Investigation on Durability Conditions of Electric Drive System Based on Real-Vehicle Operational Data

The research on the characteristics of the durability conditions in electric drive system (EDS) based on real vehicle operational data is a prerequisite and foundation for the development and evaluation of load spectrum. In this paper, an investigation on the development of integrated EDS durability test conditions has been conducted by using operational data, collected from 44 vehicles with a total mileage of 103,530 km. Firstly, the real-vehicle operational data has been pre-processed and correlated with the principal component analysis (PCA) method, in which the multidimensional vehicle operating parameters can be dimensionality reduced. Moreover, K-means clustering analysis method then used to divide the operating fragment into identifying six typical operational conditions. Then, failure mechanisms and life calculation models for the EDS core components, including shafts, gears, and bearings, are established. Based on the damage analysis under different operating conditions, accelerated test segments are further extracted. Finally, a damage objective for the entire EDS life-cycle has been constructed, and a coordinated equivalent method for multi-component damage has employed to generate load spectrum and durability test conditions with synchronous acceleration under multiple stress levels.