TY - JOUR
T1 - A Novel Nonlinear Time-dependent Hazard Extended Intelligent Reliability Prediction Approach for Electric Vehicle Motor Controller
AU - Raghavendra Rao, N.S.
AU - Chitra, A.
AU - Krishnachaitanya, Daki
AU - Al-Greer, Maher
PY - 2025
Y1 - 2025
N2 - The reliability of Electric Vehicle Motor Controllers (EVMCs) is critical for the safety and performance of electric vehicles (EVs). Traditional reliability prediction models often fall short due to their reliance on linear assumptions and time-invariant factors, failing to capture the complex, nonlinear, and time-dependent nature of real-world operational conditions. The lower Concordance index (C-index) scores limit the time-series intelligent methods that address the nonlinear time-dependent characteristic. To address this gap, this paper proposes an Extended Hazard Nonlinear and Time-Dependent (EHNTD) strategy based on the Lifelines Python library, specifically targeting the reliability of EVMC. The prediction methodology adheres to Automotive Electronics Council (AEC) standards, using estimated failure data under desired operating conditions. The EHNTD reliability prediction technique demonstrates superior accuracy for EVMC, achieving a C-index of 0.95 and outperforming other intelligent methods for nonlinear hazard data. The insights and proposed EHNTD approach presented in this paper are intended to advance EVMC reliability studies and can be applied in the broader field of data-extended reliability engineering and research.
AB - The reliability of Electric Vehicle Motor Controllers (EVMCs) is critical for the safety and performance of electric vehicles (EVs). Traditional reliability prediction models often fall short due to their reliance on linear assumptions and time-invariant factors, failing to capture the complex, nonlinear, and time-dependent nature of real-world operational conditions. The lower Concordance index (C-index) scores limit the time-series intelligent methods that address the nonlinear time-dependent characteristic. To address this gap, this paper proposes an Extended Hazard Nonlinear and Time-Dependent (EHNTD) strategy based on the Lifelines Python library, specifically targeting the reliability of EVMC. The prediction methodology adheres to Automotive Electronics Council (AEC) standards, using estimated failure data under desired operating conditions. The EHNTD reliability prediction technique demonstrates superior accuracy for EVMC, achieving a C-index of 0.95 and outperforming other intelligent methods for nonlinear hazard data. The insights and proposed EHNTD approach presented in this paper are intended to advance EVMC reliability studies and can be applied in the broader field of data-extended reliability engineering and research.
M3 - Article
SN - 2169-3536
JO - IEEE Access
JF - IEEE Access
ER -