Embedded Files
Publications
Publications
01. Generalized ML models to predict significant wave height utilizing wind and atmospheric parameters
01. Generalized ML models to predict significant wave height utilizing wind and atmospheric parameters
Fig: (a) Overall Modeling, (b) Data Pre-processing, (c) Scatter Plot on Test Data, (d) Performance on Randomly Selected Stations
Published: Energy Conversion and Management X: (2024)
Machine learning models ANN, SNN, XGBoost and LightGBM were employed to predict significant wave height.
Pioneering study to have models trained on 47 stations along the North American coast.
Model performance was evaluated on six new stations (DS3).
Deep learning models achieved R2 score above 0.90 on training data.
Gradient boosting models achieved R2 score around 0.70 on DS3.
02. Exploring SAARC’s ocean energy potential: Current status and future policies
02. Exploring SAARC’s ocean energy potential: Current status and future policies
Published: Energy Reports (2024)
Overview of ocean renewable energy (ORE) harvesting methods provided.
Analysis & summary of ORE potential in 5 SAARC countries.
Analysis of energy policy up to 2030 based on SDG and COP pledges in SAARC countries.
Discussion of drivers & barriers to ORE in SAARC countries based on existing systems.
Proposal for research & energy production direction based on the review.
03. Multi-objective optimization and 4E (energy, exergy, economy, environmental impact) analysis of a Triple Cascade Refrigeration System
03. Multi-objective optimization and 4E (energy, exergy, economy, environmental impact) analysis of a Triple Cascade Refrigeration System
Fig: (a) System Schematics, (b) & (d) Parametric Analysis, (c) Pareto Front and Optimum Point
Published: Heliyon (2024)
A triple cascade refrigeration system with hydrocarbon refrigerants was simulated.
Conducted parametric analysis based on 4E (Energy, Exergy, Economic and Environmental Impact).
Employed multi-objective optimization for optimum operating conditions.
Optimum COP, exergy efficiency, and plant cost were 0.71, 0.51, and 38262.05$/year, respectively.
HTC compressor and HTC throttle were the major contributors (35%) to exergy destruction.
04. Advanced exergy analysis and machine learning based multi-objective optimization of a modified triple cascade refrigeration system for enhanced performance
04. Advanced exergy analysis and machine learning based multi-objective optimization of a modified triple cascade refrigeration system for enhanced performance
Fig: (a) System Schematics, (b) Parametric Analysis, (c) Population diagram, (d) Pareto Front and Optimum Point
Published: Applied Thermal Engineering (2025)
A novel triple cascade refrigeration system is simulated for enhanced efficiency.•
Sensitivity analysis covers energy, exergy, economic, and environmental impacts.
ANN integrated multi-objective optimization is performed using NSGA- Ⅱ algorithm.
Optimum COP, CO2 emission and cost: 0.7, 62119 kg and 33,728 $ per year.
Advanced exergy destruction shows Evaporator has highest improvement potential.
M.Sc. Thesis
M.Sc. Thesis
Performance Forecasting and Machine Learning Based Optimization of a Concentrated Solar Energy-Powered Combined Power, Cooling, and Desalination Plant (2024-Present)
Performance Forecasting and Machine Learning Based Optimization of a Concentrated Solar Energy-Powered Combined Power, Cooling, and Desalination Plant (2024-Present)
Fig: Proposed Hybrid Cycle for Simultaneous Power, Cooling and Fresh Water
Supervisor: Prof. Dr. Mohammad Ahsan Habib
Co-Supervisor: Prof. Dr. Mohammad Monjurul Ehsan
The thesis has the following objectives:
To propose and simulate a concentrated solar energy-operated combined power, ultra-low temperature cooling, and desalination plant.
To conduct an in-depth thermal and advanced exergo-economic performance assessment and machine learning-based multi-objective optimization of the proposed system.
To forecast the performance of the integrated system under variant climate conditions using advanced machine learning algorithms.
B.Sc. Thesis
B.Sc. Thesis
Significant Wave Height Prediction Using Machine Learning Techniques (2022-2023)
Significant Wave Height Prediction Using Machine Learning Techniques (2022-2023)
Fig: Outline of the proposed thesis
Supervisor: Prof. Dr. Mohammad Ahsan Habib
Machine learning models ANN, SNN, XGBoost and LightGBM were employed to predict significant wave height from wind and atmospheric parameters.
Pioneering study to have models trained on 47 stations along the North American coast.
Deep learning models outperformed tree-based models and showed superior generalizability.
Page updated
Google Sites
Report abuse