With the rapid development of the global economy, issues such as the energy
crisis and environmental pollution have become increasingly severe. Owing to
their environmental friendliness, structural simplicity, and high energy
efficiency, electric vehicles have attracted widespread attention. Electric
drive technology serves as the most promising and versatile propulsion solution
for battery electric vehicles, hybrid electric vehicles, and fuel cell vehicles.
As an advanced mechatronic transmission system, the electric drive axle offers
high transmission efficiency, flexible packaging, and ease of digital and active
chassis control integration, and has thus been increasingly adopted in modern
vehicle architectures. The differential is a key component within the electric
drive axle, responsible for regulating the rotational speed difference between
the left and right wheels and ensuring balanced torque distribution. It plays a
decisive role in vehicle stability and traction performance. This study focuses
on the reliability testing methodology for differentials in electric drive
axles, primarily including the extraction of reliability test conditions and the
feasibility analysis of the proposed testing scheme. Specifically, based on the
parameters of a given electric vehicle, a Simulink model of the motor and
differential is established, and a complete four-wheel-drive vehicle model is
constructed. Through simulation under typical driving conditions, operational
data of the rear-drive axle differential are obtained. The collected data are
then preprocessed and subjected to dimensionality reduction using Principal
Component Analysis. The selected principal components are further analyzed using
K-means clustering to construct representative differential reliability test
conditions. The limitations of existing testing methods are analyzed based on
the simulated results and relevant literature. Finally, a reinforced fatigue
testing method for the differential is designed according to the extracted test
conditions, and the feasibility of the corresponding test bench is
evaluated.