Normal-Operation-Undisturbed Magnet Flux Linkage Monitoring in PMSM Drives via a Mechanical-Model-Based Dual Time-Scale Approach

Accurate identification of magnet flux linkage is of great importance for the condition monitoring and control optimization of permanent magnet synchronous motor (PMSM) drives. In this work, a normal-operation-undisturbed magnet flux linkage monitoring technique using a mechanical-model-based dual time-scale approach is presented. The proposed technique is composed of two Adaline-type asymptotic observers, which operate at different time scales and provide updates to each other to guarantee accuracy and cope with rank deficiency during the identification. With the aid of Lyapunov theory, a dynamic learning factor is ingeniously designed for each of the two Adaline-type asymptotic observers, which yields powerful noise immunity and helps to guarantee observer stability. In comparison to existing magnet flux linkage identification solutions, the proposed method is impervious to inverter nonlinearity and magnetic saturation, and meanwhile, it circumvents the need for harmonic signal injection and control structure alteration so as to eliminate the resulting possibility of affecting normal motor operations. Simulations, along with real-time experiments under different temperatures, loads, and speeds, are presented to demonstrate the feasibility of the proposed technique and its capability to accurately monitor the magnet flux linkage.