Gabriel Murariu

Nonlinear system identification based on internal recurrent neural networks

A novel approach for nonlinear complex system identification based on internal recurrent neural networks (IRNN) is proposed in this paper. The computational complexity of neural identification can be greatly reduced if the whole system is decomposed into several subsystems. This approach employs internal state estimation when no measurements coming from the sensors are available for the system states. A modified backpropagation algorithm is introduced in order to train the IRNN for nonlinear system identification. The performance of the proposed design approach is proven on a car simulator case study.

POLYNOMIAL FITTING APPROACH FOR THE KRAMERS KRONIG METHOD FOR OPTICAL PROPERTIES COMPUTING FOR Bi2O3 FABRIC

The optical reflectance of Bi2O3 was measured, and the optical properties were estimated by Kramers–Kronig analysis. The novelty of the present study is due to the implementation of a MAPLE software approach to the complex computations implied by this extrapolation. The analytical fit of the reflectance spectrum is applied, accompanied by a careful extrapolation, which is necessary within the Kramers–Kronig method. In this way starting from the reflectance samples data, using this transformation, a very good agreement is obtained between the main optical parameters, namely the refractive index and the absorbtion coefficient. The study is implemented for Bi2O3 films deposited by thermal vacuum evaporation at different temperatures of the glass substrates and the comparison with the experimental data set being made using the transmission and the reflection optical spectra.

Using renewable energy sources for electric vehicles charging

This paper presents several possibilities for electric power supply from renewable sources of electric vehicle charging stations. The supply system sizing and the economic analysis were performed using the specialized computing facility Homer energy and the local resources data needed to optimize the system were taken from the RETScreen databases. In this way we could choose the photovoltaic panels depending on the solar radiation and the wind turbines according to the wind class of the region. The location chosen for the charging stations is Huşi city, Vaslui county, at an average altitude of 82 m above sea level, accessible from the European road E581. The results obtained with the three systems analyzed by Homer simulation showed that the choice of an appropriate power charging station should take into account both the local resources and the policies to use and encourage renewable energy in order to establish the technically and financially feasibility of the system.

Risk for metabolic syndrome in a group of overweight children from South-East Romania

Abstract The present study aims to identify and analyse the cardiometabolic risk factors associated with the metabolic syndrome in overweight children and adolescents. The study group included 163 overweight children and adolescents, average age: 13.02 ± 3.42 years. The following evaluations were performed: anthropometrical measurements, blood pressure measurements, biochemical tests investigating the lipid and carbohydrate metabolism. Metabolic syndrome was identified in 48 subjects (29.4%). The risk to develop MS was found to be higher in males and within the 13-18 age group. The most common cardiometabolic risk factors were abdominal obesity (75.5%) and high blood pressure (41.1%), followed by low HDL-cholesterol (35%), increased fasting blood glucose (23.3%) and hypertriglyceridemia (17.8%.). The variables under analysis exhibited significant correlations with the number of metabolic syndrome diagnosis criteria. The metabolic syndrome prevalence in the paediatric population affected by excess body weight has reached high values in our geographical area. It is thus justified to initiate screening activities for the early detection and adequate treatment of the modifiable cardiometabolic risk factors, contributing to the prevention of long-term complications. Rezumat Studiul de faţa şi-a propus identificarea şi analizarea componen telor cu risc cardiometabolic asociate cu sindromul metabolic la copiii şi adolescenţii cu exces ponderal. Lotul de studiu a inclus un număr de 163 de copii şi adolescenţi cu exces ponderii, cu media de vârstă 13.02 ±3.42 ani. S-au efectuat: măsurători antropometrice, măsurarea tensiunii arteriale şi determinări biochimice pentru investigarea metabolismului lipidic şi glucidic. Prezenţa sindromului metabolic a fost înregistrată la 48 de subiecţi (29.4%). Riscul de a dezvolta SM a fost mai crescut la copiii de sex masculin şi la cei din grupa de vârstă 13-18 ani. Cele mai frecvente componente cu risc cardiometabolic au fost obezitatea abdominală (75.5%) şi hipertensiunea arterială (41.1%), urmate de HDL-colesterol scăzut 35%, glicemic ă jeun crescută (23.3%) şi hipertrigliceridemie (17.8%). Variabilele analizate au prezentat corelaţii semnificative cu numărul de criterii de diagnostic ale sindromului metabolic. Prevalenţa sin- dromului metabolic în populaţia pediatrică cu exces ponderal este crescută în zona noastră geografică. Se justifică ast- fel iniţierea activităţilor de screening pentru depistarea precoce şi intervenţia adecvată asupra factorilor de risc carclio- metabolici modificabili, contribuind în acest mod la prevenirea complicaţiilor acestora pe termen lung

Study on the influence of atmospheric parameters on the accuracy of the geodetic measurements

In the cadastral operations, geodesic measurements are extremely necessary. These measurements must be accurate and precise. In the present work, we studied the measurement accuracy of electronic measuring devices and tried to emphasize the influence of atmospheric parameters on measurement accuracy.

TRACE ELEMENTS IN FISH TISSUE WITH COMMERCIAL VALUE OF THE DANUBE DELTA BIOSPHERE RESERVE

Researches and analysis of heavy metals concentrations in the fish species exploited as a food source are of great interest due to long term toxic effects on human health even at very low concentrations. Therefore, we have conducted this study in an area of great interest in terms of aquatic biodiversity, but also of tourism and trade. Depending on the degree of toxicity in delta aquatic ichtyofauna, five heavy metals (Cr, Cu, Hg, Ni, Pb) and a metalloid (As) concentrations were measured in the muscle tissue of fifteen commercial value fish species in order to assess their impact on human health. Samples were collected seasonally (twice a year) between 2013-2014 from Somova-Parches aquatic complex (situated near Tulcea - Romania industrial area in predeltaic area) and four representative aquatic complexes for Danube Delta, respectively Sontea-Fortuna, Matita-Merhei, Gorgova-Uzlina and Rosu-Puiu. The determination of elements from collected samples was done differently, mercury was analysed by cold vapor atomic absorption, while the other five elements (arsenic, chromium, copper, nickel and lead) were determined using mass spectrometry with inductively coupled plasma. The investigations of heavy metals revealed interspecific differences, important not only because of each specie particularities and food chain, but also for intraspecific variation caused by geographical location. The statistical analysis of the 4800 measurements in correlation with the analysis of the bioaccumulation factor (BCF) reflected the degree of contamination on the exploited Danube Delta Biosphere Reserve fish resources and also showed an irregular distribution of heavy metals and metalloids depending on anthropogenic pollutants.

Increase of Regression Speed of Combustion through Oxidizer Doping

This paper presents an innovative method to increase the regression speed of a solid fuel grain by using oxidizer doping. The results obtained on a propulsion application showing the performance parameters as an output of a numerical model are provided. Two models are given that predict the regression speed increase for embedding wires and for oxidizer doping with a factor of ∼2 to 3. Also, it is shown that this increase of the regression speed by using a low oxidizer doping percentage, which is based on the assumption that at a higher doping rate the combustion instabilities, can lead to detonations. Next, a thermochemical model for the combustion within the oxidizer doping model is afforded. For this model, the burn is assumed to take place around the surface of each of the oxidizer particles, being composed of two different burning processes: one is the burning on the surface of the solid fuel grain and the other is the burning between the solid fuel and the solid oxidizer. For a non-doped fuel grain, the burning takes place only on the surface of the fuel grain. Finally, a potential application of such propulsion units to small satellite launchers is presented, and a case study of such a vehicle intended to place a 50 kg payload on a low earth orbit (300 to 500 km altitude) is described. The modelling proposed is based on the assumption that the use of solid methane makes this launcher environmentally friendly.

Sliding mode control for a dynamic system, helicopter type

This paper presents a sliding mode control structure designed for a dynamic system, helicopter type. The approach is robust to model uncertainties and disturbances. The proposed control algorithm is validated on a laboratory stand which uses carrying force to stabilize a helicopter.

Statistical and dynamical models on the Prut River state parameters. Monitoring area - Galati, Romania case study

In this paper a complementary approach of mathematical analysis is proposed. In this spirit, a systematic analysis of the recorded data on the Prut River over a period of about 8 years was made. The paper begins with a brief description of the natural habitat and positions that have made experimental determinations. The next sections present the obtained results, especially basing on the statistical analysis of these measurements. The statistical analysis was performed in two stages: a global analysis and a seasonal analysis. In the last section, the results basing on numerical analysis applied on the dynamics of the principal state parameters are presented. Future research will be conducted for a deeper analysis of aquatic systems of this type.

Temperature dependence of the optical properties of Bi2O3. A theoretical approach basing on the Kramers-Kronig transformation for polynomial mixed terms models

The Kramers-Kronig transforms (KK) constitute a powerful tool to validate experimental data. The present study is implemented for Bi2O3 thin films deposited by thermal vacuum evaporation at different temperatures of the glass substrates. Since the extraordinary properties of this fabric allow us to consider particular analytical approach as it was previously shown, the reflectance properties of Bi2O3 as a function of temperature could be studied.The novelty of this article is the studying of a global effective analytical representation, based on polynomial functions, in order to obtain a general model that includes temperature dependence of the optical properties, using the Kramers-Kronig transformation type. In the mathematical expressions, were included mix combined term in order to avoid the effects of Runge phenomenon. As a case study was chosen Bi2O3 — a substance less studied in literature. In the last part are the presented and commented the results obtained for a series of eight studied models.