Scholarlink

<p>Energy consumption is a major concern in the food industry, particularly in the frozen food production process, where cooling systems account for a major proportion of energy usage. In this study, machine learning (ML) techniques were applied to develop predictive models for energy consumption and cooling load during the frozen food production process. Six ML models were evaluated: Auto-Regressive Integrated Moving Average (ARIMA), Long Short-Term Memory (LSTM), Random Forest (RF), Support Vector Regression (SVR), Extreme Gradient Boosting (XGBoost), and Gradient Boosting Trees (GBT). Model performance was assessed using standard evaluation metrics, with XGBoost achieving the highest overall accuracy and an R<sup>2</sup> of 0.93. By capturing the seasonality and complex, non-linear patterns often found in industrial data, the resulting prediction models attain improved accuracy. In energy-intensive food production settings, our results show how ensemble learning techniques, in particular XGBoost, have the potential to enhance energy forecasting and assist energy management.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Applying Machine Learning Concepts for Forecasting Energy Consumption in the Frozen Food Production Process

  • 期刊论文

  • 发表日期：

    2025年7月30日

查看全文
  • Thawanrat Sumrit,
  • Chanawat Nitatwichit,
  • Somkiat Jaturonglumlert,
  • Jaturapatr Varith

摘要

Energy consumption is a major concern in the food industry, particularly in the frozen food production process, where cooling systems account for a major proportion of energy usage. In this study, machine learning (ML) techniques were applied to develop predictive models for energy consumption and cooling load during the frozen food production process. Six ML models were evaluated: Auto-Regressive Integrated Moving Average (ARIMA), Long Short-Term Memory (LSTM), Random Forest (RF), Support Vector Regression (SVR), Extreme Gradient Boosting (XGBoost), and Gradient Boosting Trees (GBT). Model performance was assessed using standard evaluation metrics, with XGBoost achieving the highest overall accuracy and an R2 of 0.93. By capturing the seasonality and complex, non-linear patterns often found in industrial data, the resulting prediction models attain improved accuracy. In energy-intensive food production settings, our results show how ensemble learning techniques, in particular XGBoost, have the potential to enhance energy forecasting and assist energy management.