EPE Association: EPE 1991 - 33 - Session 3.11: ESTIMATION, MEASUREMENTS AND PROTECTIONS

EPE 1991 - 33 - Session 3.11: ESTIMATION, MEASUREMENTS AND PROTECTIONS
You are here: EPE Documents > 01 - EPE & EPE ECCE Conference Proceedings > EPE 1991 - Conference > EPE 1991 - 33 - Session 3.11: ESTIMATION, MEASUREMENTS AND PROTECTIONS

   [return to parent folder]

   AN INSTANTANEOUS SPEED OBSERVER FOR HIGH PERFORMANCE CONTROL OF DC SERVOMOTOR USING DSP AND LOW PRECISION SHAFT ENCODER
By Yoichi Hori; Takaji Umeno; Toshiyuki Uchida; Yusuke Konno [View]
[Download]

Abstract: We propose the instantaneous speed observer for high performance control of servomotors which use inexpensive low precision shaft encoders. Due to the recent development of DSP (Digital Signal Processors), we have been able to control servomotors with a very short sampling period, e.g. 100[μs]. We usually obtain the speed information from the increased pulse number of a shaft encoder in a sampling period. However, in this simple method, we lose the speed resolution easily particularly in the lower speed range. If we use a longer sampling period for maintaining the speed resolution still high, the system will be easily unstable because a long deadtime is inserted in the feedback loop. To overcome these problems, we propose the novel estimation technique of the instantaneous speed by utilizing the armature current together with the encoder information. As this observer inherently includes the disturbance observer, we can use it for the effective disturbance rejection control. We implement this system by using a DSP and MOSFET chopper, and prove that it can greatly improve the system stability and achieves the excellent disturbance rejection performance through laboratory experiments.

   SPEED MEASUREMENT IN VECTOR-CONTROLLED INDUCTION MACHINE BY ADAPTIVE METHOD
By M. Boussak; G. A. Capolino; V. T. Nguyen Phuoc [View]
[Download]

Abstract: In order to remove the direct speed measurement in vector controlled induction machines drives, several techniques have been proposed during the last years. The aim of this paper is to present a new technique, based upon Model Reference Adaptive System (MRAS) method, which permits to estimate the speed by means of comparison of rotor fluxes obtained both from stator and rotor equations in static reference frame. The design has been performed using simulation tool with block-scheme representation and the tests of parameter influence have been investigated for robustness improvement. The final implementation has been checked for a large range of speed measurement from no-load to full-load on the shaft. Several remarks for numerical performance of the control microcomputer have been pointed out.

   A PROCEDURE TO ESTIMATE THE ABSOLUTE POSITION OF THE ROTOR FLUX OF A PERMANENT MAGNET SYNCHRONOUS MACHINE
By Klaus Backhaus; Harry Reinold; Thomas Kalker [View]
[Download]

Abstract: For applying field oriented control to a permanent magnet synchronous machine (PSM) it is necessary to know the rotor flux position. A method is presented, which makes it possible to estimate the location of the rotor flux of a PSM to initialize an incremental encoder without moving the rotor more than a few increments and to enable a controller to start field oriented current control after system start. The procedure is also suitable if the rotor of the machine is braked mechanically. A realization with a 1kW servo system is presented, which shows the accuracy of estimated rotor position better than 1.5% of electrical angle.

   FIELD-ORIENTED CONTROL OF SYNCHRONOUS AND ASYNCHRONOUS DRIVES WITHOUT MECHANICAL SENSORS USING A KALMAN FILTER
By G. Henneberger; B.-J. Brunsbach; Th. Klepsch [View]
[Download]

Abstract: Pole-position and rotor speed sensors are required for the field-oriented control of synchronous and induction motors. These expensive sensors can be replaced by a sensorless concept. It is based on a Kalman filter that estimates the state variables that are required for the control by only measuring the line voltages and currents of the machine. This method has been applied to a synchronous and an induction motor. Different models have been determined and practical results will be given.

   EXTENDED KALMAN OBSERVER FOR INDUCTION MACHINE ROTOR CURRENTS
By G. A. Capolino; B. Du [View]
[Download]

Abstract: In modern electrical drives such as vector controlled induction machines, it is often necessary to know exactly the machine state variables for the design of robust regulators. The Kalman filter is one of the multiple possibilities to observe non-measurable quantities such as rotor currents and its parameters. The aim of this paper is to present the design of Kalman filter (KF) and extended Kalman filter (EKF) in vector control scheme in order to estimate both rotor currents and parameters of the machine. The first part of the paper gives a short introduction about Kalman filter theory in order to provide notations for the reader and to recall the basic formulas. The second part is dedicated to development of discrete form for state equations in linear time-invariant (LTI) way for application of KF to state estimation. An extension to EKF is given adding rotor resistance as state variable in a non-linear time-invariant (NLI) way. The last part of the text is simply related to numerical implementation and tests about convergence of the algorithms. CPU evaluation in dedicated processor and some simulation results.

   ON-LINE ESTIMATION TECHNIQUE FOR ELECTRICAL AND MECHANICAL PARAMETERS OF DC AND AC MOTORS
By C. Rossi; A. Tonielli [View]
[Download]

Abstract: The design of an automatic system for parametrization of DC and AC motors directly on the production line, is considered. The specific application is constrained by timing and measurement limitations imposing indirect measurement techniques and requires careful consideration in the selection and evaluation of parameter estimation algorithms. Continuous-time and discrete-time estimation techniques are considered: the former gives better results, while the latter is easier to implement. Simulation results are reported, showing advantages and drawbacks of each method.

EPE 1991 - 33 - Session 3.11: ESTIMATION, MEASUREMENTS AND PROTECTIONS
You are here: EPE Documents > 01 - EPE & EPE ECCE Conference Proceedings > EPE 1991 - Conference > EPE 1991 - 33 - Session 3.11: ESTIMATION, MEASUREMENTS AND PROTECTIONS
	[return to parent folder]

	AN INSTANTANEOUS SPEED OBSERVER FOR HIGH PERFORMANCE CONTROL OF DC SERVOMOTOR USING DSP AND LOW PRECISION SHAFT ENCODER By Yoichi Hori; Takaji Umeno; Toshiyuki Uchida; Yusuke Konno	[View] [Download]
Abstract: We propose the instantaneous speed observer for high performance control of servomotors which use inexpensive low precision shaft encoders. Due to the recent development of DSP (Digital Signal Processors), we have been able to control servomotors with a very short sampling period, e.g. 100[μs]. We usually obtain the speed information from the increased pulse number of a shaft encoder in a sampling period. However, in this simple method, we lose the speed resolution easily particularly in the lower speed range. If we use a longer sampling period for maintaining the speed resolution still high, the system will be easily unstable because a long deadtime is inserted in the feedback loop. To overcome these problems, we propose the novel estimation technique of the instantaneous speed by utilizing the armature current together with the encoder information. As this observer inherently includes the disturbance observer, we can use it for the effective disturbance rejection control. We implement this system by using a DSP and MOSFET chopper, and prove that it can greatly improve the system stability and achieves the excellent disturbance rejection performance through laboratory experiments.

	SPEED MEASUREMENT IN VECTOR-CONTROLLED INDUCTION MACHINE BY ADAPTIVE METHOD By M. Boussak; G. A. Capolino; V. T. Nguyen Phuoc	[View] [Download]
Abstract: In order to remove the direct speed measurement in vector controlled induction machines drives, several techniques have been proposed during the last years. The aim of this paper is to present a new technique, based upon Model Reference Adaptive System (MRAS) method, which permits to estimate the speed by means of comparison of rotor fluxes obtained both from stator and rotor equations in static reference frame. The design has been performed using simulation tool with block-scheme representation and the tests of parameter influence have been investigated for robustness improvement. The final implementation has been checked for a large range of speed measurement from no-load to full-load on the shaft. Several remarks for numerical performance of the control microcomputer have been pointed out.

	A PROCEDURE TO ESTIMATE THE ABSOLUTE POSITION OF THE ROTOR FLUX OF A PERMANENT MAGNET SYNCHRONOUS MACHINE By Klaus Backhaus; Harry Reinold; Thomas Kalker	[View] [Download]
Abstract: For applying field oriented control to a permanent magnet synchronous machine (PSM) it is necessary to know the rotor flux position. A method is presented, which makes it possible to estimate the location of the rotor flux of a PSM to initialize an incremental encoder without moving the rotor more than a few increments and to enable a controller to start field oriented current control after system start. The procedure is also suitable if the rotor of the machine is braked mechanically. A realization with a 1kW servo system is presented, which shows the accuracy of estimated rotor position better than 1.5% of electrical angle.

	FIELD-ORIENTED CONTROL OF SYNCHRONOUS AND ASYNCHRONOUS DRIVES WITHOUT MECHANICAL SENSORS USING A KALMAN FILTER By G. Henneberger; B.-J. Brunsbach; Th. Klepsch	[View] [Download]
Abstract: Pole-position and rotor speed sensors are required for the field-oriented control of synchronous and induction motors. These expensive sensors can be replaced by a sensorless concept. It is based on a Kalman filter that estimates the state variables that are required for the control by only measuring the line voltages and currents of the machine. This method has been applied to a synchronous and an induction motor. Different models have been determined and practical results will be given.

	EXTENDED KALMAN OBSERVER FOR INDUCTION MACHINE ROTOR CURRENTS By G. A. Capolino; B. Du	[View] [Download]
Abstract: In modern electrical drives such as vector controlled induction machines, it is often necessary to know exactly the machine state variables for the design of robust regulators. The Kalman filter is one of the multiple possibilities to observe non-measurable quantities such as rotor currents and its parameters. The aim of this paper is to present the design of Kalman filter (KF) and extended Kalman filter (EKF) in vector control scheme in order to estimate both rotor currents and parameters of the machine. The first part of the paper gives a short introduction about Kalman filter theory in order to provide notations for the reader and to recall the basic formulas. The second part is dedicated to development of discrete form for state equations in linear time-invariant (LTI) way for application of KF to state estimation. An extension to EKF is given adding rotor resistance as state variable in a non-linear time-invariant (NLI) way. The last part of the text is simply related to numerical implementation and tests about convergence of the algorithms. CPU evaluation in dedicated processor and some simulation results.

	ON-LINE ESTIMATION TECHNIQUE FOR ELECTRICAL AND MECHANICAL PARAMETERS OF DC AND AC MOTORS By C. Rossi; A. Tonielli	[View] [Download]
Abstract: The design of an automatic system for parametrization of DC and AC motors directly on the production line, is considered. The specific application is constrained by timing and measurement limitations imposing indirect measurement techniques and requires careful consideration in the selection and evaluation of parameter estimation algorithms. Continuous-time and discrete-time estimation techniques are considered: the former gives better results, while the latter is easier to implement. Simulation results are reported, showing advantages and drawbacks of each method.