Development and validation of sensorless control strategies for PMSM drives by means of HIL and Code Generation

Modern electric drives are required to achieve higher performance and improved efficiency. At the same time, costs and development time must be reduced to meet the objectives within the time constraints imposed by the market. In this context, Hardware-in-the-Loop (HIL) simulations play a fundamental role, as they allow testing both the control board and the control algorithms without relying on the power hardware. By running a real-time model of the drive, the HIL system emulates the physical plant and reproduces realistic operating conditions. Automatic code generation can further accelerate this process by enabling a rapid transition from high-level block models to executable embedded code, simplifying both validation and deployment. These tools can therefore be used to analyze and validate advanced control strategies, such as sensorless techniques, prior to their implementation on the actual drive. Sensorless control techniques are widely used in applications where cost reduction and compact design are key requirements. Removing position sensors allows for smaller and more cost-effective systems, but also increases the need for accurate estimation and robust control algorithms. In this thesis, three sensorless Field-Oriented Control (FOC) strategies for Permanent Magnet Synchronous Motors (PMSMs) were developed and compared: an Extended EMF observer, a Permanent Magnet flux observer with Adaptive Quadrature Phase-Locked Loop (AQPLL) and an Extended Kalman Filter (EKF). The algorithms were implemented in PLECS using code generation tools and validated both through HIL testing and on a real setup including a three-phase inverter and a PMSM. The results demonstrate that combining HIL simulation and automatic code generation significantly reduces development time while ensuring reliable assessment of the performance and stability of advanced sensorless control strategies.