Sensorless Control of Permanent Magnet Synchronous Machines

Research output: ResearchPh.D. thesis

Abstract

Permanent magnet machines, with either surface mounted or embedded magnets on the rotor, are becoming more common due to the key advantages of higher energy conversion efficiency and higher torque density compared to the classical induction machine.
Besides energy efficiency the permanent magnet machine is also used for servo applications where higher dynamics is required, e.g. in industrial automation.
The energy efficiency is essential for battery powered electric vehicles where the electric storage capacity is limited by cost, mass and volume. The control system necessary to operate the synchronous machine requires knowledge of the rotor shaft position due to the synchronous and undamped nature of the machine. The rotor position may be measured using a mechanical sensor, but the sensor reduces reliability and adds cost to the system and for this reason sensorless control methods started to appear. Sensorless control implies control of the machine without using a direct measurement of the rotor position. Instead, more information is extracted from the existing controller feedback signals - often the machine currents - and this information is used together with accurate system knowledge to replace the mechanical sensor with an indirect measurement.
The main hardware components considered in a sensorless drive are the inverter, the PM synchronous machine and the current acquisition i.e transducers, interface circuits and sampling. These hardware parts are analysed with respect to their influence on sensorless operation of a drive system with special focus on the machine.
A detailed study of the machine using finite element methods is applied to provide an in-depth knowledge of the electromagnetic behaviour of the machine. This behaviour is described by the flux linkages which are dependent on the phase currents and rotor position.
Based on the flux linkages the differential inductances are determined and used to establish the inductance saliency in terms of ratio and orientation. The orientation and its dependence on the current and rotor position are used to analyse the behaviour and establish the suitability of the machine for sensorless control using inductance saliency tracking methods.
The same electromagnetic behaviour is used in the implementation of a dynamical simulation model of the machine useful for evaluation of sensorless control methods at the control design stage. Further, the influence of the inverter dead-time on the sensorless properties is established, including timing between the inverter output and acquisition of the machine currents. The current acquisition system can add a phase shift, offset and introduce a gain mismatch. These issues are analysed and their influence are considered in the domain of sensorless properties.
The sensorless characteristics of the whole system are included in the sensorless control design stage and used for analysing the consequences on the rotor position estimation dynamics and thereby the control of the machine torque. The modelling and sensorless control design are supported by experimental work for validation and confirmation of the sensorless drive capabilities, including hardware-near details.
For the experimental work a test bench was established consisting of a industrial servo drive with an interface to a computer, where a developed toolbox in Matlab was used for configuration and full automated drive testing.
The main scientific contributions are the system characterisation focused toward sensorless properties, in particular the machine modelling incorporating saturation and rotor harmonics of the flux linkage. Those issues are transferred into sensorless characteristics of the machine and used to determine the suitability of the drive for sensorless control. The sensorless suitability of the machine with focus on injection methods tracking the inductance saliency is described by the saliency ratio SL and angle θL. The two
saliency properties can be quite dependent on the applied current and the rotor position of the machine.
The saliency ratio is a measure on how well-defined i.e. robust the sensed saliency angle is towards influence from e.g. the inverter and current acquisition in a sensorless drive. A minimum ratio can be used to indicate a suitable operating region or used opposite to set recommendations to the maximum influence from the inverter and current acquisition system.
Close

Details

Permanent magnet machines, with either surface mounted or embedded magnets on the rotor, are becoming more common due to the key advantages of higher energy conversion efficiency and higher torque density compared to the classical induction machine.
Besides energy efficiency the permanent magnet machine is also used for servo applications where higher dynamics is required, e.g. in industrial automation.
The energy efficiency is essential for battery powered electric vehicles where the electric storage capacity is limited by cost, mass and volume. The control system necessary to operate the synchronous machine requires knowledge of the rotor shaft position due to the synchronous and undamped nature of the machine. The rotor position may be measured using a mechanical sensor, but the sensor reduces reliability and adds cost to the system and for this reason sensorless control methods started to appear. Sensorless control implies control of the machine without using a direct measurement of the rotor position. Instead, more information is extracted from the existing controller feedback signals - often the machine currents - and this information is used together with accurate system knowledge to replace the mechanical sensor with an indirect measurement.
The main hardware components considered in a sensorless drive are the inverter, the PM synchronous machine and the current acquisition i.e transducers, interface circuits and sampling. These hardware parts are analysed with respect to their influence on sensorless operation of a drive system with special focus on the machine.
A detailed study of the machine using finite element methods is applied to provide an in-depth knowledge of the electromagnetic behaviour of the machine. This behaviour is described by the flux linkages which are dependent on the phase currents and rotor position.
Based on the flux linkages the differential inductances are determined and used to establish the inductance saliency in terms of ratio and orientation. The orientation and its dependence on the current and rotor position are used to analyse the behaviour and establish the suitability of the machine for sensorless control using inductance saliency tracking methods.
The same electromagnetic behaviour is used in the implementation of a dynamical simulation model of the machine useful for evaluation of sensorless control methods at the control design stage. Further, the influence of the inverter dead-time on the sensorless properties is established, including timing between the inverter output and acquisition of the machine currents. The current acquisition system can add a phase shift, offset and introduce a gain mismatch. These issues are analysed and their influence are considered in the domain of sensorless properties.
The sensorless characteristics of the whole system are included in the sensorless control design stage and used for analysing the consequences on the rotor position estimation dynamics and thereby the control of the machine torque. The modelling and sensorless control design are supported by experimental work for validation and confirmation of the sensorless drive capabilities, including hardware-near details.
For the experimental work a test bench was established consisting of a industrial servo drive with an interface to a computer, where a developed toolbox in Matlab was used for configuration and full automated drive testing.
The main scientific contributions are the system characterisation focused toward sensorless properties, in particular the machine modelling incorporating saturation and rotor harmonics of the flux linkage. Those issues are transferred into sensorless characteristics of the machine and used to determine the suitability of the drive for sensorless control. The sensorless suitability of the machine with focus on injection methods tracking the inductance saliency is described by the saliency ratio SL and angle θL. The two
saliency properties can be quite dependent on the applied current and the rotor position of the machine.
The saliency ratio is a measure on how well-defined i.e. robust the sensed saliency angle is towards influence from e.g. the inverter and current acquisition in a sensorless drive. A minimum ratio can be used to indicate a suitable operating region or used opposite to set recommendations to the maximum influence from the inverter and current acquisition system.
Original languageEnglish
PublisherDepartment of Energy Technology, Aalborg University
Number of pages265
StatePublished - 2010
Publication categoryResearch
ID: 76468382