Eric Ofori-Ntow Jnr
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Eric Ofori-Ntow Jnr has a total of 24 co-authors. Here are the top co-authors by number of shared publications:
- Yao Yevenyo Ziggah
Geomatic Engineering Department, University of Mines and Technology, Tarkwa, Ghana
5 shared publications - Frederick Abangba Akendola3 shared publications
- Susana Relvas
CEG-IST, Instituto Superior Técnico, Universidade de Lisboa
3 shared publications Yao Yevenyo Ziggah
3 shared publicationsMaria Joao Rodrigues
2 shared publicationsAbdul Somad Alhassan
1 shared publication
Latest Publications
- Open Access
- 6 October 2026
- Published by Elsevier in Energy Conversion and Management: X
- Vol. 32
A novel three-level hybrid model for wind speed forecasting based on phase-space reconstruction and optimized backpropagation neural network
- 6 October 2026
- Published by Elsevier in Energy Conversion and Management: X
- Vol. 32
Open AccessA novel three-level hybrid model for wind speed forecasting based on phase-space reconstruction and optimized backpropagation neural network
- Open Access
- 24 September 2026
- Published by Elsevier in SSRN Electronic Journal
Predicting Photovoltaic Power with Confidence: A Combined Interpretable Machine Learning and Conformal Prediction Approach
Photovoltaic (PV) power is known to be environmentally friendly but difficult to integrate into the national grid because of the challenges in handling it. The conventional method of forecasting PV power availability and system reliability presents challenges by system reliability. This,study therefore proposed a transformer model called Tabular Prior-data Fitted Network (Tab-PFN) for forecasting PV power using data from Ghana. The proposed model is compared with three similar models, including the Kolmogorov-Arnold Network (KAN), Tabular-Large Language Model (Tab-LLM), and Tabular Neural Network (TabNet). The performance of the models is assessed using Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), Coefficient of Regression (R), Scatter Index (SI), and Bayesian Information Criteria (BIC). Tab-PFN model led in all indicators where it achieved RMSE of 0.8359, MAPE of 0.0453, R of 0.9999, SI of 7.7138, and BIC of -1178.69 during testing. Explainable Artificial Intelligence (XAI) techniques are further used to determine the impacts of the various features on the PV power forecasting. Global Irradiance was found to be the feature having the most impact on the PV power forecasting using the Tab-PFN model.
- 24 September 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessPredicting Photovoltaic Power with Confidence: A Combined Interpretable Machine Learning and Conformal Prediction Approach
Photovoltaic (PV) power is known to be environmentally friendly but difficult to integrate into the national grid because of the challenges in handling it. The conventional method of forecasting PV power availability and system reliability presents challenges by system reliability. This,study therefore proposed a transformer model called Tabular Prior-data Fitted Network (Tab-PFN) for forecasting PV power using data from Ghana. The proposed model is compared with three similar models, including the Kolmogorov-Arnold Network (KAN), Tabular-Large Language Model (Tab-LLM), and Tabular Neural Network (TabNet). The performance of the models is assessed using Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), Coefficient of Regression (R), Scatter Index (SI), and Bayesian Information Criteria (BIC). Tab-PFN model led in all indicators where it achieved RMSE of 0.8359, MAPE of 0.0453, R of 0.9999, SI of 7.7138, and BIC of -1178.69 during testing. Explainable Artificial Intelligence (XAI) techniques are further used to determine the impacts of the various features on the PV power forecasting. Global Irradiance was found to be the feature having the most impact on the PV power forecasting using the Tab-PFN model.
- Open Access
- 26 August 2026
- Published by Elsevier in SSRN Electronic Journal
A Novel Three-Level Hybrid Model for Wind Speed Forecasting Based on Phase-Space Reconstruction and Optimized Backpropagation Neural Network
Many countries are switching attention from traditional sources of generating electricity because of sustainability issues. Wind energy is one of the sustainable sources of generating electricity, which has received much attention. The unstable nature of wind speed is becoming a challenge to handle for power generators. This study is therefore about developing a new three-level hybrid model for forecasting unstable wind speed. The hybrid models include chaos theory of Phase Space Reconstruction (PSR), Fox Optimization Algorithm (FOA) and Backpropagation Neural Network (BPNN) to form a novel PSR-FOA-BPNN model. The developed hybrid model’s performance was compared with five different but similar hybrid models for fair competition. The statistical model assessors of Mean Absolute Error (MAE), Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), and Scatter Index (SI) were used for the comparison. When compared against PSR-AOA-BPNN, PSR-ACVO-BPNN, PSR-GTO-BPNN, PSR-RSA-BPNN, and PSR-BPNN hybrid models, the statistical findings demonstrated that the suggested PSR-FOA-BPNN model performed better and is thus regarded as an effective wind speed prediction tool. For training, testing, and validation, the proposed hybrid model had the lowest MAPE values of 0.516 %, 0.232 %, and 0.1923 % respectively. While the SI scores for training, testing, and validation were 0.3219, 0.2302, and 0.2032, respectively, the RMSE values were 1.2662 m/s for training, 1.2404 m/s for testing, and 0.9626 m/s for validation. Therefore, taking into account the PSR-FOA-BPNN results, the indication confirms that this model may be used for effective wind speed forecasting for national grid operations.
- 26 August 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessA Novel Three-Level Hybrid Model for Wind Speed Forecasting Based on Phase-Space Reconstruction and Optimized Backpropagation Neural Network
Many countries are switching attention from traditional sources of generating electricity because of sustainability issues. Wind energy is one of the sustainable sources of generating electricity, which has received much attention. The unstable nature of wind speed is becoming a challenge to handle for power generators. This study is therefore about developing a new three-level hybrid model for forecasting unstable wind speed. The hybrid models include chaos theory of Phase Space Reconstruction (PSR), Fox Optimization Algorithm (FOA) and Backpropagation Neural Network (BPNN) to form a novel PSR-FOA-BPNN model. The developed hybrid model’s performance was compared with five different but similar hybrid models for fair competition. The statistical model assessors of Mean Absolute Error (MAE), Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), and Scatter Index (SI) were used for the comparison. When compared against PSR-AOA-BPNN, PSR-ACVO-BPNN, PSR-GTO-BPNN, PSR-RSA-BPNN, and PSR-BPNN hybrid models, the statistical findings demonstrated that the suggested PSR-FOA-BPNN model performed better and is thus regarded as an effective wind speed prediction tool. For training, testing, and validation, the proposed hybrid model had the lowest MAPE values of 0.516 %, 0.232 %, and 0.1923 % respectively. While the SI scores for training, testing, and validation were 0.3219, 0.2302, and 0.2032, respectively, the RMSE values were 1.2662 m/s for training, 1.2404 m/s for testing, and 0.9626 m/s for validation. Therefore, taking into account the PSR-FOA-BPNN results, the indication confirms that this model may be used for effective wind speed forecasting for national grid operations.
- Open Access
- 3 July 2026
- Published by Elsevier in Results in Surfaces and Interfaces
- Vol. 24
Experimental and computational analysis of ultrasonic clutch disc wear detection system for vehicles
- 3 July 2026
- Published by Elsevier in Results in Surfaces and Interfaces
- Vol. 24
Open AccessExperimental and computational analysis of ultrasonic clutch disc wear detection system for vehicles
- Open Access
- 24 June 2026
- Published by Elsevier in SSRN Electronic Journal
Multi-scale Morphological Evolution and Instability Decoupling Mechanisms of Turbulent Ammonia/Methanol Premixed Flames
Ammonia (NH3) is a critical zero-carbon energy carrier, but its practical application is limited by poor reactivity and complex turbulent flame dynamics. This study investigates the multi-scale morphological evolution and instability decoupling mechanisms of turbulent NH3/CH3OH premixed flames in a constant-volume combustion chamber. Experiments spanned equivalence ratios of 0.8-1.2, with in-depth morphological analyses strictly anchored at the stoichiometric condition (ϕ = 1.0) under initial pressures of 2–4 bar, methanol fractions of 20%–40%, and turbulence intensities of 2.682, 4.023, and 5.364 m/s. An analytical framework—integrating contour unfolding, curvature statistics, and Power Spectral Density (PSD) analysis—quantified the spatial energy cascade. Observations reveal a counterintuitive scale-segregated evolution: rather than amplifying global wrinkling, intensified intrinsic instabilities (via elevated pressure or methanol) suppress macroscopic fluctuation amplitudes while triggering significant micro-scale fragmentation. This establishes a spatiotemporal decoupling framework where external turbulence and intrinsic instabilities evolve independently. Linear stability theory indicates high pressure compresses laminar flame thickness, compromising thermal-diffusive smoothing and causing spatial geometric saturation via cell merging. Concurrently, methanol enrichment shortens the evolution time window, truncating the temporal growth of macroscopic modes. PSD analysis confirms frequency-domain segregation: macroscopic turbulence dominates kinetic energy injection at low wavenumbers, while intrinsic physicochemical properties govern high-frequency breakdown. These insights physically explain the bending effect (early saturation of turbulent burning velocity), providing geometric constraints for predictive combustion models in zero-carbon heavy-duty engines.
- 24 June 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessMulti-scale Morphological Evolution and Instability Decoupling Mechanisms of Turbulent Ammonia/Methanol Premixed Flames
Ammonia (NH3) is a critical zero-carbon energy carrier, but its practical application is limited by poor reactivity and complex turbulent flame dynamics. This study investigates the multi-scale morphological evolution and instability decoupling mechanisms of turbulent NH3/CH3OH premixed flames in a constant-volume combustion chamber. Experiments spanned equivalence ratios of 0.8-1.2, with in-depth morphological analyses strictly anchored at the stoichiometric condition (ϕ = 1.0) under initial pressures of 2–4 bar, methanol fractions of 20%–40%, and turbulence intensities of 2.682, 4.023, and 5.364 m/s. An analytical framework—integrating contour unfolding, curvature statistics, and Power Spectral Density (PSD) analysis—quantified the spatial energy cascade. Observations reveal a counterintuitive scale-segregated evolution: rather than amplifying global wrinkling, intensified intrinsic instabilities (via elevated pressure or methanol) suppress macroscopic fluctuation amplitudes while triggering significant micro-scale fragmentation. This establishes a spatiotemporal decoupling framework where external turbulence and intrinsic instabilities evolve independently. Linear stability theory indicates high pressure compresses laminar flame thickness, compromising thermal-diffusive smoothing and causing spatial geometric saturation via cell merging. Concurrently, methanol enrichment shortens the evolution time window, truncating the temporal growth of macroscopic modes. PSD analysis confirms frequency-domain segregation: macroscopic turbulence dominates kinetic energy injection at low wavenumbers, while intrinsic physicochemical properties govern high-frequency breakdown. These insights physically explain the bending effect (early saturation of turbulent burning velocity), providing geometric constraints for predictive combustion models in zero-carbon heavy-duty engines.
- Open Access
- 15 December 2025
- Published by Springer Nature in Discover Mechanical Engineering
- Vol. 4 (1)
An integrated approach for characterizing cocoa pod cutting dynamics using uniaxial shear loading technique
- 15 December 2025
- Published by Springer Nature in Discover Mechanical Engineering
- Vol. 4 (1)
Open AccessAn integrated approach for characterizing cocoa pod cutting dynamics using uniaxial shear loading technique
- Open Access
- 11 August 2025
- Published by Elsevier in Solar Compass
- Vol. 15
Design and integration of a solar photovoltaic system for mechanized and sustainable cocoa pod splitting
- 11 August 2025
- Published by Elsevier in Solar Compass
- Vol. 15
Open AccessDesign and integration of a solar photovoltaic system for mechanized and sustainable cocoa pod splitting
- 25 January 2025
- Published by Elsevier in Next Research
- Vol. 2 (1)
A novel multidimensional figure of merit for the evaluation of transparent electrode materials in optoelectronic devices
- 25 January 2025
- Published by Elsevier in Next Research
- Vol. 2 (1)
A novel multidimensional figure of merit for the evaluation of transparent electrode materials in optoelectronic devices
- Open Access
- 21 September 2023
- Published by Elsevier in e-Prime - Advances in Electrical Engineering, Electronics and Energy
- Vol. 6
Electricity demand forecasting based on feature extraction and optimized backpropagation neural network
- 21 September 2023
- Published by Elsevier in e-Prime - Advances in Electrical Engineering, Electronics and Energy
- Vol. 6
Open AccessElectricity demand forecasting based on feature extraction and optimized backpropagation neural network
- 18 April 2022
- Published by Springer Nature in Natural Resources Research
- Vol. 31 (3) , 1265-1287
A New Long-Term Photovoltaic Power Forecasting Model Based on Stacking Generalization Methodology
- 18 April 2022
- Published by Springer Nature in Natural Resources Research
- Vol. 31 (3) , 1265-1287
A New Long-Term Photovoltaic Power Forecasting Model Based on Stacking Generalization Methodology