Article
Computer Science, Interdisciplinary Applications
Jinping Liu, Juanjuan Wu, Yongfang Xie, Wang Jie, Pengfei Xu, Zhaohui Tang, Huazhan Yin
Summary: This article proposes a robust nonlinear process monitoring scheme based on a denoising sparse auto-encoder (DSAE). The scheme addresses the challenges posed by high-dimensional, nonlinear, and complex industrial processes. The proposed method demonstrates promising results in fault monitoring experiments.
JOURNAL OF INDUSTRIAL INFORMATION INTEGRATION
(2022)
Article
Biotechnology & Applied Microbiology
Konrad Miller, Even Arefaine, Ardic Arikal, Annegret Cantu, Raul Cauduro Girardello, Anita Oberholster, Hildegarde Heymann, David E. Block
Summary: This study achieved significant time reduction in white wine fermentation by optimizing the relationship between yeast nutrient concentration and fermentation progression, resulting in wines with no significant sensory differences compared to traditional methods. This intensified fermentation process is of great importance for improving equipment utilization and resolving bottlenecks in fermentation facilities.
FERMENTATION-BASEL
(2022)
Article
Automation & Control Systems
Dazi Li, Fuqiang Zhu, Xiao Wang, Qibing Jin
Summary: Many real-world control problems involve conflicting objectives, and obtaining Pareto optimal solution sets for each objective is necessary. This study proposes a soft proximal policy optimization algorithm combined with a hybrid weight-generation method to find the Pareto front approximation of the fed-batch fermentation process. The algorithm aims to find a single policy for the multi-objective reinforcement learning problem and use a hybrid weight-generation method to find a set of Pareto optimal solutions.
JOURNAL OF PROCESS CONTROL
(2022)
Article
Biotechnology & Applied Microbiology
Nisa Saelee
Summary: Using squeezed sap from old oil palm trunks as a substrate, the LA yield and productivity can be improved by employing a modified constant feed fermentation mode.
FERMENTATION-BASEL
(2022)
Article
Computer Science, Interdisciplinary Applications
Hye Ji Lee, Jaehan Bae, Dong Hwi Jeong, Jong Min Lee
Summary: This paper proposes a real-time synchronization scheme based on relative proceeding rates to reduce synchronization errors and increase the accuracies of state estimation and fault detection. The utility of the proposed method is verified through a case study of industrial penicillin production with a real-time unmeasurable variable.
COMPUTERS & CHEMICAL ENGINEERING
(2021)
Article
Engineering, Multidisciplinary
Chongyang Liu, Chao Sun
Summary: This study presents a nonlinear impulsive time-delay system to describe the fed-batch culture in microbial 1,3-propanediol production, discusses important properties of the solution, and proposes a robust parameter identification model and solution. Numerical results indicate that both the system and algorithm perform well.
APPLIED MATHEMATICAL MODELLING
(2022)
Article
Computer Science, Software Engineering
Christopher Reinartz, Thomas T. Enevoldsen
Summary: pyTEP is an open-source simulation API for the Tennessee Eastman process in Python. It simplifies the setup of complex simulation scenarios and provides the option of interactive simulation. Through the pyTEP API, users can easily configure and operate simulations without needing to understand the underlying mechanics.
Article
Energy & Fuels
Mingfei Hu, Xinyi Hu, Zhenzhou Deng, Bing Tu
Summary: In this paper, a kernel extreme learning machine (KELM) based on an adaptive variation sparrow search algorithm (AVSSA) is proposed for fault detection and diagnosis in large industrial systems. The performance of the fault classifier is improved by optimizing the dataset and the network hyperparameters, and the effectiveness of the proposed method is verified using multidimensional diagnostic metrics in a chemical process.
Article
Computer Science, Interdisciplinary Applications
Ildar Lomov, Mark Lyubimov, Ilya Makarov, Leonid E. Zhukov
Summary: This paper investigates advanced approaches using deep learning methods in the field of fault detection in chemical processes, showing that with the recent advent of deep learning neural network methods and abundance of available sensor data, it became possible to develop advanced approaches to early fault detection and prediction that do not require feature engineering and provide more accurate and timely results. The proposed temporal CNN1D2D architecture achieved overall better performance on the dataset than any referenced method, and the use of Generative Adversarial Network GAN was suggested to extend and enrich data used in training.
JOURNAL OF INDUSTRIAL INFORMATION INTEGRATION
(2021)
Article
Automation & Control Systems
Carlos Andres Elorza Casas, Mahshad Valipour, Luis A. Ricardez Sandoval
Summary: This study applies the multi-scenario nonlinear model predictive control (MSc-NMPC) and multi-stage nonlinear model predictive control (MS-NMPC) to the Tennessee-Eastman (TE) challenge, using extended Kalman filter (EKF) and moving horizon estimation (MHE) as state estimators. The robust NMPC formulation ensures constraint violation prevention and close tracking of the process set-point under parameter uncertainty, even in cases where traditional NMPC leads to an unstable response. Unlike unconstrained state estimators like EKF, MHE considers process constraints in its formulation, overcoming the challenge of estimates falling outside the feasible region of the process. The additional computational time required for solving robust NMPC and MHE does not cause significant delays, demonstrating their applicability to complex industrial chemical processes.
CONTROL ENGINEERING PRACTICE
(2023)
Article
Food Science & Technology
Hao Chen, Jinjing Wang, Qi Li, Xin Xu, Chengtuo Niu, Feiyun Zheng, Chunfeng Liu
Summary: This study developed a mutant yeast strain with high RNA yield and optimized the fermentation process to increase RNA production. Genome sequencing revealed the metabolic mechanisms underlying high RNA production. This work has the potential to reduce the cost of RNA production and shorten the fermentation cycle.
Article
Engineering, Multidisciplinary
Yiman Li, Tian Peng, Wei Sun, Chunlei Ji, Yuhan Wang, Zihan Tao, Chu Zhang, Muhammad Shahzad Nazir
Summary: This study proposes a soft sensor model based on convolutional neural network - bidirectional long short-term memory (CNN-BiLSTM) and improved Harris hawk optimization (IHHO) algorithm for the Tennessee Eastman (TE) process. The model uses random forest (RF) algorithm to select auxiliary variables and incorporates circle mapping initialization and simulated annealing (SA) search strategy to improve the performance of the HHO algorithm. The CNN-BiLSTM model is constructed for soft sensor modeling of key variables in the TE process.
Article
Thermodynamics
Vijaya Chandgude, Teemu Valisalmi, Juha Linnekoski, Tom Granstrom, Bruna Pratto, Tero Eerikainen, German Jurgens, Sandip Bankar
Summary: This study successfully improved the product titer and yield of ABE fermentation production by combining reducing agents and a controlled feeding strategy, leading to efficient utilization of glucose and significant enhancement in solvent production.
ENERGY CONVERSION AND MANAGEMENT
(2021)
Article
Energy & Fuels
Rikus Styger, Kenneth R. Uren, George van Schoor
Summary: This paper proposes and evaluates five reduction techniques for graph-based fault detection and isolation (FDI) methods. Unlike existing graph reduction techniques, the proposed techniques can be easily applied to the type of attributed graph used by FDI methods. The results show that using these reduction techniques significantly reduces the size and complexity of the attributed graph while maintaining a similar level of FDI performance.
Article
Multidisciplinary Sciences
Avinash Maran Beena, Ajaya Kumar Pani
Summary: In the past decade, data-driven machine learning techniques have become popular in process monitoring, with Gaussian process regression as a promising but underexplored method. This research applied GPR to the Tennessee Eastman challenge problem and found its performance to be superior to other techniques in a comparative study.
ARABIAN JOURNAL FOR SCIENCE AND ENGINEERING
(2021)
Article
Engineering, Industrial
Chengyi Zhang, Jianbo Yu, Shijin Wang
Summary: The paper introduces a hybrid deep learning model (1-DCNN + SDAE) for extracting high level features from complex process signals, enhancing the performance of process fault detection and diagnosis. The model takes advantage of the characteristics of one-dimensional process signals and shows effectiveness in feature learning and fault diagnosis on multivariate manufacturing processes in experiments. This study provides guidance for the development of hybrid deep learning-based multivariate control models.
INTERNATIONAL JOURNAL OF PRODUCTION RESEARCH
(2021)
Article
Computer Science, Artificial Intelligence
Jianbo Yu, Chengyi Zhang, Shijin Wang
Summary: This study introduces a new deep neural network model MC1-DCNN, which learns fault features from high-dimensional process signals using wavelet transform, achieving remarkable feature extraction and fault diagnosis performance.
NEURAL COMPUTING & APPLICATIONS
(2021)
Article
Computer Science, Artificial Intelligence
Jianbo Yu, Guoliang Liu
Summary: This paper introduces a new deep neural network model KBSDAE, which enhances the understanding of representations learned by the deep network and improves the learning performance of stacked denoising auto-encoder. It achieves this by inserting knowledge into the deep network structure, offering a novel method for knowledge insertion and showing better feature learning performance compared to typical DNNs.
Article
Engineering, Industrial
Changhui Liu, Kun Chen, Sun Jin, Yuan Qu, Jianbo Yu, Binghai Zhou
Summary: This study proposes an automatic and integrated method for recognizing control chart patterns, which consists of three main modules: wavelet denoising, feature extraction, and classifier. Through comparison with other methods and application in a practical case, the high recognition accuracy of the integrated method is validated.
INTERNATIONAL JOURNAL OF PRODUCTION RESEARCH
(2022)
Article
Engineering, Industrial
Zhuang Ye, Jianbo Yu
Summary: The paper proposes a new deep neural network (DNN) model, adaptive kernel sparse network (AKSNet), to extract multi-scale fault features from vibration signals. AKSNet integrates key techniques such as adaptive kernel selection, channel attention, and spatial attention to effectively improve fault diagnosis performance of the classifier.
JOURNAL OF MANUFACTURING SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Xing Liu, Jianbo Yu, Lyujiangnan Ye
Summary: The paper introduces a new deep neural network RACAE for process monitoring, significantly improving monitoring performance in nonlinear processes. A new process monitoring model is developed with two statistics for fault detection, and the effectiveness of this method is evaluated through numerical cases and benchmark processes.
NEURAL COMPUTING & APPLICATIONS
(2021)
Article
Computer Science, Artificial Intelligence
Zhuang Ye, Jianbo Yu
Summary: Machine health assessment is crucial for prognostics and health management, and the proposed LSTMCAE demonstrates effectiveness in feature learning and generating health index using multivariate Gaussian distribution. Experimental results show the superiority of LSTMCAE in machine health assessment compared to other unsupervised learning methods.
APPLIED SOFT COMPUTING
(2021)
Article
Computer Science, Interdisciplinary Applications
Junjie Zhou, Jianbo Yu
Summary: A machine vision method is proposed for precise measurement of chisel edge wear in high-speed steel twist drills, aiming to improve measurement accuracy and reduce testing costs. Experimental results demonstrate that the system exhibits high response speed and detection accuracy, showing great potential for real-time monitoring of tool wear in industry.
COMPUTERS IN INDUSTRY
(2021)
Article
Automation & Control Systems
Chengyi Zhang, Jianbo Yu, Lyujiangnan Ye
Summary: This paper proposes a fault detection method for complex multivariate processes using sparsity and manifold regularized convolutional auto-encoders (SMRCAE) to extract features and evaluates its performance on an industrial benchmark.
CONTROL ENGINEERING PRACTICE
(2021)
Article
Automation & Control Systems
Jianbo Yu, Jiatong Liu
Summary: This article proposes a novel deep neural network, PCACAE, for wafer map defect recognition in semiconductor manufacturing process. Experimental results demonstrate that PCACAE outperforms other well-known convolutional neural networks in WMPR.
IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS
(2021)
Article
Computer Science, Artificial Intelligence
Zhuang Ye, Jianbo Yu
Summary: A novel deep neural network (AKRNet) is proposed for multi-scale feature learning from vibration signals, which performs better on gearbox fault diagnosis compared to other typical DNNs.
Article
Engineering, Electrical & Electronic
Mengqi Miao, Changhui Liu, Jianbo Yu
Summary: This article introduces a new DNN model, ADCAE, for feature extraction from vibration signals in an unsupervised way, which utilizes adaptive attention mechanism and multiscale convolution to enhance performance. Experimental results demonstrate that ADCAE performs well on gearbox vibration signals.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2021)
Article
Engineering, Electrical & Electronic
Jianbo Yu, Xing Liu, Lyujiangnan Ye
Summary: This article proposes a new deep neural network, CLSTM-AE, for feature learning from process signals in modern industrial processes. By embedding selective residual block in the deep network, it improves the training accuracy and performs feature selection effectively. The hybrid learning technique with convolutional LSTM and autoencoder provides a new way for feature learning and fault detection for complex industrial processes.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2021)
Article
Engineering, Electrical & Electronic
Yuanhang Sun, Jianbo Yu
Summary: A new method called SR-ASD is proposed for extracting fault features from bearing vibration signals. By retaining the sparsity of the signal and its difference, this method effectively eliminates noise interference and extracts impulsive features.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2021)
Article
Engineering, Electrical & Electronic
Xun Cheng, Jianbo Yu
Summary: The study introduces a new deep neural network model, DEA_RetinaNet, for steel surface defect detection, which utilizes methods like differential evolution search-based anchor optimization and channel attention mechanism to improve detection accuracy and achieve better recognition performance.
IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
(2021)