Lavinius Ioan Gliga

Lavinius Ioan Gliga

Auto-generated profile, based on publicly available information

Lavinius Ioan Gliga

University ‘Politehnica’ of Bucharest

Research Focus & Metrics

Academic Career Length

12Years (2015-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
19 August 2021 in Wireless Personal Communications
Verified
Journal Article
Closed access
PDF file
LGLavinius Ioan Gliga
Lavinius Ioan Gliga
RKRedouane Khemmar
Redouane Khemmar
HCHoucine Chafouk
Houcine Chafouk
DPDumitru Popescu
Dumitru Popescu

1 October 2018
Verified
Conference Paper
Closed access
LGLavinius I. Gliga
Lavinius I. Gliga
HCHOUCINE CHAFOUK
HOUCINE CHAFOUK
DPDumitru Popescu
Dumitru Popescu
CLCiprian Lupu
Ciprian Lupu
The most common faults which can affect a Permanent Magnet Synchronous Generator are the rotor demagnetization, eccentricity (static, dynamic and mixed) and inter-turn short circuit. Their effect is noticeable on the spectrum of the stator currents, which is computed using the FFT. However, for a wind turbine, the spectrum of the currents changes with the wind speed. Therefore, the obtained results may not be accurate. In this paper, the residuals between an Extended Kalman Filter and the measured currents are proposed to be used for fault diagnosis and identification, via the FFT. The spectrum of the residuals is invariant to changes in the wind speed, but sensitive to faults.

1 May 2017
Verified
Conference Paper
Closed access
LGLavinius Ioan Gliga
Lavinius Ioan Gliga
CLCiprian Lupu
Ciprian Lupu
HCHOUCINE CHAFOUK
HOUCINE CHAFOUK
DPDumitru Popescu
Dumitru Popescu
The challenges related to the monitoring, diagnosis and control of a wind farm are presented. First, Wireless Sensors Networks are investigated, from an Internet of Things point of view: the most used and recent routing protocols, and the methods which can secure the network against attacks. As the volume of collected information is very large and diverse, big data processing frameworks needed to aggregate and interpret the information are then looked into. Computing paradigms which can ease the storing and the processing of big data, while lowering the costs for power companies are also considered. Wind farm level fault detection methods are used to look for defects, both from the model based perspective, as well as from the signal processing one. Control algorithms which are tolerant to faults are needed for this large system of systems. A quick summary of this new domain is given, even if it still is in an incipient phase. Perspectives on how all the previous components can be integrated with one another, to control a wind farm, are given.

1 October 2018
Verified
Conference Paper
Closed access
LGLavinius Ioan Gliga
Lavinius Ioan Gliga
HCHOUCINE CHAFOUK
HOUCINE CHAFOUK
DPDumitru Popescu
Dumitru Popescu
CLCiprian Lupu
Ciprian Lupu
A Permanent Magnet Synchronous Generator is chosen as a case study for this comparison. A continous model with a discrete integrator is shown to be the best solution for the discretization of the nonlinear model. Then, the Kalman Filter (KF), the Extended Kalman Filter (EKF) and the Unscented Kalman Filter are compared. The results are presented and it is shown that the EKF is the best suited for this application. Moreover, a discussion regarding the behaviour of the filters follows, where all are shown to act like proportional controllers.

1 June 2015
Verified
Conference Paper
Closed access
LGLavinius Ioan Gliga
Lavinius Ioan Gliga
CMCosmin - Constantin Mihai
Cosmin - Constantin Mihai
CLCiprian Lupu
Ciprian Lupu
DPDumitru Popescu
Dumitru Popescu
The adaptive-robust control strategy, proposed in this paper, combines the advantages of robust and adaptive control, while minimizing the computational load. Usually, a lot of mathematical operations are necessary, which limit the real time applicability of such strategies. By developing a simple, yet useful rule, the number of operations needed to determine if a new controller is needed has been greatly reduced. That is the most repetitive part of an adaptive - robust algorithm, because the rule check has to be done at every sampling moment. To test the new rule, a simulation software was developed, which proves the efficiency of the new rule. There are no disadvantages for the new adaptive - robust strategy, when compared to the usual approach.

1 June 2017
Verified
Conference Paper
Closed access
HCHOUCINE CHAFOUK
HOUCINE CHAFOUK
LGLavinius Ioan Gliga
Lavinius Ioan Gliga
A real time fire and explosion detection system is presented in this paper, using data acquired from a Wireless Sensor Network inside a room. First it is filtered to remove the noise. Then, the stochastic process is modelled in real time. The model is also used to predict the future temperature. The outputs of the model are used to detect sensor faults, this way assuring the reliability of the data. Fires are detected using a change detection method three being proposed in this paper, but just one being recommended. Finally, explosions are identified using the predicted data.