Faculty Publications

Permanent URI for this communityhttps://idr.nitk.ac.in/handle/123456789/18736

Publications by NITK Faculty

Browse

Author: search.filters.author.Moger, T.Subject: search.filters.subject.Linear programming

Search Results

Now showing 1 - 2 of 2
  • No Thumbnail Available
    Item
    Fuzzy logic approach for reactive power coordination in grid connected wind farms to improve steady state voltage stability
    (Institution of Engineering and Technology journals@theiet.org, 2017) Moger, T.; Dhadbanjan, T.
    This study presents a fuzzy logic approach for reactive power coordination in grid connected wind farms with different types of wind generator units to improve steady state voltage stability of power systems. The load bus voltage deviation is minimised by changing the reactive power controllers according to their sensitivity using fuzzy set theory. The proposed approach uses only few controllers of high sensitivity to achieve the desired objectives. The 297-bus and 417-bus equivalent grid connected wind systems are considered to present the simulation results. To prove the effectiveness of the proposed approach, a comparative analysis is carried out with the conventional linear programming based reactive power optimisation technique. Results demonstrated that the proposed approach is more effective in improving the system performance as compared with the conventional existing technique. © 2016 The Institution of Engineering and Technology.
  • No Thumbnail Available
    Item
    A Novel Transformer-Based Approach for Reliability Evaluation of Composite Systems With Renewables and Plug-in Hybrid Electric Vehicles
    (Institute of Electrical and Electronics Engineers Inc., 2025) Yarramsetty, C.; Moger, T.; Jena, D.; Rao, V.S.
    This paper proposes a novel hybrid framework that integrates machine learning (ML) techniques with Sequential Monte Carlo Simulation (SMCS) to enhance the reliability assessment of modern power systems incorporating renewable energy resources (RER) and plug-in hybrid electric vehicle (PHEVs) integration. While PHEVs can leverage RER to significantly reduce greenhouse gas emissions, the increased energy demand from large PHEVs fleets poses potential challenges to power system reliability. To address these issues, this research presents an advanced mixed-integer linear programming (MILP) based algorithm for optimizing EV charging. The algorithm prioritizes clean energy utilization through intelligent power allocation strategies while considering cost-revenue trade-offs. A probabilistic model is developed to account for factors such as driving distance, charging times, locations, battery state of charge, and charging needs of PHEVs. The proposed approach is tested on the IEEE RTS-79 test system and evaluates multiple ML architectures, including Long Short-Term Memory (LSTM), Bidirectional LSTM (BiLSTM), and Transformer models, often combined with boosting algorithms, across three scenarios: base case, uncontrolled charging, and intelligent charging. Results highlight that ML-based approaches, particularly the Transformer model, achieve computational time reductions of up to 49% compared to traditional SMCS methods while maintaining comparable accuracy. The Transformer model identified 1,788 loss-of-load states compared to 1,510 actual instances, requiring only 176 minutes of computation. Among all models, the BiLSTM with Adaptive Boosting (BiLSTM+AB) achieved the lowest overestimation, exceeding actual instances by just 256 states. Performance metrics such as Loss of Load Probability (LOLP) and Expected Demand Not Supplied (EDNS) validate the effectiveness of the proposed ML approaches in balancing accuracy and computational efficiency. © 2013 IEEE.