Next Article in Journal
European LNG Import Network Analysis and Investigation of Supply Security
Next Article in Special Issue
Few-Shot Load Forecasting Under Data Scarcity in Smart Grids: A Meta-Learning Approach
Previous Article in Journal
Numerical Analysis of Gas Flow Distribution Characteristics in a 5 kW Molten Carbonate Fuel Cell Stack
Previous Article in Special Issue
Transfer Learning for the Prediction of Energy Performance of Water-Cooled Electric Chillers: Grey-Box Models Versus Deep Neural Network (DNN) Models
 
 
Article
Peer-Review Record

Predictive Modeling of Energy Consumption for Cooling Ventilation in Livestock Buildings: A Machine Learning Approach

Energies 2025, 18(3), 633; https://doi.org/10.3390/en18030633
by Carlos Alejandro Perez Garcia, Patrizia Tassinari, Daniele Torreggiani and Marco Bovo *
Reviewer 1: Anonymous
Reviewer 2: Anonymous
Reviewer 3:
Energies 2025, 18(3), 633; https://doi.org/10.3390/en18030633
Submission received: 19 December 2024 / Revised: 15 January 2025 / Accepted: 28 January 2025 / Published: 30 January 2025
(This article belongs to the Special Issue Machine Learning for Energy Load Forecasting)

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This study utilizes NeuralProphet to predict the energy consumption for ventilation in a dairy cattle farm. While the study presents novel insights, the manuscript requires significant improvements in writing.

  • Abstract: The abstract should concisely summarize the main results of the study. Currently, it lacks a clear presentation of the key findings.

  • Introduction: The introduction should provide a comprehensive overview of recent research of this area. Additionally, the introduction should focus on content that is directly related to the study, as it currently includes unnecessary information that makes it difficult for readers to grasp the main objective.

  • Materials and Methods: This section should be dedicated to explaining the methodology, not presenting results. For example, lines 180-192 should be moved to the appropriate results section.

  • Figure 5: The figure lacks units.

  • Results and Discussion: This section should make comparisons to similar studies in the field. Currently, the authors have not compared their results to any existing research.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors The author did an interesting job to solve a problem about Energy Consumption for Cooling Ven-2 tilation in Livestock Buildings, by building upon concepts from the Prophet model and integrating neural net- work architectures. The methodology section describes the research design and result analysis, particularly providing detailed explanations on the data, design, and analysis of time series. Although the author provided explanations on the machine learning section, training data, and conclusions in sections 2.4 and 2.5, they were relatively simple, making it unclear how the author combined time series analysis with machine learning to strengthen the conclusions of time series analysis. I hope the author can add a framework diagram for this issue in the revised manuscript, providing a more detailed description of the connection between machine learning and time series analysis.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript entitled “Predictive Modeling of Energy Consumption for Cooling Ventilation in Livestock Buildings: A Machine Learning Approach” proposed a machine-learning-based framework for the energy consumption for cooling ventilation in Livestock buildings. In general, this paper used novel methods for Innovative scenario applications. With some modifications, the paper would be ready for publication. Here are my comments:

1. The heavy focus on time series analytics limits the scenario's innovation. More emphasis on location-specific feature engineering and scenario analysis is recommended.

2. NeuralProphet has limitations, mainly relying on temporal features, often univariate. Consider models that account for both temporal and spatial dependencies. Additionally, benchmarking for your model is missing.

3. Related prediction research can be compared: a: Short-term Load Forecasting of Distribution Transformer Supply Zones Based on Federated Model-Agnostic Meta Learning and b: Multi-node load forecasting based on multi-task learning with modal feature extraction

4. Figure plots require improvement.

5. More references are needed to validate the work.

6. The conclusions should offer deeper insights rather than merely summarizing the work.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

This manuscript can be accepted.

Back to TopTop