Tunan Shen aims to increase the availability of powertrain systems for
autonomous electric vehicles by improving the diagnostic capability for
critical faults. Following the fault analysis of powertrain systems in
battery electric vehicles, the focus is on the electrical and mechanical
faults of the electric machine. A multi-level diagnostic approach is
proposed, which consists of multiple diagnostic models, such as a
physical model, a data-based anomaly detection model, and a neural
network model. To improve the overall diagnostic capability, a decision
making function is designed to derive a comprehensive decision from the
predictions of various operating points and different models.