Kaidi Ji
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Kaidi Ji has a total of 22 co-authors. Here are the top co-authors by number of shared publications:
- Yinlan Ruan
School of Optoelectronic Engineering, Guilin University of Electronic Technology, Guilin 541004, China
2 shared publications - Zhen Li
Subphotonic (Shenzhen) Technologies Co., Ltd., Nanning 530000, China
2 shared publications - Aofei Li
School of Life & Environmental Sciences, Guilin University of Electronic Technology, Guilin 541004, China
1 shared publication Biao Chen
School of Life and Environmental Science Guilin University of Electronic Technology Guilin P.R. China
1 shared publication- Biao Chen
School of Life & Environmental Sciences, Guilin University of Electronic Technology, Guilin 541004, China
1 shared publication - Boxue Chang
School of Life & Environmental Sciences, Guilin University of Electronic Technology, Guilin 541004, China
1 shared publication
Latest Publications
- 8 June 2025
- Published by Wiley in Biotechnology and Bioengineering
- Vol. 122 (9) , 2366-2376
Data‐Augmented Deep Learning Algorithm for Accurate Control of Bioethanol Fermentation Using an Online Raman Analyzer
Fed‐batch fermentation has become the preferred strategy in many industrial biomanufacturing processes. However, a key challenge remains in optimizing the feeding strategy to achieve stable maximum yields. In this study, we present an online Raman spectroscopy‐based monitoring and control system, using bioethanol production by Saccharomyces cerevisiae as a case study. To address the issue of limited labeled data, a pseudo‐labeling approach based on semi‐supervised learning was employed, expanding the available training data set by 100‐fold compared to conventional labeling methods. In addition, we developed a spectral‐temporal concatenation convolutional neural network (STC‐CNN) that incorporates sequential spectral features. Comparative evaluations with multiple machine learning algorithms demonstrated the superior performance of STC‐CNN, achieving a root mean square error (RMSE) of 3.63 g/L for glucose prediction. The system enabled rapid and automated glucose feeding to maintain various target concentrations. Notably, a glucose setpoint of 30 g/L yielded the highest ethanol concentration of 140.68 g/L—an increase of 3.85% over traditional Fed‐batch fermentation—while reducing glycerol by 6.67%. These results highlight the significant potential of Raman spectroscopy combined with deep learning for automated bioprocess optimization and discovery of optimal operating strategies.
- 8 June 2025
- Published by Wiley in Biotechnology and Bioengineering
- Vol. 122 (9) , 2366-2376
Data‐Augmented Deep Learning Algorithm for Accurate Control of Bioethanol Fermentation Using an Online Raman Analyzer
Fed‐batch fermentation has become the preferred strategy in many industrial biomanufacturing processes. However, a key challenge remains in optimizing the feeding strategy to achieve stable maximum yields. In this study, we present an online Raman spectroscopy‐based monitoring and control system, using bioethanol production by Saccharomyces cerevisiae as a case study. To address the issue of limited labeled data, a pseudo‐labeling approach based on semi‐supervised learning was employed, expanding the available training data set by 100‐fold compared to conventional labeling methods. In addition, we developed a spectral‐temporal concatenation convolutional neural network (STC‐CNN) that incorporates sequential spectral features. Comparative evaluations with multiple machine learning algorithms demonstrated the superior performance of STC‐CNN, achieving a root mean square error (RMSE) of 3.63 g/L for glucose prediction. The system enabled rapid and automated glucose feeding to maintain various target concentrations. Notably, a glucose setpoint of 30 g/L yielded the highest ethanol concentration of 140.68 g/L—an increase of 3.85% over traditional Fed‐batch fermentation—while reducing glycerol by 6.67%. These results highlight the significant potential of Raman spectroscopy combined with deep learning for automated bioprocess optimization and discovery of optimal operating strategies.
- Open Access
Rapid Quantitative Determination of Adulteration of Camellia Oil Using Portable Raman Spectroscopy and Chemometrics
Over the past decade, Raman spectroscopy and chemometrics have been extensively utilized in the food industry for the research and development of new products but have failed to establish a strong foothold in quality control and the assessment of food items. To bridge this gap, we introduce a novel application of Raman spectroscopy capable of swiftly identifying free fatty acids (FFAs) in cooking oil and quantifying adulteration. This advanced method was validated using camellia oil, a highly esteemed cooking oil in China and various Asian countries known for its nutritional richness and diverse culinary applications. With its growing popularity among high-end food consumers in Asia, camellia oil has increasingly become a target for adulteration, causing dissatisfaction among both consumers and genuine producers. In this study, we employed Raman spectroscopy to characterize the FFA profiles of cooking oil samples, complemented by principal component analysis (PCA) and partial least squares-discriminant analysis (PLS-DA) for sample categorization and adulteration detection in camellia oil. By segregating camellia oil from other vegetable oils and differentiating genuine from adulterated samples using the partial least squares (PLS) method, we achieved a high determination coefficient (R2) of over 0.98 and a low root mean square error of prediction (RMSEP) of less than 1.45%. These findings offer a robust predictive model for rapid camellia oil adulteration assessment, potentially augmenting traditional qualitative tests and streamlining sampling procedures in the food industry.
Open AccessRapid Quantitative Determination of Adulteration of Camellia Oil Using Portable Raman Spectroscopy and Chemometrics
Over the past decade, Raman spectroscopy and chemometrics have been extensively utilized in the food industry for the research and development of new products but have failed to establish a strong foothold in quality control and the assessment of food items. To bridge this gap, we introduce a novel application of Raman spectroscopy capable of swiftly identifying free fatty acids (FFAs) in cooking oil and quantifying adulteration. This advanced method was validated using camellia oil, a highly esteemed cooking oil in China and various Asian countries known for its nutritional richness and diverse culinary applications. With its growing popularity among high-end food consumers in Asia, camellia oil has increasingly become a target for adulteration, causing dissatisfaction among both consumers and genuine producers. In this study, we employed Raman spectroscopy to characterize the FFA profiles of cooking oil samples, complemented by principal component analysis (PCA) and partial least squares-discriminant analysis (PLS-DA) for sample categorization and adulteration detection in camellia oil. By segregating camellia oil from other vegetable oils and differentiating genuine from adulterated samples using the partial least squares (PLS) method, we achieved a high determination coefficient (R2) of over 0.98 and a low root mean square error of prediction (RMSEP) of less than 1.45%. These findings offer a robust predictive model for rapid camellia oil adulteration assessment, potentially augmenting traditional qualitative tests and streamlining sampling procedures in the food industry.
- Open Access
Automatic Control of Nucleation and Crystal Growth Using Online Raman Analyzer
The accurate determination of crystal formation during crystallization is crucial for obtaining crystal products with consistent quality and quantity. In this study, we aimed to identify the feasibility of using Raman spectroscopy to monitor the crystal growth stage in the crystallization process using cephalosporin intermediate 7-ACT as an example molecule. By observing the changes in the characteristic peak of the 7-ACT crystal (504 cm−1) and the characteristic peak of the solvent acetonitrile (914 cm−1), a correlation between the crystal growth stage and the change in the Raman intensity of the crystal solution was discovered. The determination of the optimal starting time for the crystal growth stage through a Raman analyzer significantly improves the consistency of crystal product quality. This led to a fivefold reduction in the variation in the weight and water content of the final 7-ACT crystal products compared to those obtained via manual control. In addition, our experiments also indicated that Raman monitoring could be more efficient at enabling the chemical synthesis reaction to be completed compared to manual control. Thus, our work demonstrates the potential of Raman spectroscopy in the real-time control of chemical synthesis reactions and crystallization processes.
Open AccessAutomatic Control of Nucleation and Crystal Growth Using Online Raman Analyzer
The accurate determination of crystal formation during crystallization is crucial for obtaining crystal products with consistent quality and quantity. In this study, we aimed to identify the feasibility of using Raman spectroscopy to monitor the crystal growth stage in the crystallization process using cephalosporin intermediate 7-ACT as an example molecule. By observing the changes in the characteristic peak of the 7-ACT crystal (504 cm−1) and the characteristic peak of the solvent acetonitrile (914 cm−1), a correlation between the crystal growth stage and the change in the Raman intensity of the crystal solution was discovered. The determination of the optimal starting time for the crystal growth stage through a Raman analyzer significantly improves the consistency of crystal product quality. This led to a fivefold reduction in the variation in the weight and water content of the final 7-ACT crystal products compared to those obtained via manual control. In addition, our experiments also indicated that Raman monitoring could be more efficient at enabling the chemical synthesis reaction to be completed compared to manual control. Thus, our work demonstrates the potential of Raman spectroscopy in the real-time control of chemical synthesis reactions and crystallization processes.