Industrial Control Research Articles

Improving Industrial Quality Control: A Transfer Learning Approach to Surface Defect Detection

To automate the quality control of painted surfaces of heating devices, an automatic defect detection and classification system was developed by combining deflectometry and bright light-based illumination on the image acquisition, deep learning models for the classification of non-defective (OK) and defective (NOK) surfaces that fused dual-modal information at the decision level, and an online network for information dispatching and visualization. Three decision-making algorithms were tested for implementation: a new model built and trained from scratch and transfer learning of pre-trained networks (ResNet-50 and Inception V3). The results revealed that the two illumination modes employed widened the type of defects that could be identified with this system, while maintaining its lower computational complexity by performing multi-modal fusion at the decision level. Furthermore, the pre-trained networks achieved higher accuracies on defect classification compared to the self-built network, with ResNet-50 displaying higher accuracy. The inspection system consistently obtained fast and accurate surface classifications because it imposed OK classification on models trained with images from both illumination modes. The obtained surface information was then successfully sent to a server to be forwarded to a graphical user interface for visualization. The developed system showed considerable robustness, demonstrating its potential as an efficient tool for industrial quality control.

Research on the Application of PID Control Algorithm and Fuzzy Control Theory in the Field of Temperature Control

In the field of industrial automation control systems, PID control algorithms are widely used in various control fields such as temperature, speed and position. Among them, fuzzy PID control can improve control accuracy and stability, can significantly improve product quality and production efficiency, in food processing, chemical engineering, pharmaceutical and textile industries have a wide range of applications. Therefore, this paper comprehensively discusses the theory of fuzzy PID control algorithm and its application in the field of temperature control. The advantages of fuzzy PID control in complex industrial environments are emphasized through a comparative analysis with other temperature control methods as well as traditional linear PID control. The research indicates that by optimizing the control algorithm and hardware design, the fuzzy PID control algorithm can reduce the system overshoot and steady state error, thus improving the control accuracy and response speed. At the same time, the introduction of multivariable PID controllers or the use of advanced multivariable control strategies can achieve coordinated control of multiple variables, thereby improving the performance of the entire system. In addition, the integration of advanced energy management system and data analysis technology can comprehensively monitor and optimize the management of energy consumption, further improving the energy efficiency of the system

Machine Learning-Based Multilevel Intrusion Detection Approach

In this paper, we propose a multilevel-based intrusion detection model. Firstly, we design an integrated shared feature technique, which filters the features to create a general dataset, retaining fewer but more significant features to enhance the detection accuracy of the model and reduce computational costs. The first stage employs OC-SVM to achieve the efficient classification of normal and abnormal traffic based on a general dataset. Additionally, the first stage is deployed close to the monitored system to enable low-latency prediction and privacy-preserving operations, thus enhancing flexibility and improving global classification performance. The second stage proposes a novel Edge Attention Network (EGAT) with a Multi-Head Dynamic Mechanism (MHD) framework, which introduces the graph attention mechanism and considers edge information as the only element, assigning greater weights to nodes and edges exhibiting high similarity, emphasizing their relationships and thereby improving the model’s accuracy and expressiveness. The MHDEGAT model facilitates additional weight learning by integrating the multi-head attention mechanism with edge features, while the weighted aggregation process enhances the data utilization across different network traffic. Finally, the model is trained and tested using the method of on-network data from a gas industrial control system, with an accuracy of 96.99%, a precision of 97.11%, a recall of 96.99%, and an F1 score of 96.93%, all of which outperform the comparison method.

A System Control and Waveform Acquisition Framework for the Kicker System at SHINE

A lumped kicker magnet system is being developed as part of the electron beam switchyard of the Shanghai High Repetition Rate X-Ray Free-Electron Laser (XFEL) and Extreme Light (SHINE) facility at Shanghai Advanced Research Institute (SARI), Chinese Academy of Sciences (CAS). SHINE is a linac-based facility with an energy of 8 GeV and a repetition rate of 1 MHz for multiple users. The kicker system is a critical element of the deflection system and determines the stability performance of the beamlines. To ensure autonomous and reliable operation in a harsh environment, a dedicated control system—consisting of a custom-designed programmable logic controller (PLC) and waveform acquisition system—has been developed based on the Experimental Physics and Industrial Control System (EPICS). The PLCsystem enables status acquisition, control implantations, and alarm handling, allowing for automated operation. The waveform acquisition system provides an intuitive demonstration and enables the real-time diagnosis of kicker waveforms. The results demonstrate a half quasi-sinuous waveform with a peak current of 11.5 A and a full width at half maximum of 218 ns at a repetition rate of 10 kHz. The detailed design, overall concept, achieved performance, and initial test results of two modular controls are provided for further demonstration.

DCFE-YOLO: A novel fabric defect detection method.

Accurate detection of fabric defects is crucial for quality control in the textile industry. However, the task of fabric defect detection remains highly challenging due to the complex textures and diverse defect patterns. To address the issues of inaccurate localization and false positives caused by complex textures and varying defect sizes, this paper proposes an improved YOLOv8-based fabric defect detection method. First, Dynamic Snake Convolution is introduced into the backbone network to enhance sensitivity to elongated and subtle defects, improving the extraction of edge and texture details. Second, a Channel Priority Convolutional Attention mechanism is incorporated after the Spatial Pyramid Pooling layer to enable more precise defect localization by leveraging multi-scale structures and channel priors. Finally, the feature fusion network integrates Partial Convolution and Efficient Multi-scale Attention, optimizing the fusion of information across different feature levels and spatial scales, which enhances the richness and accuracy of feature representations while reducing computational complexity. Experimental results demonstrate a significant improvement in detection performance. Specifically, mAP@0.5 increased by 2.9%, precision improved by 3.5%, and mAP@0.5:0.95 rose by 2.3%, highlighting the model's superior capability in detecting complex defects. The project is available at https://github.com/lilian998/fabric.

TDLAS based gas absorption spectrum detection system for the college physics experimental teaching

Abstract Laser absorption spectroscopy, grounded in the Lambert–Beer law, stands out as a prevalent non-destructive testing method within optical gas sensing technology. It finds extensive applications across atmospheric environment monitoring, industrial process control, combustion diagnosis, and biomedicine. Within this study, a pioneering gas absorption spectrum detection system is introduced, leveraging LabVIEW programming and tunable diode laser absorption spectroscopy. This system is tailored for college physics education, enabling undergraduates to grasp the fundamental workings of optical gas sensing technology. Methane (CH4) is selected as the target gas molecule for its relevance in environmental dynamics and the push towards carbon neutrality. Investigative experiments delve into diverse spectral line parameters impacting the gas absorption signal profile, as prescribed by the theoretical underpinnings of laser spectroscopy. The teaching methodology proposed herein blends theoretical concepts with hands-on experimentation, fostering a deep student interest in the exploration of gas sensing technologies within the college physics domain.

Evaluating the internal control strategy in the AI industry: using the COSO framework

ABSTRACT This study explores the application of the COSO Internal Control Integrated Framework (COSO ICIF) within Taiwan’s artificial intelligence (AI) industry, addressing challenges such as rapid technological advancements, complex risk landscapes, and dynamic global markets. By employing a hybrid methodology that integrates Decision-Making Trial and Evaluation Laboratory (DEMATEL) with the DEMATEL-based Analytic Network Process (DANP), the research constructs an Influence Network Relation Map (INRM) to analyse interdependencies and prioritize internal control improvements. The findings emphasize the foundational role of the control environment in shaping other internal control components, highlighting the critical importance of monitoring activities and internal communication in mitigating risks and enhancing operational efficiency. The research also identifies emerging challenges, including stricter data privacy regulations, demands for transparent AI algorithms, and heightened global competition, stressing the necessity for adaptive internal controls that ensure compliance while supporting strategic agility. By extending the applicability of COSO ICIF to technology-driven industries, this research contributes valuable insights for both scholars and practitioners. The proposed hybrid model offers a robust framework for addressing the complexities of internal control in the AI industry, providing a systematic approach to enhance organizational resilience and performance in an evolving technological landscape.

Deep Learning Algorithm-Based Appearance Detection of Cigarette Strip Packs

Each local feature in the appearance image of cigarette packs is a key element to reflect the corresponding brand information. If only a single convolutional neural network is used, the context information of sequence data may be lost, resulting in an insufficient grasp of the overall information. In order to realize the deep-level feature extraction of the appearance of cigarette packs and realize the appearance detection of cigarette packs with higher accuracy and speed, a new method for the appearance detection of cigarette packs was proposed by combining the convolutional neural network and the cyclic neural network methods in the deep learning algorithm. The image acquisition card is used to collect the appearance images of cigarette packs, and contrast enhancement and rotation correction are performed on the collected images to effectively improve their quality and provide a good guarantee for subsequent feature extraction and detection. The preprocessed cigarette package image is input into the convolutional neural network to realize the deep-level feature extraction of the cigarette package appearance image. The cigarette package appearance features’ output from the convolutional neural network is input into the short-term and long-term memory unit, as well as the gate-controlled cyclic unit, of the corresponding recurrent neural network, in order to process time sequence information based on the efficient extraction of details from the image, retain the sequence and context information of the input data, and ultimately achieve accurate detection of the cigarette package appearance. Through experimental analysis, this method can effectively identify the appearance defects of cigarette packs, mark them immediately, and present them in an intuitive way, so that staff can quickly locate the problem and take corresponding measures. The method can detect appearance defects larger than [Formula: see text] with high accuracy. For various appearance defects, the detection rate can be guaranteed to be over 98%, providing strong support for quality control and product upgrading in the tobacco industry.

ANÁLISE MATEMÁTICA E SIMULAÇÃO DO SISTEMA DE AMORTECIMENTO VEICULAR COM MATLAB

Abstract: In modern control theory, the concept of state variables plays a crucial role in the modeling and analysis of dynamic systems. These variables provide a comprehensive description of the system’s future behavior over time. The organization into a state vector enables a compact and complete representation of the system’s internal state. The state-space function describes the temporal evolution of these variables in response to system inputs, offering an efficient mathematical approach for the design and analysis of control systems. Applications range from industrial process control to autonomous vehicle control and robotics. By understanding and applying these concepts, control engineers develop effective solutions for challenges in complex systems, ensuring performance, stability, and efficiency in dynamic environments. Keywords: state variables; dynamic systems; state-space; control; stability.

An integrated cyber-physical framework for worst-case attacks in industrial control systems

Industrial Control Systems (ICSs) are widely used in critical infrastructures that face various cyberattacks causing physical damage. With the increasing integration of the ICSs and information technology (IT), ensuring the security of ICSs is of paramount importance. In an ICS, cyberattacks exploit vulnerabilities to compromise sensors and controllers, aiming to cause physical damage. Maliciously accessing different components poses varying risks, highlighting the importance of identifying worst-case cyberattacks. This aids in designing effective detection schemes and mitigation strategies. This article proposes an optimization framework that integrates cyber and physical systems of ICSs to identify the worst-case attack. The framework models cyberattacks with varying resources by (i) maximizing physical impact in terms of time to failure of the physical system, (ii) quickly accessing the sensors and controllers in the cyber system while exploiting limited vulnerabilities, (iii) avoiding detection in the physical system, and (iv) complying with the cyber and physical restrictions. These objectives enable us to model the interactions between the cyber and physical systems jointly and study the critical cyberattacks that cause the highest impact on the physical system under certain resource constraints. Our framework serves as a tool to understand the critical vulnerabilities of an ICS by holistically considering the interactions between cyber and physical systems. It also assesses the robustness of existing detection schemes and mitigation strategies by generating worst-case attack strategies. We illustrate and verify the effectiveness of our proposed method in both a numerical and a case study. The results show that a worst-case strategic attacker causes almost 19% further acceleration in the time to failure of the physical system while remaining undetected compared with a random attacker. Moreover, by considering the interactions and interdependencies between cyber and physical systems, our framework identifies worst-case attacks that reveal critical vulnerabilities and deficiencies in existing detection schemes, which can remain hidden when the two layers are analyzed separately.

Enhancing DNP3 Security Using CNN Deep Learning Techniques

Industrial Automation and Control Systems (IACS) are necessary for enabling secure information exchange between smart devices; ensuring security in Industrial Control Systems (ICS) is of importance due to the presence of these devices at distant locations and their control over vital plant activities. Intelligent devices and hosts use protocols such as Modbus, DNP3, IEC 60870, IEC 61850, and others. This paper focuses on the analysis and development of techniques for detecting of network traffic within the industrial environment, more specifically anomalies in the application ZZZAlayer in the to the protocol called Distribution Network Protocol (DNP3) is an open-source protocol used in Supervisory Control and Data Acquisition (SCADA) systems and widely recognized as the standard for the water, sewage, and oil and gas industries. it is used in the realm of industrial automation; they are critical facilities for the population and must be secured against any security breaches. One of the main objectives of cyber attackers is related with these systems. In This paper presents an architecture that, classification system by Deep Learning algorithm with (CNN). The proposed model was evaluated using standard Intrusion Detection Dataset for DNP3, with 7326) and 86field. The CNN algorithm obtained the best results accuracy

Differences in Salmonella Serovars Response to Lactic Acid and Peracetic Acid Treatment Applied to Pork.

Pathogen control in the meat industry relies on the effectiveness of postharvest interventions in reducing microbial populations. This study investigated differences in the survival of Salmonella serovars when exposed to organic acids used as antimicrobials on raw pork meat. Seven serovars were included in this study (S. Newport, S. Kentucky, S. Typhimurium, S. Dublin, S. Heidelberg, S. Infantis, and S. Enteritidis). Multistrain serovar cocktails were prepared and tested against lactic acid (LA) and peracetic acid PAA at two concentrations, LA 2 and 4% and PAA 200 and 400ppm. Pork samples were assigned to each serovar, inoculated with 6.0 Log CFU/cm2Salmonella (one serovar at a time), and treated with the corresponding antimicrobials. A two-way analysis of variance was conducted to examine the effects of serovar and antimicrobial concentrations on Salmonella survival. A significant main effect of serovar was identified, indicating that Salmonella concentration and reduction rate were significantly affected by serovar. Similarly, a significant main effect of antimicrobials was observed, suggesting that the treatment types impacted Salmonella concentration and reduction rate. However, the interaction effect between serovar and antimicrobial was not significant. Posthoc comparisons indicate that PAA 400ppm is more effective at reducing Salmonella concentrations and that S. Dublin may be more susceptible than S. Newport to antimicrobial sprays. Additionally, under PAA exposure, only S. Dublin, S. Kentucky, and S. Heidelberg showed statistically significant differences (P<0.05) compared with the control, indicating that these three serovars are more susceptible to PAA treatments than the rest. The behavior of different Salmonella serovars under stress conditions can give us an insight into how these pathogens survive processing.

ColorNet: An AI-based framework for pork freshness detection using a colorimetric sensor array.

Pork freshness is crucial for flavour, nutrition and consumer health. The current colorimetric sensor array (CSA) detection systems face challenges related to high sensor development costs, low recognition accuracy and limitations in the platform use. Herein, we developed a CSA and ColorNet framework to detect pork freshness. The 53-point CSA was designed by selecting sensitised pH indicators and aldehyde/ketone indicators. To optimize the sensor, the Euclidean distance method was used to identify 24 array points with dyes that exhibited more sensitive responses. The ColorNet captured the color information of pork freshness, allowing real-time detection with a 99.5% accuracy. For practical deployment and mobile applications, a refined 12-point CSA was developed using gradient activation mapping, maintaining a 99% recognition rate, which is comparable with the 24-point CSA. The proposed CSA and model ensure consumer health and safety, providing strong technical support for quality monitoring and control in the pork industry.

Industrial robot control system research based on the direction of programmable logic controllers

Industrial robot control system research based on the direction of programmable logic controllers

Machine Learning-Based Detection of Anomalies, Intrusions and Threats in Industrial Control Systems

Machine Learning-Based Detection of Anomalies, Intrusions and Threats in Industrial Control Systems

Enhancing Anomaly Detection in Industrial Control Systems through Supervised Learning and Explainable Artificial Intelligence

Enhancing Anomaly Detection in Industrial Control Systems through Supervised Learning and Explainable Artificial Intelligence

A Framework for Runtime Safety of Industrial Control Systems Through Runtime Verification

A Framework for Runtime Safety of Industrial Control Systems Through Runtime Verification

Artificial Intelligence and Machine Learning Driven Adaptive Control Applications

This article explores the transformative role of Artificial Intelligence (AI) and Machine Learning (ML) in enhancing adaptive control systems across various industries. Adaptive control systems adjust their parameters in real-time to maintain optimal performance in dynamic environments, and the integration of AI and ML significantly improves their effectiveness. AI enables these systems to learn, recognize patterns, and make autonomous decisions, while ML algorithms allow adaptive controllers to generalize from data, enhancing performance through techniques such as supervised, reinforcement, and unsupervised learning. Applications in autonomous vehicles, industrial process control, and aerospace illustrate the impact of AI and ML on adaptive control. Despite challenges such as computational demands, safety concerns, and regulatory hurdles, the future of adaptive control will be shaped by advances in AI and ML technologies, driving innovations in automation, precision, and reliability across diverse sectors.

Getting democracy wrong

Recent developments in large language models and computer automated systems more generally (colloquially called ‘artificial intelligence’) have given rise to concerns about potential social risks of AI. Of the numerous industry-driven principles put forth over the past decade to address these concerns, the Future of Life Institute’s Asilomar AI principles are particularly noteworthy given the large number of wealthy and powerful signatories. This paper highlights the need for critical examination of the Asilomar AI Principles. The Asilomar model, first developed for biotechnology, is frequently cited as a successful policy approach for promoting expert consensus and containing public controversy. Situating Asilomar AI principles in the context of a broader history of Asilomar approaches illuminates the limitations of scientific and industry self-regulation. The Asilomar AI process shapes AI’s publicity in three interconnected ways: as an agenda-setting manoeuvre to promote longtermist beliefs; as an approach to policy making that restricts public engagement; and as a mechanism to enhance industry control of AI governance.

Microstructural characterization of a wheat-based food material using image analysis and pore network modeling during baking.

Microstructural properties of wheat-based food materials change during baking. These alterations affect the final product's mechanical properties, physical attributes, and consumer satisfaction. Image processing and pore network modeling were used to analyze the variations in a cookie's microstructural properties during baking. Cookies were baked in an oven set at 184±2°C for 10-60min and then freeze-dried. Over 1500 two-dimensional images were collected for each sample using X-ray micro-computed tomography. Image processing and pore network modeling were utilized to obtain porosity, pore size distribution, tortuosity, pore coordination number, permeability, and other properties. A gas pycnometer was used to validate the obtained porosity values and calculate the solid density. Three stages of expansion, shrinkage, and subsequent expansion were observed during baking. The total porosity ranged between 0.145 and 0.373. The largest pores were observed in the 20-min baked sample, which had a moisture content of 8.53g/100g solids. Cookies baked longer had more interconnected pores than the 10-min sample, enabling them to conduct fluid at a higher rate. The absolute permeability values were between 1.75 × 10-11 m2 and 4.68 × 10-11 m2. The results suggested that porosity and pore coordination number had a similar trend as cookie permeability. Both heat and mass transfer during baking are expected to influence the development of the porous structure. PRACTICAL APPLICATION: Understanding the microstructural properties of wheat-based food materials is expected to help the baking industry control processing conditions, final product quality, and visual texture by adjusting the baking parameters.