基于用户数据的云端用户失效风险等级辨识

Identification of Cloud User Failure Risk Level Based on User Data

摘要: 相较于传统汽车，电动汽车的驱动电机展现出更宽广的调速范围、更大的启动扭矩、更高的功率密度及效率。然而，随着转速、扭矩等载荷强度的提升，电驱动系统面临更高的失效风险。随着新能源汽车的不断发展，市面上有大量用户，开展不同用户等级同一部件损伤风险评估，开展同一用户等级不同部件损伤风险评估，揭示用户等级与电驱损伤关联规律。结合聚类后用户风险等级与部件损伤关联结果，设计以用户数据核心部件损伤为输入、用户风险等级为输出的机器学习模型，实现云端新增用户的失效风险等级快速辨识。

Abstract: Compared with traditional vehicles, the drive motor of electric vehicles shows a wider speed range, larger starting torque, higher power density and higher efficiency. However, with the increase of load strength such as speed and torque, the electric drive system faces a higher risk of failure. With the continuous development of new energy vehicles, there are a large number of users on the market to carry out damage risk assessment of the same component at different user levels, and carry out damage risk assessment of different components at the same user level, so as to reveal the correlation law between user level and electric drive damage. Combined with the results of the association between the user risk level and the component damage after clustering, a machine learning model with the damage of the core component of the user data as the input and the user risk level as the output is designed to realize the rapid identification of the failure risk level of the new users in the cloud.