A Novel Low-Speed Rotor Position Estimation Algorithm for Permanent Magnet Synchronous Motors Based on a Simplified Fundamental PWM Excitation Method

Currently, the mainstream method for low-speed rotor position estimation in permanent magnet synchronous motors (PMSMs) is the high-frequency signal injection technique. This method superimposes high-frequency voltage or current signals onto the fundamental waveform and extracts the corresponding high-frequency signals carrying rotor position information to achieve position estimation. Although it is effective, this approach introduces additional losses and noise, which degrade low-speed control performance. On the other hand, fundamental PWM excitation (FPE) method is an attractive alternative for low and zero speed sensorless control, in that it does not need to inject high frequency signals. But it suffers from complicated current sampling scheme, which prohibits its practical application. This paper therefore proposes a low-speed rotor position estimation algorithm based on a simplified FPE method. By synchronously sampling the mean current derivative (MCD) within two adjacent PWM cycles, the rotor position can be estimated. The effectiveness of the proposed position estimation algorithm is verified through intensive simulations of an IPMSM drive typical for EVs.