一项基于多变量数据融合和KNN-Attention-LSTM-XGBoost建模的Daqu时间序列预测和酸度分析研究
Haili Yang1, Hao Xia2, Sai Liu3
1School of Mechanical Engineering, Sichuan University of Science and Engineering, Yibin, 644000, China.
这项研究引入了一种新的KNN-Attention-LSTM-XGBoost模型,用于实时预测发酵期间的Daqu酸度,改善传统中国造品质的质量控制.
科学领域:
- 生物技术和发酵科学 生物技术和发酵科学
- 数据科学和机器学习
背景情况:
- 达库对于中国造酒至关重要,但由于传统方法的局限性,其酸度监测具有挑战性.
- 实时酸度预测对于优化Daqu发酵和确保产品质量至关重要.
研究的目的:
- 开发和验证一个时间序列预测模型,用于发酵过程中的Daqu酸度.
- 为了解决达库造中传统酸度测量方法的局限性.
主要方法:
- 在发酵过程中收集了Daqu的微环境参数.
- 使用XGBoost进行最佳数据归算 (LFBI,KNN) 和PLSR进行关键参数提取.
- 应用延迟和滚动窗口方法用于时间特征提取.
- 开发和评估了一个KNN-Attention-LSTM-XGBoost模型用于酸度预测.
主要成果:
- 该KNN-Attention-LSTM-XGBoost模型实现了高预测精度 (R2高达0.9790).
- 这个模型在不同的Daqu层中显著优于LSTM-XGBoost和XGBoost模型.
- 通过KNN预处理,提高了Attention-LSTM-XGBoost模型的预测性能.
结论:
- 拟议的模型提供了一种有效的解决方案,用于实时监测Daqu酸度.
- 这一进步可以带来优化的Daqu发酵过程和提高造质量.
- 该研究强调了机器学习在传统发酵工业中的潜力.
更多相关视频
06:19Integration of Animal Behavioral Assessment and Convolutional Neural Network to Study Wasabi-Alcohol Taste-Smell Interaction
Published on: August 16, 2024
10:25Construction of Models for Nondestructive Prediction of Ingredient Contents in Blueberries by Near-infrared Spectroscopy Based on HPLC Measurements
Published on: June 28, 2016
相关概念视频
End Point Prediction: Gran Plot
For potentiometric titration, the Gran plot is created by plotting...
Acid–Base Titration: Overview
Acid-Base Titration Curves
For a titration carried out for 25.00 mL of...
Acidity and Basicity of Carboxylic Acid Derivatives
The relative acidic strength of the derivatives can be explained based on the extent of resonance stabilization of the conjugate base. The...
Acidity of Carboxylic Acids
Ladder Diagrams: Acid–Base Equilibria
A ladder diagram for acid-base equilibria consists of a vertical axis that represents pH and horizontal bars (steps on the ladder) that help position all the pKa values in the system. At equilibrium, the pH value of the system corresponds to one of...
