Marijana Pavlov-Kagadejev

Marijana Pavlov-Kagadejev

Auto-generated profile, based on publicly available information

Marijana Pavlov-Kagadejev

Mining and Metallurgy Institute Bor Alberta Ajnštajna 1 19210 Bor Serbia

Research Focus & Metrics

Academic Career Length

17Years (2010-Present)

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

Representative Publications

View All Publications
12 February 2024 in Artificial Intelligence Review
Verified
Journal Article
Open access
PDF file
MPMarijana Pavlov-Kagadejev
Marijana Pavlov-Kagadejev
Luka Dragoljub JovanovicLuka Dragoljub Jovanovic
Luka Dragoljub Jovanovic
NEBOJSA BACANINNEBOJSA BACANIN
NEBOJSA BACANIN
MDMuhammet Deveci
Muhammet Deveci
Miodrag ZivkovicMiodrag Zivkovic
Miodrag Zivkovic
WPWitold Pedrycz
Witold Pedrycz
Power supply from renewable energy is an important part of modern power grids. Robust methods for predicting production are required to balance production and demand to avoid losses. This study proposed an approach that incorporates signal decomposition techniques with Long Short-Term Memory (LSTM) neural networks tuned via a modified metaheuristic algorithm used for wind power generation forecasting. LSTM networks perform notably well when addressing time-series prediction, and further hyperparameter tuning by a modified version of the reptile search algorithm (RSA) can help improve performance. The modified RSA was first evaluated against standard CEC2019 benchmark instances before being applied to the practical challenge. The proposed tuned LSTM model has been tested against two wind production datasets with hourly resolutions. The predictions were executed without and with decomposition for one, two, and three steps ahead. Simulation outcomes have been compared to LSTM networks tuned by other cutting-edge metaheuristics. It was observed that the introduced methodology notably exceed other contenders, as was later confirmed by the statistical analysis. Finally, this study also provides interpretations of the best-performing models on both observed datasets, accompanied by the analysis of the importance and impact each feature has on the predictions.

1 January 2015 in Serbian Journal of Electrical Engineering
Verified
Journal Article
Open access
PDF file
RRRadojle Radetic
Radojle Radetic
MPMarijana Pavlov-Kagadejev
Marijana Pavlov-Kagadejev
NMNikola Milivojevic
Nikola Milivojevic
The most exact temperature measurement can be made by using platinum sensors. Temperatures from -254.3?C up to +850?C can be measured with Pt100 sensor. The relationship between resistance and temperature is relatively linear, but for measurements of very high precision, Pt100 working curve should be a little bit improved. The paper describes an efficient way of measurement characteristic linearization by using the analogue electric circuits. The obtained results proved the initial considerations and the Pt100 becomes rather transducer than pure sensor.

1 January 2013 in Serbian Journal of Electrical Engineering
Verified
Journal Article
Open access
PDF file
RRRadojle Radetic
Radojle Radetic
MPMarijana Pavlov-Kagadejev
Marijana Pavlov-Kagadejev
DBDarko Brodic
Darko Brodic
NMNikola Milivojevic
Nikola Milivojevic
The paper presents the real instrument functional characteristics and describes the way of practical solutions of its performance improvement. It presents the design process of the instrument made for resistance measuring. In order to achieve desired objectives, a great number of experiments have been carried out during the development. Basically, the comparison method has been applied. At first, it was intended for the small resistor measuring as a single range unit. Later, the device has been improved and upgraded for a wide range resistance measuring. Finally, some of the difficulties have been detected and explained as well. The paper contains solutions developed and applied for their overcoming.

1 January 2010 in Measurement Science Review
Verified
Journal Article
Open access
RRR. Radetic
R. Radetic
DMD. Milivojevic
D. Milivojevic
MPM. Pavlov
M. Pavlov

1 January 2016 in Mining and Metallurgy Engineering Bor
Verified
Journal Article
Closed access
PDF file
MPMarijana Pavlov-Kagadejev
Marijana Pavlov-Kagadejev
RRRadojle Radetic
Radojle Radetic
VDVladimir Despotovic
Vladimir Despotovic
The Mining and Smelting Complex Bor (RTB) is one of the biggest electrical energy consumers in Serbia, with costs that exceed 3 million USD per month, or between 30 and 40 million USD per year. A complex technological process of copper production comprises mining, concentrate production (flotation), smelting and electrolytic refining. In all these stages a large amount of energy is consumed. The electricity is supplied from the power plants using 5 substations: two of them located in Bor, and the remaining three in Majdanpek. The conditions for delivery and billing of the electrical energy were defined by the contract with the supplier. The methods of billing (calculation) the electrical energy costs were changed several times in the previous 25 years. Although the consumed energy is the largest part of the overall costs, there were also some indirect costs affecting the final monthly price. It is clear that even a minor percentage decrease of costs can lead to substantial savings in the total amount. Hence, different technological and organizational activities were undertaken to reduce these costs. The paper presents primarily the effects of administrative measures on the amount of total electrical energy costs in copper production.

1 January 2024 in Bakar
Verified
Journal Article
Open access
PDF file
MPMarijana Pavlov-Kagadejev
Marijana Pavlov-Kagadejev
MRMilan Radivojević
Milan Radivojević
PSPredrag Stolic
Predrag Stolic
This paper describes an example of applying the machine learning to predict the SO2 blower vibrations. The SVR (Support Vector Regression) algorithm was used to form the model. The paper includes an overview of machine learning and mathematical fundamentals of the SVM model. Based on a large amount of data, obtained from the monitoring system, a mathematical model was constructed. After testing, the measures of model accuracy such as the MSE (Mean Squared Error) and R2 were provided.