Sensorless control system by doubly fed machine on based the Kalman filter
Year: 2024
Type: article
Abstract: Modern sensorless asynchronous electric drive systems for stator control widely use Kalman filters as observers of coordinates, parameters and disturbing influences. For doubly fed machines where control actions are formed in the rotor circuit when the stator is directly connected to the power supply network, the problem of vector control is complicated by the need to identify not only the reference vector of the flux linkage and the angular speed of the rotor of the asynchronous machine but also the angle rotation of the rotor relative to the stator. If we integrate the speed estimation signal from the observer's output, then the regulation process will be self-oscillating. Therefore, in the paper, the rotor rotation angle equation is included in the Kalman observer model. In this case only the control voltages of the rotor circuit are accepted as input influences of the observer, and the projections of the unregulated stator voltage vector are included in the state matrix of the Kalman observer as quantities that functionally depend on the angle of rotation of the rotor. This approach made it possible to synthesize a Kalman observer in which there are no self-oscillations in the process of speed control and the control system maintains asymptotic stability at various levels of speed stabilization. The Kalman filter implemented as a computer program on a microcontroller has insignificant inertia, does not introduce a phase shift into feedback signals, and with the correct choice of its parameters and structure has virtually no effect on the dynamics of the control system. The control system provides a good decomposition of the asynchronous machine into two control channels, and the Kalman filter, when implementing direct relay control in the active and reactive power channels, ensures high accuracy and stability of the process of identifying the coordinates of the electric drive, regardless of whether, there are random noises in the measured currents and voltages.
Institution: Dniprovsk State Technical University
Language: English
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W4408527436