effects and exponential function structure has a strong long-term prediction capability and is suitable for the online remaining life prediction of electric vehicle reducers.By comparing and analyzing the influence of different truncation test times on the accuracy of life evaluation,an endurance test termination criterion is established.The validity of the proposed method is verified through experiments,which provides a method and means to reduce the endurance testing cost of electric vehicle reducers.