High-performance Permanent Magnet Synchronous Motor (PMSM) control systems typically incorporate a position and/or speed sensor. In these control systems, precise rotor position data is crucial for converting the stator currents into two independent components: one dedicated to torque control and the other aimed at regulating the magnetic flux. Nevertheless, the effectiveness of the control system declines markedly when inaccuracies arise in the measurement of rotor position.

Rotary position sensors encompass resolvers, incremental encoders, and absolute encoders. The precision of both resolvers and encoders is influenced by the variability inherent in the manufacturing process associated with a specific design. A key advantage of the resolver is its robustness. Nevertheless, its implementation tends to be relatively complex, and the processing of output signals to ascertain the rotor position presents additional challenges. Furthermore, a resolver-to-digital converter is necessary as an electronic interface to digitize the resolver signals for the controller. Position deviations may arise following rotor movement caused by voltage fluctuations.

This paper presents a fault diagnosis and fault-tolerant control scheme for handling position or speed sensor faults in salient pole PMSM drives using estimation techniques. The approach integrates a conventional vector controller with virtual sensors to ensure continuous operation across the full speed range. Fault detection relies on comparing measured and delayed rotor speed signals. This software-based method provides a reliable backup against sensor failures such as signal loss or drift. While sensorless control strategies cannot match the accuracy of resolver-based systems, they must maintain position estimation errors within acceptable limits to prevent excessive torque ripple and current.