EPE Association: EPE-PEMC 2000 - Topic 07a: Sensing and Observing of Variables

EPE-PEMC 2000 - Topic 07a: Sensing and Observing of Variables
You are here: EPE Documents > 04 - EPE-PEMC Conference Proceedings > EPE-PEMC 2000 - Conference > EPE-PEMC 2000 - Topic 07: Sensing and Observing > EPE-PEMC 2000 - Topic 07a: Sensing and Observing of Variables

   [return to parent folder]

   A Speed Observer System of the Induction Motor with Magnetising Curve Identification
By Krzemiñski Z.*, Jaderko A. [View]
[Download]

Abstract: A new speed observer for control system of induction motor is proposed in the paper. Disturbances are defined, estimated and used in observer instead of products of rotor angular velocity and rotor flux vector components. The disturbances and rotor flux vector components are used to calculate the rotor angular velocity. Very low errors between the estimated and real rotor angular velocity in steady states and transients have been observed in the new observer for values very close or equal to zero. A nonlinearity of the main magnetic path is expressed as a function of a properly chosen factor versus the position vector length. The value of the mutual inductance has been calculated on the ground of the estimated values of the state variables in the fuzzy - logic algorithm of identification. The factors of the differential equations of the observer system have been modernized in each step of the observer on the ground of the calculated mutual inductance. Properties of the control system with new speed observer have been investigated and results of simulations are presented.

   Back-EMF Estimation Methods for Permanent Magnet Motors with Non-sinusoidal Flux Distribution
By Kabziñski J., Krawiecki M., Wasiak G. [View]
[Download]

Abstract: Permanent magnet synchronous motor has sinusoidal flux distribution (or back EMF waveform) and is able to produce a ripple-free torque. The precise and continuous rotor position is re-quired for the control algorithm. All concepts of speed/position observers are based on the as-sumption that the flux (or back EMF) is sinusoidal. How-ever, owing to manufacturing tolerances, even identical machines from the same production batch will differ in their back-EMF waveform, which does not necessarily ex-hibit a perfect sinusoid. We concentrate on modelling flux distribution deformity by introducing higher harmonics and propose several observer-based techniques for back-EMF on-line identification.

   Extended Kalman Filter Based Speed Sensorless AC Motor Control
By Fodor D., Vass J., Juhász Zs., Bíró K. A. [View]
[Download]

Abstract: The paper describes a method for speed sensorless control of an AC motor. Properly the extended Kalman filter is used for estimating the speed. The EKFT algorithm and its adaptation to the motor model are described by rigorous mathematical representation and they were realised in MATLAB Simulink. The simulation results show good dynamic behaviour suitable for DSP implementation.

   Flux Observers Based on Combined Voltage-Current Models for Control of PMSM Drives
By Andreescu G.D. [View]
[Download]

Abstract: This paper develops a generalised approach on flux observer structures in wide speed range for permanent magnet synchronous motors (PMSM) drives using combined topologies of the voltage and current models. Typical applications are in the direct torque and flux control (DTFC) and in the field oriented vector control (FOVC). A critical analysis on the flux estimators structures without corrective terms concludes to combine the advantages of the current models at low speed and the voltage models at medium-high speed. To realise a smooth and monotone transition depending on speed between these models, Luenberger observer structures having the dynamic compensator proportional integral type are used. A critical comparison on eight flux observer structures gives a top classification and useful recommendations. Simulation and experimental results using a DTFC of PMSM drive, including representative flux observers, evaluate the dynamic performances of the observers with real parameter deviations and demonstrate a good global robustness in wide speed range. An improvement will be obtained by on line parameters identifications.

   Neural Network Estimation of the Rotor Speed for Direct Field Oriented Control
By Žalman M., Jovankoviè J., Bélai I. [View]
[Download]

Abstract: In the article there are described artificial neural networks (ANN) as a rotor speed estimator in the direct field oriented control structure. The two structures of the neural network estimators are qualitative compared with the based MRAS (rotor flux based MRAS) structure of the rotor speed estimation. The first structure is based on the on-line learning principle using back propagation algorithm with gradient descent learning method. The second structure is based on the off-line learning principle using back propagation algorithm with Levenberg-Marquardt learning method. The simulation using MATLAB with Simulink realizes the application results. The structures are verified and tested in term of controlling, loading and the sensitivity on parametric variation of the stator resistance.

   On-Line Induction Motor Speed Estimation: Two Schemes Based on Voltages and Currents Measurement and Pseudoinverse Calculation
By Grzesiak L., Beliczynski B. [View]
[Download]

Abstract: A pseudoinverse based method and two schemes of an on-line induction motor speed estimation are presented and discussed. We have applied nonrecursive and recursive formulae for pseudoinverse calculation. Both implementation schemes utilise voltages and currents measurement only, both are easily implementable. Simulation results are attached.

   Robust Speed and Position Calculation from Back-EMF Estimation - Permanent Magnet Drives with Non-sinusoidal Flux Distribution
By Kabziñski J., Krawiecki M. [View]
[Download]

Abstract: We consider a permanent magnet brushless motor with an arbitrary flux distribution. The drive is equipped with an observer producing estimates of back-EMFs in stator coordinates. We propose several methods of speed and position calculations from back-EMF estimations. We consider robustness of these methods and their applicability.

   Rotor Flux Observer Based Torque Control of Induction Motor Drive for EV
By Hren A., Jezernik K., Šabanoviæ A.* [View]
[Download]

Abstract: An approach to the torque control of induction motor drive, where the position of the rotor flux is calculated by a new rotor flux observer, is presented in this paper. The proposed new rotor flux observer is constructed so that the convergence of the observation error to zero is assured. The same principle of the non-linear control based on the combination of the VSS control theory and Lyapunov stability theory is used in the rotor flux observer and in the stator current control loop. The performance of the proposed IM torque control scheme was verified through the simulations and experiments.

   Sensorless Detection of Speed of Induction Machines Using Recursive Short Time Furrier Transform
By Djuroviæ M., Obradoviæ A., Joksimoviæ G. [View]
[Download]

Abstract: It has been shown that by using the recursive function in conjunction with short time Furrier transformation (STFT) and measurements of stator current one can determine the speed of induction machines without using any sensor. The use of the recursive function enables to overcome all disadvantages which are brought in calculations using STFT. It has produced results in shorter time for the same accuracy.

   Sensorless Position Estimation in Induction Machines
By Spiteri Staines C., Cilia J., Asher G.M. [View]
[Download]

Abstract: This paper describes various methods aimed at estimating the rotor position for an AC 3-phase induction motor without a shaft sensor using specially designed asymmetric rotors. A High Frequency (HF) signal needs to be injected so as to detect the rotor position. As is common knowledge, saturation causes estimation problems when using HF injection, this effect and ways to overcome this problem are discussed. The methods considered assume a saturated induction motor having an engineered cage saliency spanning one pole pitch. The performance of these methods is investigated at all loads at low speed.

   Sliding Mode Observer Scheme for Induction Motor Control
By Rodiè M., Sabanoviæ A.*, Jezernik K. [View]
[Download]

Abstract: A novel approach to the speed sensorless torque and flux tracking sliding mode control of induction motor is presented in the paper. The main contribution of the paper is a new structure of induction machine rotor flux observer aimed for the sensorless operation. Proposed approach addresses the problem sensorless induction machine servo-drive at both, low and high speed, where rapid speed changes can occur. The idea is realized by introducing an non-linear stator frequency dependent gain in the rotor flux observer and thus enabling the operation at low and high speeds. The adaptation algorithm for stator resistance is included. Observer is robust to the parameter uncertainties, especially to rotor resistance. The performance is investigated and verified with simulations and experiments.

EPE-PEMC 2000 - Topic 07a: Sensing and Observing of Variables
You are here: EPE Documents > 04 - EPE-PEMC Conference Proceedings > EPE-PEMC 2000 - Conference > EPE-PEMC 2000 - Topic 07: Sensing and Observing > EPE-PEMC 2000 - Topic 07a: Sensing and Observing of Variables
	[return to parent folder]

	A Speed Observer System of the Induction Motor with Magnetising Curve Identification By Krzemiñski Z., Jaderko A.*	[View] [Download]
Abstract: A new speed observer for control system of induction motor is proposed in the paper. Disturbances are defined, estimated and used in observer instead of products of rotor angular velocity and rotor flux vector components. The disturbances and rotor flux vector components are used to calculate the rotor angular velocity. Very low errors between the estimated and real rotor angular velocity in steady states and transients have been observed in the new observer for values very close or equal to zero. A nonlinearity of the main magnetic path is expressed as a function of a properly chosen factor versus the position vector length. The value of the mutual inductance has been calculated on the ground of the estimated values of the state variables in the fuzzy - logic algorithm of identification. The factors of the differential equations of the observer system have been modernized in each step of the observer on the ground of the calculated mutual inductance. Properties of the control system with new speed observer have been investigated and results of simulations are presented.

	Back-EMF Estimation Methods for Permanent Magnet Motors with Non-sinusoidal Flux Distribution By Kabziñski J., Krawiecki M., Wasiak G.	[View] [Download]
Abstract: Permanent magnet synchronous motor has sinusoidal flux distribution (or back EMF waveform) and is able to produce a ripple-free torque. The precise and continuous rotor position is re-quired for the control algorithm. All concepts of speed/position observers are based on the as-sumption that the flux (or back EMF) is sinusoidal. How-ever, owing to manufacturing tolerances, even identical machines from the same production batch will differ in their back-EMF waveform, which does not necessarily ex-hibit a perfect sinusoid. We concentrate on modelling flux distribution deformity by introducing higher harmonics and propose several observer-based techniques for back-EMF on-line identification.

	Extended Kalman Filter Based Speed Sensorless AC Motor Control By Fodor D., Vass J., Juhász Zs., Bíró K. A.	[View] [Download]
Abstract: The paper describes a method for speed sensorless control of an AC motor. Properly the extended Kalman filter is used for estimating the speed. The EKFT algorithm and its adaptation to the motor model are described by rigorous mathematical representation and they were realised in MATLAB Simulink. The simulation results show good dynamic behaviour suitable for DSP implementation.

	Flux Observers Based on Combined Voltage-Current Models for Control of PMSM Drives By Andreescu G.D.	[View] [Download]
Abstract: This paper develops a generalised approach on flux observer structures in wide speed range for permanent magnet synchronous motors (PMSM) drives using combined topologies of the voltage and current models. Typical applications are in the direct torque and flux control (DTFC) and in the field oriented vector control (FOVC). A critical analysis on the flux estimators structures without corrective terms concludes to combine the advantages of the current models at low speed and the voltage models at medium-high speed. To realise a smooth and monotone transition depending on speed between these models, Luenberger observer structures having the dynamic compensator proportional integral type are used. A critical comparison on eight flux observer structures gives a top classification and useful recommendations. Simulation and experimental results using a DTFC of PMSM drive, including representative flux observers, evaluate the dynamic performances of the observers with real parameter deviations and demonstrate a good global robustness in wide speed range. An improvement will be obtained by on line parameters identifications.

	Neural Network Estimation of the Rotor Speed for Direct Field Oriented Control By Žalman M., Jovankoviè J., Bélai I.	[View] [Download]
Abstract: In the article there are described artificial neural networks (ANN) as a rotor speed estimator in the direct field oriented control structure. The two structures of the neural network estimators are qualitative compared with the based MRAS (rotor flux based MRAS) structure of the rotor speed estimation. The first structure is based on the on-line learning principle using back propagation algorithm with gradient descent learning method. The second structure is based on the off-line learning principle using back propagation algorithm with Levenberg-Marquardt learning method. The simulation using MATLAB with Simulink realizes the application results. The structures are verified and tested in term of controlling, loading and the sensitivity on parametric variation of the stator resistance.

	On-Line Induction Motor Speed Estimation: Two Schemes Based on Voltages and Currents Measurement and Pseudoinverse Calculation By Grzesiak L., Beliczynski B.	[View] [Download]
Abstract: A pseudoinverse based method and two schemes of an on-line induction motor speed estimation are presented and discussed. We have applied nonrecursive and recursive formulae for pseudoinverse calculation. Both implementation schemes utilise voltages and currents measurement only, both are easily implementable. Simulation results are attached.

	Robust Speed and Position Calculation from Back-EMF Estimation - Permanent Magnet Drives with Non-sinusoidal Flux Distribution By Kabziñski J., Krawiecki M.	[View] [Download]
Abstract: We consider a permanent magnet brushless motor with an arbitrary flux distribution. The drive is equipped with an observer producing estimates of back-EMFs in stator coordinates. We propose several methods of speed and position calculations from back-EMF estimations. We consider robustness of these methods and their applicability.

	Rotor Flux Observer Based Torque Control of Induction Motor Drive for EV By Hren A., Jezernik K., Šabanoviæ A.*	[View] [Download]
Abstract: An approach to the torque control of induction motor drive, where the position of the rotor flux is calculated by a new rotor flux observer, is presented in this paper. The proposed new rotor flux observer is constructed so that the convergence of the observation error to zero is assured. The same principle of the non-linear control based on the combination of the VSS control theory and Lyapunov stability theory is used in the rotor flux observer and in the stator current control loop. The performance of the proposed IM torque control scheme was verified through the simulations and experiments.

	Sensorless Detection of Speed of Induction Machines Using Recursive Short Time Furrier Transform By Djuroviæ M., Obradoviæ A., Joksimoviæ G.	[View] [Download]
Abstract: It has been shown that by using the recursive function in conjunction with short time Furrier transformation (STFT) and measurements of stator current one can determine the speed of induction machines without using any sensor. The use of the recursive function enables to overcome all disadvantages which are brought in calculations using STFT. It has produced results in shorter time for the same accuracy.

	Sensorless Position Estimation in Induction Machines By Spiteri Staines C., Cilia J., Asher G.M.	[View] [Download]
Abstract: This paper describes various methods aimed at estimating the rotor position for an AC 3-phase induction motor without a shaft sensor using specially designed asymmetric rotors. A High Frequency (HF) signal needs to be injected so as to detect the rotor position. As is common knowledge, saturation causes estimation problems when using HF injection, this effect and ways to overcome this problem are discussed. The methods considered assume a saturated induction motor having an engineered cage saliency spanning one pole pitch. The performance of these methods is investigated at all loads at low speed.

	Sliding Mode Observer Scheme for Induction Motor Control By Rodiè M., Sabanoviæ A., Jezernik K.*	[View] [Download]
Abstract: A novel approach to the speed sensorless torque and flux tracking sliding mode control of induction motor is presented in the paper. The main contribution of the paper is a new structure of induction machine rotor flux observer aimed for the sensorless operation. Proposed approach addresses the problem sensorless induction machine servo-drive at both, low and high speed, where rapid speed changes can occur. The idea is realized by introducing an non-linear stator frequency dependent gain in the rotor flux observer and thus enabling the operation at low and high speeds. The adaptation algorithm for stator resistance is included. Observer is robust to the parameter uncertainties, especially to rotor resistance. The performance is investigated and verified with simulations and experiments.