Marton Kuslits

57192821267

Publications - 5

Model-based control algorithm development of induction machines by using a well-defined model architecture and rapid control prototyping

Krisztian Horvath Marton Kuslits Szilárd Lovas

Publication Name: Electrical Engineering

Publication Date: 2020-09-01

Volume: 102

Issue: 3

Page Range: 1103-1116

Description:

This paper presents a new control algorithm development approach for induction machines by using model-based design and a systematically built model architecture implemented in MATLAB/Simulink. The model architecture follows a three-layer structure, and it is developed according to the principle of functional decomposition and the needs of reusability and expandability. The first model layer consists of elementary model and algorithm components, the second contains a machine simulation model and a field-oriented control (FOC) algorithm, built upon the first layer’s components, and the third realises the executable models by connecting the models and algorithms defined in the second layer. Furthermore, rapid control prototyping (RCP) is discussed as an experimental validation method, and an experimental setup with RCP is also introduced. The application of the presented methods is demonstrated by simulations as well as by experiments, and by using a control algorithm based on FOC as an example.

Open Access: Yes

DOI: 10.1007/s00202-020-00935-6

Parameter Sensitivity Analysis Method for Speed Sensorless Induction Machine Drives Based on Unscented Kalman Filter

Krisztian Horvath Marton Kuslits

Publication Name: Proceedings 2018 IEEE 18th International Conference on Power Electronics and Motion Control Pemc 2018

Publication Date: 2018-11-03

Volume: Unknown

Issue: Unknown

Page Range: 744-749

Description:

This paper presents a new method for determining the steady-state machine parameter sensitivities of induction machine speed sensorless control algorithms, which are utilizing unscented Kalman filter (UKF) and direct rotor field oriented control (DRFOC). Sensitivity influence of the UKF is investigated by alternative noise parameters and pointed out that selection of noise parameters may influence the flux and torque sensitivities of the entire system with respect to stator resistance variations. In order to obtain a better picture, the sensitivities are calculated for various operating points resulting in sensitivity maps. For this, a computationally efficient calculation method is applied by using the steady-state system equations.

Open Access: Yes

DOI: 10.1109/EPEPEMC.2018.8521843

Optimization-based parameter tuning of unscented Kalman filter for speed sensorless state estimation of induction machines

Krisztian Horvath Marton Kuslits

Publication Name: Proceedings 2017 5th International Symposium on Electrical and Electronics Engineering Iseee 2017

Publication Date: 2017-12-07

Volume: 2017-December

Issue: Unknown

Page Range: 1-7

Description:

State estimation of induction machines may be a difficult problem, due to the non-linear behavior of theirs. For non-linear state estimation, the unscented Kalman filter (UKF) is a well-known extension of the linear Kalman filter. Operation of the UKF algorithm strongly depends on the process and measurement noise covariance parameters of the estimator. Determination of these parameters is not straightforward and can be difficult, especially if the number of state variables and hence the system complexity is relatively high. In this paper, the UKF algorithm is applied for speed sensorless state estimation of induction machines in such a way that seven state variables are estimated from the measured stator currents and from the known excitation voltages. In order to tune the noise parameters of the UKF, a new, optimization-based method is presented. This tuning method provides adequate behavior for the observer beside difficult operating conditions as it has been shown by simulation experiment.

Open Access: Yes

DOI: 10.1109/ISEEE.2017.8170649

Speed sensorless field oriented control of induction machines using unscented kalman filter

Krisztian Horvath Marton Kuslits

Publication Name: Proceedings 2017 International Conference on Optimization of Electrical and Electronic Equipment Optim 2017 and 2017 Intl Aegean Conference on Electrical Machines and Power Electronics Acemp 2017

Publication Date: 2017-07-11

Volume: Unknown

Issue: Unknown

Page Range: 523-528

Description:

In this paper, an observer-based speed sensorless field oriented control (FOC) algorithm is presented for induction machines. The state observer is based on a new, detailed observer model which describes the machine with seven state variables, i.e. with seven equations. Since these equations are strongly non-linear, the applied observer algorithm is the unscented Kalman filter (UKF). Using the advantages of the detailed non-linear model and the UKF algorithm, the state variables can be estimated adequately, including the rotor flux position. Using these variables a speed sensorless FOC structure has been developed.

Open Access: Yes

DOI: 10.1109/OPTIM.2017.7975021

Model-Based development of induction motor control algorithms with modular architecture

Krisztian Horvath Marton Kuslits

Publication Name: Proceedings 2016 IEEE International Power Electronics and Motion Control Conference Pemc 2016

Publication Date: 2016-11-21

Volume: Unknown

Issue: Unknown

Page Range: 133-138

Description:

Development of control algorithms for electrical machines may be a difficult procedure if functional safety, software quality, reusability and expandability come into scope. These properties might be required in both of research and industrial development projects. State of the art methods and tools like model-based design (MBD) and automated code generation may help to meet these requirements. In this paper, MBD methods and a modular, reusable model architecture are presented for implementation of field oriented control (FOC)-based controller software for induction motors.

Open Access: Yes

DOI: 10.1109/EPEPEMC.2016.7751987