EPE Association: EPE 2014 - DS1e: High Performance Drives

EPE 2014 - DS1e: High Performance Drives
You are here: EPE Documents > 01 - EPE & EPE ECCE Conference Proceedings > EPE 2014 ECCE Europe - Conference > EPE 2014 - Topic 04: Electrical Machines and Drive Systems > EPE 2014 - DS1e: High Performance Drives

   [return to parent folder]

   A Compensation Method for a Motor Phase Current Sensor Offset Error using a Voltage-Source-Inverter Output Voltage Reference Value
By Hiroshi TAMURA, Toshiyuki AJIMA, Yasuo NOTO, Jun-ichi ITOH [View]
[Download]

Abstract: This paper presents a method for reducing a motor torque ripple resulting from a motor phase current sensor offset error for motor drive applications such as a hybrid electric vehicle. The proposed method estimates the motor phase current sensor offset error on the basis of the direct voltage value included in the voltage source inverter output voltage reference value calculated by a discrete Fourier transform. It then compensates for the motor phase current sensor offset error by subtracting the estimated motor phase current sensor offset error value from the detected motor phase current value. Its main feature is its ability to compensate for the motor phase current sensor offset error that varies due to temperature change while a voltage source inverter and a motor are driven. The effectiveness of the proposed method for reducing the motor torque ripple with reference to simulation results is discussed in detail. These simulation results reveal that the proposed method can compensate for the motor phase current sensor offset error with less than or equal to +-1[A] difference and reduce the average value of the motor torque ripple to half (1/2) of that without the proposed method in steady state in which the motor torque reference is from 10[Nm] to 80[Nm] and the motor rotation speed is 1000[rpm]. Also, in the transient conditions in which the motor rotation speed and the motor torque reference vary, the motor phase current sensor offset error can be compensated for normally, and the stable motor torque without a motor torque ripple of a fundamental frequency component can be obtained.

   A New Direct Slip-position Estimator MRAS Self-Sensing Control Method for Application on the DFIG Connected to a DC Network Trough a Diode Rectifier on the Stat
By Gil MARQUES, João SANTANA, Matteo IACCHETTI [View]
[Download]

Abstract: This paper presents a new self-sensing method for the direct estimation of the slip position angle in the DFIG connected to a dc network. This connection is made by the stator through a diode bridge and by the rotor through a single voltage source inverter. Considerable savings are obtained when this system is used in networks where several systems should work together, because only a dc-link is needed. Additionally to the slip position estimation, required to implement field orientation, it is also necessary to control the stator frequency because the system is connected to a dc link, and consequently the mains do not impose the stator frequency as in the traditional DFIG. The slip angle estimation method is based on the MRAS methodology and compares the stator current amplitude with the q-axis rotor current component computed using the estimation angle. Small estimation errors are obtained using this method.

   Computed Voltage Control of a robot arm including mechanical link and electrical actuator dynamics
By Pierre-philippe ROBET, Maxime GAUTIER [View]
[Download]

Abstract: To evaluate the efficiency of the computed voltage controller, tracking position and velocity errors are compared by simulation with the computed torque control and two conventional laws of regulation. At the end, another comparison is done with a non linear control law of regulation where the derivatives of the controlled variable q are tracked. The design of computed voltage controller, which includes the current dynamics of a synchronous driver, is derived. The control law is implemented and validated on an experimental robot joint.

   Estimator comparison for Resolver Motivated Sensorless Rotor Position Estimation of Wound Rotor Synchronous Motors
By David UZEL, Vaclav SMIDL, Zdenek PEROUTKA [View]
[Download]

Abstract: This paper compares the developed estimation approaches to sensorless control of wound rotor synchronous motor drives. The proposed rotor position estimation strategy is motivated by principle function of a resolver. The rotor circuit is fed by controlled three-phase bridge rectifier which produces an ac component of frequency of 300Hz. This is an injection signal. Its response on the stator can be evaluated by Phase lock loop (PLL) or simplified Kalman filter which estimates the rotor position. These two estimation strategies are in detail compared in this contribution. The comparison, theoretical foundations and conclusions are supported by both simulation and experimental results made on developed drive prototype.

   Global Linear Model of Sensorless Vector Control Systems for Induction Machines
By Kazuhiro OHYAMA, Ryosuke KUBOTA [View]
[Download]

Abstract: The linear models of Sensorless Vector Control Systems (SVCSs) for induction machines have common and non-common control blocks. The peculiar variables exist in the differential equations of common control blocks. This paper proposes the method eliminating peculiar differential variables from the differential equations of common control blocks. The proposed method and obtained linear models can simplify the calculation process of the linear model of other SVCSs. The linear model is defined as a Global Linear Model (GLM). The GLM can simplify the process of stability analysis to compare between the various SVCSs, since the linear models of other SVCSs can be derived easily by swapping the linear models of non-common parts.

   Identification of Magnetic Characteristics of Induction Motors Based on the Jiles-Atherton Model
By Giampaolo TORRISI, Sébastien MARIÉTHOZ, Roy SMITH [View]
[Download]

Abstract: Precise estimation of flux and accurate modeling are of paramount importance for improving control performance in induction machines. Non-linearities limit the performance of control techniques, which in general assume linearity of the components. The paper proposes a procedure to identify magnetic saturation and hysteresis losses, which uses the available current and voltage measurements from the inverter and is based on the Jiles-Atherton theory. The theoretical achievements are tested on an experimental test rig, showing practical feasibility of the method.

   Losses Estimation and Modelling in Active Magnetic Bearings
By Aleksei ROMANENKO, Alexander SMIRNOV, Rafal JASTRZEBSKI, Olli PYRHÖNEN [View]
[Download]

Abstract: Active magnetic bearings (AMBs) are becoming very popular in high speed electrical machines applicationsin the last decade. As the interest in them grows the necessity to optimise all the aspects of theiroperation arises. This paper presents the analytical way of calculating losses of active magnetic bearingsbased on reluctance network method. The method provides fast estimation reducing development timeand effort with accurate results. In the article the proposed method is compared with other popular lossesestimation methods and verified with experimental results.

   Multiple-objective optimization of a flux switching permanent magnet machine considering power converter constraints
By Nils LARSEN, Albrecht GENSIOR, Peter HEIN [View]
[Download]

Abstract: Among electrical machines with permanent magnets, the class of flux-switching permanent magnet machines has especially good properties with regard to mechanical robustness and cooling. It is a relatively new class of machines with only few industrial applications so far. The design of such machines is thus not well established and known publications on the topic focused on optimizing single objectives or on varying single parameters only. A systematic way to find a trade-off between multiple objectives is proposed while also considering constraints---in particular such constraints introduced by the power converter. To practically demonstrate the usefulness of multiple-objective optimization, a hypothetical redesign is performed with reference to a machine, which was originally designed and commissioned without explicit use of multiple-objective optimization. It is found that the original machine is optimal in a certain sense, but that the strength of multiple-objective optimization does not lie in finding one particular optimal solution but in visualizing trade-offs between multiple objectives/goals. The visualization reveals, for example, that 11 \% magnet material could have been saved at the cost of 16 \% higher copper losses at full load.

EPE 2014 - DS1e: High Performance Drives
You are here: EPE Documents > 01 - EPE & EPE ECCE Conference Proceedings > EPE 2014 ECCE Europe - Conference > EPE 2014 - Topic 04: Electrical Machines and Drive Systems > EPE 2014 - DS1e: High Performance Drives
	[return to parent folder]

	A Compensation Method for a Motor Phase Current Sensor Offset Error using a Voltage-Source-Inverter Output Voltage Reference Value By Hiroshi TAMURA, Toshiyuki AJIMA, Yasuo NOTO, Jun-ichi ITOH	[View] [Download]
Abstract: This paper presents a method for reducing a motor torque ripple resulting from a motor phase current sensor offset error for motor drive applications such as a hybrid electric vehicle. The proposed method estimates the motor phase current sensor offset error on the basis of the direct voltage value included in the voltage source inverter output voltage reference value calculated by a discrete Fourier transform. It then compensates for the motor phase current sensor offset error by subtracting the estimated motor phase current sensor offset error value from the detected motor phase current value. Its main feature is its ability to compensate for the motor phase current sensor offset error that varies due to temperature change while a voltage source inverter and a motor are driven. The effectiveness of the proposed method for reducing the motor torque ripple with reference to simulation results is discussed in detail. These simulation results reveal that the proposed method can compensate for the motor phase current sensor offset error with less than or equal to +-1[A] difference and reduce the average value of the motor torque ripple to half (1/2) of that without the proposed method in steady state in which the motor torque reference is from 10[Nm] to 80[Nm] and the motor rotation speed is 1000[rpm]. Also, in the transient conditions in which the motor rotation speed and the motor torque reference vary, the motor phase current sensor offset error can be compensated for normally, and the stable motor torque without a motor torque ripple of a fundamental frequency component can be obtained.

	A New Direct Slip-position Estimator MRAS Self-Sensing Control Method for Application on the DFIG Connected to a DC Network Trough a Diode Rectifier on the Stat By Gil MARQUES, João SANTANA, Matteo IACCHETTI	[View] [Download]
Abstract: This paper presents a new self-sensing method for the direct estimation of the slip position angle in the DFIG connected to a dc network. This connection is made by the stator through a diode bridge and by the rotor through a single voltage source inverter. Considerable savings are obtained when this system is used in networks where several systems should work together, because only a dc-link is needed. Additionally to the slip position estimation, required to implement field orientation, it is also necessary to control the stator frequency because the system is connected to a dc link, and consequently the mains do not impose the stator frequency as in the traditional DFIG. The slip angle estimation method is based on the MRAS methodology and compares the stator current amplitude with the q-axis rotor current component computed using the estimation angle. Small estimation errors are obtained using this method.

	Computed Voltage Control of a robot arm including mechanical link and electrical actuator dynamics By Pierre-philippe ROBET, Maxime GAUTIER	[View] [Download]
Abstract: To evaluate the efficiency of the computed voltage controller, tracking position and velocity errors are compared by simulation with the computed torque control and two conventional laws of regulation. At the end, another comparison is done with a non linear control law of regulation where the derivatives of the controlled variable q are tracked. The design of computed voltage controller, which includes the current dynamics of a synchronous driver, is derived. The control law is implemented and validated on an experimental robot joint.

	Estimator comparison for Resolver Motivated Sensorless Rotor Position Estimation of Wound Rotor Synchronous Motors By David UZEL, Vaclav SMIDL, Zdenek PEROUTKA	[View] [Download]
Abstract: This paper compares the developed estimation approaches to sensorless control of wound rotor synchronous motor drives. The proposed rotor position estimation strategy is motivated by principle function of a resolver. The rotor circuit is fed by controlled three-phase bridge rectifier which produces an ac component of frequency of 300Hz. This is an injection signal. Its response on the stator can be evaluated by Phase lock loop (PLL) or simplified Kalman filter which estimates the rotor position. These two estimation strategies are in detail compared in this contribution. The comparison, theoretical foundations and conclusions are supported by both simulation and experimental results made on developed drive prototype.

	Global Linear Model of Sensorless Vector Control Systems for Induction Machines By Kazuhiro OHYAMA, Ryosuke KUBOTA	[View] [Download]
Abstract: The linear models of Sensorless Vector Control Systems (SVCSs) for induction machines have common and non-common control blocks. The peculiar variables exist in the differential equations of common control blocks. This paper proposes the method eliminating peculiar differential variables from the differential equations of common control blocks. The proposed method and obtained linear models can simplify the calculation process of the linear model of other SVCSs. The linear model is defined as a Global Linear Model (GLM). The GLM can simplify the process of stability analysis to compare between the various SVCSs, since the linear models of other SVCSs can be derived easily by swapping the linear models of non-common parts.

	Identification of Magnetic Characteristics of Induction Motors Based on the Jiles-Atherton Model By Giampaolo TORRISI, Sébastien MARIÉTHOZ, Roy SMITH	[View] [Download]
Abstract: Precise estimation of flux and accurate modeling are of paramount importance for improving control performance in induction machines. Non-linearities limit the performance of control techniques, which in general assume linearity of the components. The paper proposes a procedure to identify magnetic saturation and hysteresis losses, which uses the available current and voltage measurements from the inverter and is based on the Jiles-Atherton theory. The theoretical achievements are tested on an experimental test rig, showing practical feasibility of the method.

	Losses Estimation and Modelling in Active Magnetic Bearings By Aleksei ROMANENKO, Alexander SMIRNOV, Rafal JASTRZEBSKI, Olli PYRHÖNEN	[View] [Download]
Abstract: Active magnetic bearings (AMBs) are becoming very popular in high speed electrical machines applicationsin the last decade. As the interest in them grows the necessity to optimise all the aspects of theiroperation arises. This paper presents the analytical way of calculating losses of active magnetic bearingsbased on reluctance network method. The method provides fast estimation reducing development timeand effort with accurate results. In the article the proposed method is compared with other popular lossesestimation methods and verified with experimental results.

	Multiple-objective optimization of a flux switching permanent magnet machine considering power converter constraints By Nils LARSEN, Albrecht GENSIOR, Peter HEIN	[View] [Download]
Abstract: Among electrical machines with permanent magnets, the class of flux-switching permanent magnet machines has especially good properties with regard to mechanical robustness and cooling. It is a relatively new class of machines with only few industrial applications so far. The design of such machines is thus not well established and known publications on the topic focused on optimizing single objectives or on varying single parameters only. A systematic way to find a trade-off between multiple objectives is proposed while also considering constraints---in particular such constraints introduced by the power converter. To practically demonstrate the usefulness of multiple-objective optimization, a hypothetical redesign is performed with reference to a machine, which was originally designed and commissioned without explicit use of multiple-objective optimization. It is found that the original machine is optimal in a certain sense, but that the strength of multiple-objective optimization does not lie in finding one particular optimal solution but in visualizing trade-offs between multiple objectives/goals. The visualization reveals, for example, that 11 \% magnet material could have been saved at the cost of 16 \% higher copper losses at full load.