Work

Safety and Performance Enhancement Technologies for PMSM-Based Electric Powertrains

Year: 2025
Type: dissertation
Abstract: The global transition toward electrified transportation has positioned electric vehicles (EVs) as a cornerstone of sustainable mobility, yet the reliability and safety of their powertrains remain critical challenges.This thesis addresses errors within permanent magnet synchronous motor (PMSM)-based electric powertrains, which propagate through tightly coupled subsystems, degrading performance and functional safety.Unlike catastrophic faults, these errors originate from non-idealities in sensors, power electronic inverters, analog-to-digital converters, and microcontrollers during normal operation and when mission-critical components such as inverters and position sensors operate in degraded or fault-tolerant modes.In such scenarios, measurement and estimation inaccuracies become more pronounced and, in multi-motor architectures, are further amplified by interactions between drive units.Initially, a comprehensive literature review examines error sources in vital powertrain components during regular and fault-tolerant operation.Building on this foundation, three distinct error mitigation strategies are developed.For single-motor PMSM drives, an Enhanced Adaptive Selective Harmonic Annulation algorithm embedded within a Finite Control Set Model Predictive Current Control framework provides unified, real-time compensation for hybrid errors by suppressing harmonic distortions in dq-axis currents.For quad-motor drivetrains, a correlative correction strategy exploits inherent coupling through a shared DClink, enabling simultaneous estimation and compensation of offset and scaling errors using algebraic relationships between inverter switching states and current flow paths.For dual-PMSM all-wheel-drive powertrains, a robust fault-tolerant framework addresses rotor position sensor failure using a hyperbolic tangent-based sliding mode observer, while a blended weighting transition algorithm enables seamless handover from sensor-based to sensorless operation, leveraging front motor speed feedback to enhance low-speed estimation.Experimental and comprehensive simulation results establish that system-level error mitigation, rather than component-level isolation, is essential for ensuring the reliability, efficiency, and functional safety of next-generation electrified propulsion systems.The research contributes a comprehensive software-based safety framework addressing sensor imperfections from gradual degradation to complete failure, advancing the development of safer and more dependable EVs.Nomenclature 𝑎 1 , 𝑏 1 Scaling ratios in motor subsystem 1 𝑐 𝑟 Centre of the rotor 𝑐 𝑠 Centre of the stator 𝐶 𝑠/ℎ Internal Sampling capacitor 𝑐 𝜔 Centre of rotation 𝑒 𝑓𝛼,𝛽 Component of extended EMF associated with the fundamental 𝑒 ℎ𝛼,𝛽 Component of extended EMF associated with the harmonics 𝑒 𝛼,𝛽 Extended EMF in the 𝛼-axis and 𝛽-axis respectively 𝑒2 𝛼 , 𝑒2 𝛽 Back-EMF estimates 𝐹 𝑑𝑟𝑎𝑔 Aerodynamic drag force 𝑓 𝑒 Stator electrical frequency 𝐻𝑉 𝐷𝐶 High DC-bus voltage 𝑖 𝐴 Load current 𝑖 AM1_t1_1 , 𝑖 AM1_t1_2 Currents sampled by the phase-A current sensor 𝑖 𝑎_𝑚𝑒𝑎 , 𝑖 𝑏_𝑚𝑒𝑎 Measured currents in the abc reference frame 𝑖 B1_t3_1 , 𝑖 B1_t3_2 Actual phase-B currents in motor subsystem 1 𝑖 𝑑𝑐 DC-bus current 𝑖 DC1_t1_1 , 𝑖 DC1_t1_2 Currents sampled by the DC-bus sensor 𝑖 𝑑 , 𝑖 𝑞 Stator dq-axis currents 𝑖 𝑞_𝐴𝑆𝐻𝐴 , 𝑖 𝑑_𝐴𝑆𝐻𝐴 q-axis and d-axis compensation currents 𝑖 A 1 ′ , 𝑖 𝐵1 ′ Compensated phase-A and phase-B currents
Institution: University of York
Language: English
Open Access status: green
OpenAlex ID: W7214268277
Cites: 294
Cited by:
Related to: 10