EPE Association: NORpie 2000 - Topic 09: VECTOR CONTROL OF MACHINES

NORpie 2000 - Topic 09: VECTOR CONTROL OF MACHINES
You are here: EPE Documents > 05 - EPE Supported Conference Proceedings > NORpie - Proceedings > NORpie 2000 > NORpie 2000 - Topic 09: VECTOR CONTROL OF MACHINES

   [return to parent folder]

   Kalman Filter - Based Control of Interior Permanent Magnet Synchronous Machine
By V. Comnac; M. Cernat; F. Moldoveanu; I. Draghici; R.M. Cernat [View]
[Download]

Abstract: The application of vector control techniques in a.c. drives demands accurate position and speed feedback information for the current control and servo-control loops. The paper describes al sensorless speed digital control system for interior permanent magnet synchronous machines (IPMSM). A Kalman filter is used for the mechanical state estimation of the motor. The observer was developed based on non-linear model of the synchronous motor that employs a q d − rotating reference frame attached to the rotor. The control system includes a rotor frame vector current controller with feed-forward decoupling circuit and an integral+integral proportional (I + PI) speed controller.

   Generalized Indirect Field Orientation with Analysis of Computational Requirements
By Lennart Harnefors; Magnus Jansson; Johnny Holmberg [View]
[Download]

Abstract: In this paper it is shown that any ux estimator|not just the \current model"|can be implemented using indirect eld orientation (IFO). We use the \voltage model" as working example. A generalized slip rela- tion, replacing the standard slip relation, results. Im- plementation of the \voltage model" using indirect vs. direct eld orientation (DFO) is compared with respect to computational complexity. While DFO re- quires about 40% less computations, it is shown that the computational overhead is, when using a mod- ern xed-point digital signal processor, insignicant in both cases. The benets of implementing the \voltage model" in IFO are twofold, and both con- cern sensorless control. First, the \voltage model" can be successfully used also at low speeds, avoid- ing transition to the \current model," which would require speed measurement. Second, a very simple method for speed estimation, which does not require a dynamic estimator such as an observer, can be employed.

   Backstepping Strategy for Induction Motor Control
By P. Vadstrup; H. Börsting; H. Rasmussen [View]
[Download]

Abstract: Using backstepping, which is a recursive non-linear design method, a novel approach to control of induction motors is developed. The resulting scheme leads to a nonlinear controller for the torque and the amplitude of the field. A combination of nonlinear damping and observer backstepping with a simple flux observer is used in the design. Assuming known motor parameters the design achieves stability with guaranteed region of attraction. It is also shown how a conventional field oriented controller may be obtained by omitting parts of the nonlinear controller.

NORpie 2000 - Topic 09: VECTOR CONTROL OF MACHINES
You are here: EPE Documents > 05 - EPE Supported Conference Proceedings > NORpie - Proceedings > NORpie 2000 > NORpie 2000 - Topic 09: VECTOR CONTROL OF MACHINES
	[return to parent folder]

	Kalman Filter - Based Control of Interior Permanent Magnet Synchronous Machine By V. Comnac; M. Cernat; F. Moldoveanu; I. Draghici; R.M. Cernat	[View] [Download]
Abstract: The application of vector control techniques in a.c. drives demands accurate position and speed feedback information for the current control and servo-control loops. The paper describes al sensorless speed digital control system for interior permanent magnet synchronous machines (IPMSM). A Kalman filter is used for the mechanical state estimation of the motor. The observer was developed based on non-linear model of the synchronous motor that employs a q d − rotating reference frame attached to the rotor. The control system includes a rotor frame vector current controller with feed-forward decoupling circuit and an integral+integral proportional (I + PI) speed controller.

	Generalized Indirect Field Orientation with Analysis of Computational Requirements By Lennart Harnefors; Magnus Jansson; Johnny Holmberg	[View] [Download]
Abstract: In this paper it is shown that any ux estimator\|not just the \current model"\|can be implemented using indirect eld orientation (IFO). We use the \voltage model" as working example. A generalized slip rela- tion, replacing the standard slip relation, results. Im- plementation of the \voltage model" using indirect vs. direct eld orientation (DFO) is compared with respect to computational complexity. While DFO re- quires about 40% less computations, it is shown that the computational overhead is, when using a mod- ern xed-point digital signal processor, insignicant in both cases. The benets of implementing the \voltage model" in IFO are twofold, and both con- cern sensorless control. First, the \voltage model" can be successfully used also at low speeds, avoid- ing transition to the \current model," which would require speed measurement. Second, a very simple method for speed estimation, which does not require a dynamic estimator such as an observer, can be employed.

	Backstepping Strategy for Induction Motor Control By P. Vadstrup; H. Börsting; H. Rasmussen	[View] [Download]
Abstract: Using backstepping, which is a recursive non-linear design method, a novel approach to control of induction motors is developed. The resulting scheme leads to a nonlinear controller for the torque and the amplitude of the field. A combination of nonlinear damping and observer backstepping with a simple flux observer is used in the design. Assuming known motor parameters the design achieves stability with guaranteed region of attraction. It is also shown how a conventional field oriented controller may be obtained by omitting parts of the nonlinear controller.