Reliability Analysis of High-Voltage Drive Motor Systems in Terms of the Polymorphic Bayesian Network

Abstract

The reliability of the high-voltage drive motor system for pure electric commercial vehicles is in premium demand. Conventional reliability based on fault tree analysis methods is not suitable for the quantitative assessment of polymorphic systems. As an example of a pure electric commercial vehicle, this paper combines polymorphic theory and Bayesian theory to establish a polymorphic Bayesian network model of a high-voltage drive motor system in terms of a polymorphic fault tree and to quantitatively judge the system. The polymorphic Bayesian network (BN) model can accurately depict the high-voltage drive motor system’s miscellaneous fault states and solve the top event’s probability in every state, also solving the system and drawing the consistent conclusion that the presence of abrasive particles, high-temperature gluing, moisture, and localized high temperatures are the system’s weak links by solving the critical importance, probabilistic importance, and posterior probability of the underlying event, which provides a theoretical reference for structure contrive optimization and fault diagnosis. This is extremely important in terms of improving pure electric commercial vehicles’ high-voltage drive motor systems.

Keywords

• quantitative assessment
• polymorphic fault tree
• polymorphic Bayesian network
• probability importance
• critical importance