Accurate flux linkage characterization is essential for the design, control, performance and efficiency optimization of permanent magnet (PM) traction motors in automotive applications. Precise knowledge of flux linkage across varying load, speed, and temperature conditions directly impacts torque production, field-weakening capability, overall drive system efficiency and torque security.
This paper presents a critical review and classification of flux-linkage characterization methods, encompassing offline laboratory mapping, standstill signal injection, self-commissioning inverter-only routines, and online real-time estimation. Each method exhibits distinct trade-offs in terms of accuracy, robustness to inverter nonlinearities, temperature adaptability, cost, and scalability for production and in-vehicle use. With the increasing complexity of automotive electrified traction systems, understanding these trade-offs is crucial for optimal motor design and control.
To enable systematic comparison, a performance evaluation framework is introduced based on key performance indicators (KPIs), including accuracy (below base speed, field weakening, and regeneration conditions), robustness to stator resistance variations, inverter nonlinearity, thermal adaptability, test complexity, time efficiency, cost, and applicability across the motor development lifecycle. These KPIs were selected based on their direct impact on motor performance, manufacturability, and suitability for various stages of automotive traction motor development, ensuring that analytical, dyno testing, and in-vehicle considerations are addressed.