Abstract
Among the various types of electric motors, Permanent Magnet Synchronous Motors (PMSM) have gained more attention in recent years in industrial, domestic and traction applications due to the recent advances in Permanent Magnet (PM) technology, control systems and electric motor design. In addition, the gradual entry of European energy efficiency regulations for new electric motor installations (European Minimum Energy Performance Standard, EU MEPS) and similar standards in other regions has helped to highlight the importance of reducing energy consumption. Due to the intrinsic characteristics of PMSM, the PMSM based electric drives can significantly improve efficiency, torque and power density compared to other solutions. However, the adoption of these machines still poses some technical challenges regarding the control aspects, which are the focus of this study. As an example, modern Electric and Hybrid Vehicles (which often adopt PMSMs) require robustness, smooth and stable operation across different conditions and minimal power consumption. Ensuring the fulfilment of such requirements by design, even at the “corners” of the operating range (e.g. at very high speed), still stimulates the research on motor control. On the other hand, the use of PMSMs in certain simple industrial applications (e.g. fans, pumps, compressors, which amount for a relevant share of total consumption of electrical energy) is still lagging, even if the potential improvement in energy consumption would have an important impact. The main reason is probably the higher cost, with respect to other solutions (e.g. vs. induction machines). However, the complexity in the tuning of control parameters (which in turn require identification of motor and load parameters) plays an important role. This thesis describes the work carried out between 2018 and 2020 at the Free University of Bolzano/Bozen and partially at the Power Electronics Machines and Drives laboratory of the University of Udine. Considering both the relevance of industrial applications and the evolving importance of PMSMs in electric vehicles (EVs), the proposals presented addresses some important technical challenges in the control of PMSM particularly regarding high-speed operation and parameter identification. The main practical aspects of typical applications and the related constraints have been considered throughout the development of the work, in order to accelerate their possible deployment in real world scenarios. Interior Permanent Magnet Synchronous Motors (IPMSM) are widely adopted in electric vehicles, where operation over a wide speed range is essential. In fact, IPMSMs allow “constant power" operation up to multiples of the base speed, thanks to Flux Weakening (F W) control. This thesis reports on the study carried out on the dynamic characteristics of typical F W operation, analyzing the voltage variations during high-speed transients and their correlation with the control parameters. The design constraints are obtained in terms of closed loop control bandwidth with respect to the voltage margin to be imposed, which guarantee stable operation of the control. This analysis is expected to have an impact on the development of automatic calibration procedures for industrial drives and a robust design for applications where high reliability and/or safety is required, for example in the automotive sector. In the high-speed 3 operation of PMSM, this thesis also analyzed a detailed aspect in the PMSM control using inverters, namely the systematic current measurement error with synchronous sampling due to the induced voltage. The current measurement error due to the induced voltage can become significant at high speed (in particular, with high ratio between electrical and switching frequency). Better analysis and understanding of this aspect allow appropriate design choices, especially for very high-speed drives. Accurate identification of electrical and mechanical parameters is important for the control of PMSM based drives, particularly for the design of controllers and for monitoring the condition of Permanent Magnets (PM). In fact, all the parameters of the machine to be controlled are not always available to the user (i.e., to the person who is commissioning the drive). In addition, the mechanical parameters are generally related to the characteristics of the load, which is often not known a priori. Therefore, in industrial applications it is preferable to have a control system that estimates these parameters at least off-line, i.e. during the commissioning phase. The new automatic ("self-commissioning") and sensorless parameter identification techniques proposed in this thesis aim to fill a gap in the previous PMSM parameter identification techniques, especially when it is intended to apply so-called "auto tuning" (i.e. the automatic setting of control parameters by the system itself). In this thesis, novel methods for identifying PM flux amplitude and moment of inertia of PMSM are proposed. The proposed methods allow the parameter identification techniques to be applied when the rotation of rotor in a continuous and uncontrolled manner is restricted due to the connected load. In fact, the position variation required in the proposed methods is limited to a fraction of a mechanical revolution. Moreover, unlike many other methods in literature, the speed regulator is not involved, so it does not need to be tuned, i.e. no initial assumption on the mechanical parameters is needed.