JBJelena Markovic Brankovic

Jelena Markovic Brankovic

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Jelena Markovic Brankovic has a total of 18 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    AN INTEGRATED KPI-BASED FRAMEWORK FOR THE EVALUATION AND CLASSIFICATION OF NATURE-LIKE FISH PASSES – A CASE STUDY OF THE FREUDENAU HYDROPOWER PLANT

    by Milica Markovic, Jelena Markovic Brankovic

    Nature-like fish passes are complex hydraulic–ecological systems whose performance is often assessed through fragmented monitoring approaches. This paper proposes an integrated KPI-based framework for the evaluation and classification of fish pass functionality, combining biological, hydraulic, and operational indicators within a structured methodology supported by defined thresholds. A segment-based analytical approach enables classification as fully functional, selectively functional, or insufficiently functional. The framework is applied to the Freudenau fish pass (Danube River, Vienna), demonstrating how hydraulic and operational data can be integrated into a transparent and reproducible evaluation model for regulatory verification and long-term management.

  • Open Access

    ECOLOGICAL AND HYDROTECHNICAL ASPECTS OF FISH PASSAGE DESIGN – A CASE STUDY OF THE DJERDAP 2 HYDROPOWER PLANT

    by Jelena Markovic Brankovic, Milica Markovic, Tomislav Stojkovic, Dragan Radivojevic

    Fish passages represent a key engineering and ecological measure for restoring the longitudinal connectivity of watercourses disrupted by hydraulic structures. This paper presents the conceptual design of a nature-like fish passage at the Djerdap 2 hydropower plant on the Danube River. The study includes route definition, hydraulic sizing under complex site constraints, and assessment of ecological compatibility using literature-based criteria for relevant migratory fish species. The adopted concept was developed with reference to the fish passage at Freudenau in Vienna. The results indicate hydraulically feasible and ecologically acceptable internal flow conditions, while additional modelling of the entrance zone is recommended for final design verification.

  • Open Access

    INTEGRATED HYDRAULIC, ENERGY AND ECONOMIC ASSESSMENT OF TWO DIVERSION SYSTEM ALTERNATIVES FOR THE SHP “MERĆEZ”

    by Jelena Markovic Brankovic, Milica Markovic, Miljan Bankovic

    This paper presents an integrated hydraulic, energy and economic analysis of two diversion-system alternatives for the small hydropower plant “Merćez” on the Lukovska River. Two technical concepts are considered: Alternative I – water conveyance through a pressurized steel pipeline, and Alternative II – an open-channel diversion system with a forebay tank. The comparison is based on hydraulic losses, operating head, installed capacity, annual electricity production and investment requirements. The open-channel diversion option yields better energy metrics but demands higher investment. The economic analysis is restricted to comparative cost metrics.

  • Open Access

    Enhancing Multi-Horizon Probabilistic Water Level Forecasting Using Horizon- and Event-Aware Deep Learning Models

    by Jelena Marković Branković, Milica Marković, Bojan Branković

    Accurate multi-horizon forecasting of reservoir water levels is essential for effective water resource management and flood risk mitigation. While deep learning models have demonstrated strong predictive capabilities, they often struggle to adequately represent uncertainty and extreme hydrological events, particularly at longer forecast horizons. This study proposes four variants of a Conv1D–LSTM–Temporal Attention (CLTA) architecture for probabilistic multi-horizon forecasting, differing exclusively in loss function design. The models incorporate non-crossing constraints, horizon-aware weighting, and event-aware weighting to address key limitations of standard quantile regression approaches. All models are trained on hourly water level data from May 2021 to October 2022 and evaluated on a fully unseen dataset spanning December 2022 to May 2023. The results demonstrate that horizon-aware weighting achieves the lowest average RMSE (0.0149) and the most stable performance across forecast horizons on unseen data, while event-aware weighting improves representation of extreme hydrological events and achieves the highest coefficient of determination (R2=0.9961). However, a controlled experiment further reveals that model performance is strongly influenced by the data partitioning strategy, even when architecture and loss formulation are held constant. Overall, the findings indicate that loss function design, in interaction with data partitioning strategy, is a key contributing factor to model performance in deep learning-based hydrological forecasting. A Multi-Criteria Decision Analysis (MCDA) framework identifies the horizon-weighted model as the most robust general-purpose solution, while the event-aware model is preferable for applications focused on extreme event representation. These results highlight the importance of integrating domain knowledge into both model design and evaluation strategy, offering a scalable and computationally efficient alternative to increasing architectural complexity.

  • Open Access

    Hybrid Conv1D–LSTM Modelling of Short-Term Reservoir Water-Level Dynamics for Scenario-Based Operational Analysis

    by Jelena Marković Branković, Milica Marković, Bojan Branković

    Accurate representation of short-term reservoir water-level dynamics is essential for operational analysis and scenario-based assessment under prescribed inflow–outflow conditions. In many practical applications, physically based modelling is limited by incomplete process knowledge, unavailable boundary conditions, or insufficient temporal resolution of input data. This study presents a data-driven framework for hourly conditional simulation of reservoir water level based on a hybrid Conv1D–LSTM architecture. The model learns nonlinear relationships among hydraulic forcing, operational control, and system state from historical observations, and is evaluated in a recursive multi-step simulation (rollout) mode to reflect its intended use and capture error accumulation over time. A systematic analysis of input sequence length and activation function is performed to identify a robust model configuration. On the test set, the selected configuration (L = 24, GELU) achieved RMSE = 0.1057 m, MAE = 0.0881 m, and R2 = 0.972 in rollout evaluation. The proposed framework is designed for scenario-based simulation rather than one-step deterministic forecasting, enabling rapid operational screening of alternative inflow–outflow regimes. Unlike many previous studies that emphasize one-step predictive accuracy, this work explicitly assesses model stability in recursive multi-step simulation, which is more relevant for reservoir scenario analysis.

  • Open Access

    Hybrid Conv1D–LSTM Modelling of Short-Term Reservoir Water-Level Dynamics for Scenario-Based Operational Analysis

    by Jelena Markovic Brankovic, Milica Marković, Bojan Branković

    Accurate representation of short-term reservoir water-level dynamics is essential for operational analysis and scenario-based assessment under prescribed inflow–outflow conditions. In many practical applications, physically based modelling is limited by incomplete process knowledge, unavailable boundary conditions, or insufficient temporal resolution of input data. This study presents a data-driven framework for hourly conditional simulation of reservoir water level based on a hybrid Conv1D–LSTM architecture. The model learns nonlinear relationships among hydraulic forcing, operational control, and system state from historical observations, and is evaluated in a recursive multi-step simulation (rollout) mode to reflect its intended use and capture error accumulation over time. A systematic analysis of input sequence length and activation function is performed to identify a robust model configuration. On the test set, the selected configuration (L=24, GELU) achieved RMSE = 0.1057 m, MAE = 0.0881 m, and R² = 0.972 in rollout evaluation. The proposed framework is designed for scenario-based simulation rather than one-step deterministic forecasting, enabling rapid operational screening of alternative inflow–outflow regimes. Unlike many previous studies that emphasize one-step predictive accuracy, this work explicitly assesses model stability in recursive multi-step simulation, which is more relevant for reservoir scenario analysis.


  • Open Access

    Reservoir sedimentation as a green-dam performance indicator: Bovan case study

    by Milica Marković, Jelena Marković-Branković, Mladen Milanović

    Existing dams are usually assessed through structural safety, operation and maintenance, while the long-term performance of their reservoirs is often treated as a secondary issue. This paper places the reservoir at the centre of green-dam assessment and uses storage-capacity loss, storage-zone depletion and proximity to functional thresholds as complementary performance and risk indicators. Bovan Reservoir is analysed as the main case study and quantitatively compared with Bresnica and Rastovnica Reservoirs. The comparison shows that absolute sediment volume alone is not sufficient, because the consequences of sedimentation depend on reservoir size, useful storage, dead storage, intake elevation and operating rules. Bovan exhibits a substantial long-term reduction in storage, while Bresnica illustrates rapid depletion of dead storage and Rastovnica shows a sedimentation trend affecting the intake-related lower storage zone. Initial and updated storage-capacity curves are used to visualise these differences and demonstrate why repeatable bathymetric surveys are essential for reservoir management. A low-resource green/smart framework is proposed, prioritising updated reservoir geometry, reliable monitoring, data continuity, sediment-aware operation, catchment erosion-control measures and appropriate outlet management before advanced digital modelling is introduced. The results indicate that storage loss and reservoir sedimentation can serve as practical indicators of long-term reservoir performance when interpreted relative to reservoir size, storage zones and functional thresholds.

  • Open Access

    THERMAL STRATIFICATION AS AN OPERATIONAL INDICATOR FOR MULTIPURPOSE RESERVOIR MANAGEMENT: A SCADA-BASED CASE STUDY OF THE ĆELIJE RESERVOIR

    by Jelena Markovic Brankovic, Milica Markovic, Jelica Protic

    Reservoir thermal stratification is often described as a physical and limnological process, but its direct relevance for day-to-day reservoir operation is less frequently quantified using routinely available monitoring data. This paper investigates whether high-frequency Supervisory Control and Data Acquisition (SCADA) records can be used to identify periods in which thermal stratification becomes operationally significant for the Ćelije multipurpose reservoir in Serbia. The analysis combines hourly water-temperature measurements at six elevations, 255, 265, 268, 271, 274 and 277 m a.s.l., with reservoir level and outflow data for the period September 2022–September 2023. Thermal conditions were quantified using a vertical stratification index, ΔT = T277 − T255, while reservoir operation was interpreted through level–outflow relationships and total discharge intensity. Strong stratification, defined as ΔT ≥ 10°C, occurred for 109 days, with a maximum daily value of 17.38°C recorded on 17 July 2023. The longest uninterrupted strongly stratified period lasted 64 days, from 30 June to 4 September 2023. The results show that discharge intensity was primarily governed by reservoir level, but the ΔT index provided essential additional information for identifying periods when increased release coincided with a strongly stratified water column. These operating windows are particularly relevant for assessing outlet use, withdrawal depth and potential thermal effects on released water. The study demonstrates that simple indicators derived from SCADA data can support thermally informed reservoir operation and provide a practical basis for future selective-withdrawal and water-quality assessments.

  • ANALIZA POSTOJEĆEG STANJA I MOGUĆNOSTI SANACIJE MALE ZEMLJANE BRANE „POPOVA“ KOD BLACA

    by Jelena Marković Branković, Milica Marković

    U radu su prikazani rezultati analize postojećeg stanja male zemljane brane „Popova“ kod Blaca. Istraživanje je obuhvatilo geotehničke i hidrotehničke aspekte objekta, sa posebnim akcentom na pojavu klizišta na uzvodnoj kosini, degradaciju kosina usled erozionih procesa, zasutost prelivnog sistema i nefunkcionalnost temeljnog ispusta. Na osnovu raspoložive tehničke dokumentacije i terenskih opažanja izvršena je identifikacija mogućih uzroka nastalih deformacija i oštećenja, kao i procena njihovog uticaja na stabilnost tela brane i bezbednost akumulacije. U radu je data ocena postojećeg stanja objekta, funkcionalnosti evakuacionih organa i mogućnosti primene sanacionih mera u realnim uslovima eksploatacije. Rezultati sprovedene analize ukazuju na značaj adekvatnog izbora materijala, kontrole kvaliteta pri ugradnji i sistematskog održavanja u cilju obezbeđenja dugoročne stabilnosti i funkcionalne pouzdanosti malih zemljanih brana.

  • Open Access

    Analysis of Hollow Section “Y” Connections with the Application of Non-Linear Material Modeling

    by Srđan Živković, Nenad Stojković, Dragana Turnić, Marija Spasojević Šurdilović, Jelena Marković Branković

    The use of welded structural hollow sections in civil engineering is relatively new. Design of connections in steel structures with welded structural hollow sections requires a more specified approach, compared to traditional connections achieved with gusset plates. Stress and local deformations at the connection area are non-linear and very complex. This paper presents the application of non-linear numerical analysis for the determination of the ultimate bearing capacity of “Y” connections. In the analysis, a non-linear material model was applied. In order to validate the applied method of steel modelling, comparative analysis of the analytically and numerically determined ultimate bearing capacity with the experimentally obtained results has been conducted.