Industrial Process Control Research Articles

Sensor Systems for Extremely Harsh Environments

Sensors are key elements for capturing environmental properties and are today indispensable in the industry for monitoring and control of industrial processes. Many applications are demanding for highly integrated intelligent sensors to meet the requirements on safety, clean and energy efficient operation or to gain process information in the context of industry 4.0. While in many everyday objects highly integrated sensor systems are already state of the art, the situation in an industrial environment is clearly different. Frequently the use of sensor systems is impossible, due to the fact that the extreme ambient conditions of industrial processes like high operating temperatures or strong mechanical loads do not allow a reliable operation of sensitive electronic components. Eight Fraunhofer Institutes have bundled their competencies and have run the Fraunhofer Lighthouse Project ‘eHarsh’ to overcome this situation. The project goal was to realize sensor systems for extremely harsh environments, whereby sensor systems are not only pure sensor elements, rather containing one or multiple sensor elements and integrated readout electronics. Various technologies which are necessary for the realization of such sensor systems have been identified, developed and finally bundled in a technology platform. These technologies are e. g. MEMS and ceramic based sensors, SOI-CMOS based integrated electronics, board assembly and laser based joining technologies. All these developments have been accompanied by comprehensive tests, material characterization and reliability simulations. Based on the platform a pressure sensor for turbine applications has been realized to prove the performance of the eHarsh technology platform.

Correlation of Steam Velocity and Pipe Diameter with Heat Transfer Performance on 120° Half-Pipe Jacket

Double jacket mixing tanks offer temperature control and versatility for industrial processes where temperature-sensitive materials need to be mixed, stored, or processed. The selection of the appropriate jacket type in a double-jacket mixing tank is based on the structural strength and the optimal heat transfer performance. The type of jackets can be conventional, dimple, and half-pipe. The conventional jacket is easier to manufacture, but it is not resistant to high temperature and pressure. The dimple jacket has good heat transfer, but it is susceptible to damage. The half-pipe jacket has good structural strength, but its heat transfer is affected by the central angle and pipe diameter. 120° central angle has greater heat transfer and pressure drop than 180° central angle. In order to determine the effects of steam velocity on heat transfer performance, heating time, and pressure drop of the 120° half-pipe jacket, research occurred on pipes with 2, 2.5, and 3 inches of diameter. The calculating method of heat transfer in agitated jacketed vessels is applied to visualize the relationships. The effects of steam velocity on heat transfer, heating duration, and pressure drop respectively are polynomials of order 2, power, and polynomials of order 2 with an average R2 close to 1. The greater R2, the better the relationship between two variables, according to the equation. The 120° half-pipe jacket performance will be highly effective, with 1774 W, 2.1 minutes heating duration, 8.94 kPa pressure drop if the steam velocity is 10.50 m/s with 2.5 inches pipe diameter.

A novel fractional-order dead-time compensating controller for the wireless networks

Wireless technology is becoming increasingly critical in industrial environments in recent years, and the popular wireless standards are WirelessHART, ZigBee, WLAN and ISA100.11a, commonly used in closed-loop systems. However, wireless networks in closed-loop control experience packet loss or drops, system delay and data threats, leading to process instability and catastrophic system failure. To prevent such issues, it is necessary to implement dead-time compensation control. Traditional techniques like model predictive and predictive PI controllers are frequently employed. However, these methods’ performance is sluggish in wireless networks, with processes having long dead times and set-point variations, potentially affecting network and process performance. Therefore, this paper proposes a fractional calculus-based predictive PI compensator for wired and wireless networks in the process control industries. The proposed technique has been simulated and evaluated on industrial process models, including pressure, flow, and temperature, where measurement and control are carried out wirelessly. The wireless network’s performance has been evaluated based on packet loss, reduced throughput, and increased system latency. The proposed compensator outperformed traditional methods, demonstrating superior set-point tracking, disturbance rejection, and delay compensation characteristics in the performance evaluations of the first, second, and third-order systems. Overall, the findings indicate that the proposed compensator enhances wireless networks’ performance in the process control industry and improves system stability and reliability by reducing almost half of the overshoot and settling an average of 8.3927% faster than the conventional techniques in most of the systems.

A New Texture Spectrum Based on Parallel Encoded Texture Unit and Its Application on Image Classification: A Potential Prospect for Vision Sensing.

In industrial applications based on texture classification, efficient and fast classifiers are extremely useful for quality control of industrial processes. The classifier of texture images has to satisfy two requirements: It must be efficient and fast. In this work, a texture unit is coded in parallel, and using observation windows larger than 3×3, a new texture spectrum called Texture Spectrum based on the Parallel Encoded Texture Unit (TS_PETU) is proposed, calculated, and used as a characteristic vector in a multi-class classifier, and then two image databases are classified. The first database contains images from the company Interceramic®® and the images were acquired under controlled conditions, and the second database contains tree stems and the images were acquired in natural environments. Based on our experimental results, the TS_PETU satisfied both requirements (efficiency and speed), was developed for binary images, and had high efficiency, and its compute time could be reduced by applying parallel coding concepts. The classification efficiency increased by using larger observational windows, and this one was selected based on the window size. Since the TS_PETU had high efficiency for Interceramic®® tile classification, we consider that the proposed technique has significant industrial applications.

Tunable Diode Lasers for Analytics and Diagnostics

Continuous-wave diode lasers (DLs) with tunable emission wavelengths have become extensively used in various fields of analytical spectroscopy and diagnostics. Working in the near and mid-IR spectral region, tunable diode lasers are particularly effective in detecting simple molecules, making them invaluable for environmental monitoring, industrial process control, and diagnostics of subsonic and supersonic gas flows. However, the lack of commercial diode lasers operating in the spectral region shorter than 400 nm has restricted their applicability to elemental analysis, as many resonance lines of free atoms of elements lie in the region 250–400 nm. This review aims to highlight various applications of continuous diode lasers, which are lesser-known to analytical chemists. We briefly overview their main characteristics and discuss their advantages, enabling their successful implementation in traditional analytical spectroscopy tasks, as well as for diagnosing parameters of remote gas objects, including combustion processes in mixing gas flows.

Tunable Diode Lasers for Analytics and Diagnostics

Continuous-wave diode lasers (DLs) with tunable emission wavelengths have become extensively used in various fields of analytical spectroscopy and diagnostics. Working in the near and mid-IR spectral region, tunable diode lasers are particularly effective in detecting simple molecules, making them invaluable for environmental monitoring, industrial process control, and diagnostics of subsonic and supersonic gas flows. However, the lack of commercial diode lasers operating in the spectral region shorter than 400 nm has restricted their applicability to elemental analysis, as many resonance lines of free atoms of elements lie in the region 250–400 nm. This review aims to highlight various applications of continuous diode lasers, which are lesser-known to analytical chemists. We briefly overview their main characteristics and discuss their advantages, enabling their successful implementation in traditional analytical spectroscopy tasks, as well as for diagnosing parameters of remote gas objects, including combustion processes in mixing gas flows.

Dual Antipodal Vivaldi Antenna-Based Moisture Sensor for Industrial Process Control

Dual Antipodal Vivaldi Antenna-Based Moisture Sensor for Industrial Process Control

Wine Production through Virtual Environments with a Focus on the Teaching–Learning Process

This paper focuses on the application of the hardware-in-the-loop (HIL) technique in the winemaking process. The HIL technique provides an effective methodology to test and verify the automatic control of industrial processes in 3D laboratory environments. Two parts are considered: (i) software, which consists of the virtualization of the wine process in order to generate a realistic work environment that allows the student to manipulate the system while visualizing the changes in the process; and (ii) hardware, through which the process control is implemented in ladder language in a PLC S7 1200 AC/DC/RLY (programmable logic controller). Bidirectional Ethernet TCP/IP communication is established, achieving a client–server architecture. This article highlights the main advantages of the HIL technique, such as its ability to simulate complex and extreme scenarios that would be difficult or expensive to recreate in a real environment. In addition, real-time testing of the hardware and software to implement the control system is performed, allowing for fast and accurate responses. Finally, a usability table is obtained that demonstrates the benefits of performing industrial process control work in virtual work environments, focusing the development on meaningful learning processes for engineering students.

Digital twin-enabled process control in the food industry: proposal of a framework based on two case studies

Nowadays many processes in the food industry are monitored in an automatic way, with the purpose of minimising the need for workforce and of ensuring the proper control of the quality and safety of the foodstuff. All the sensors share data with a centralised management unit, where often a Manufacturing Execution System collects and evaluates them. As reported in recent research, however, a further step that can be undertaken, exploiting Industry 4.0 enabling technologies, is the implementation of digital twin approaches, with the additional aim to prevent possible issues during production. In line with these considerations, this work aims at showing two different digital twin models intended for improving the control of as many real food systems. Liquid and powder fluids are taken as examples for highlighting the differences in the optimization of the two food processes, as well as for fully exploring the potential of the digital twin approach. Finally, based on the real data taken from two pilot plants, a framework for the selection of the best digital twin tool in the food sector is delineated.

Approach to multispectral thermometry with Planck formula and hybrid metaheuristic optimization algorithm.

Accurate temperature measurement has significant implications for product quality, industrial process control, and scientific research. As a non-contact temperature measurement method with broad application prospects, multispectral thermometry still poses significant challenges in data processing. Currently, most multispectral thermometry methods use the Wien approximation equation to construct the objective function. However, the use of the Wien approximation equation is conditional and generally applicable only to low temperatures or short wavelengths. In this paper, what we believe is a new data processing model of multispectral thermometry is established based on the Planck formula; Additionally, a feasible region constraint method is proposed to constrain the emissivity range; By utilizing a hybrid metaheuristic optimization algorithm based on differential evolution (DE) and multi-population genetic (MPG) algorithms, the simulation results of six different models and experimental results of silicon carbide demonstrate that the proposed algorithm achieves an average relative error in temperature measurement within 0.42% and a random relative error within 0.79%. The average computation time for each temperature inversion is approximately 0.26 seconds. The accuracy and efficiency of the algorithm ensure that it can be applied to real-time temperature measurement in industrial field.

Optimization of Third Order System with a Novel Fuzzy-based Fractional PID Controller

The conventional Integer Order-based Proportional Integral Derivative (IO-PID) control has been widely employed in process control and manufacturing industries. However, upon thorough review, it is evident that traditional IO-PID control falls short when applied to non-linear and higher-order systems. This paper addresses the challenge of controlling higher-order systems prevalent in numerous industrial plants. The primary goal of controllers in intricate systems is to minimize settling time, overshoot, and related parameters for higher-order systems. To tackle these issues, a novel Fuzzy fractional order-based PID (FFOPID) controller is introduced through this work. This controller aims to provide solutions for the significant challenges encountered in control systems. By subjecting both the traditional controller and the novel controller to the same unit step input, simulations are carried out using MATLAB. The comparative analysis of the output responses highlights the superior performance of the designed novel controller works over the traditional IO-PID controllers. There is a reduction of overshoot by 15% and settling time by more than one second by the FFOPID controller compared to the IO-PID controller.

Formula omitted]output feedback fault-tolerant control of industrial processes based on zero-sum games and off-policy Q-learning

Traditional model-based control methods are often not applicable in industrial processes given the typical situation that model parameters are unknown, coefficient matrices are difficult to obtain, and system states are unpredictable. Accordingly, an output feedback fault-tolerant control method based on zero-sum game theory and off-policy Q-learning is presented in this study, with the aim of achieving smooth operation and good tracking performance for industrial processes that often contain sensor faults and disturbances. The specific steps are as follows. First, a system tracking error is introduced into the system to realize a novel extended model. Second, by establishing a performance index function and combining it with minimax theory, the fault-tolerant tracking control problem is converted into a zero-sum game problem. The Bellman and Riccati equations can be established after analyzing the relationship between the performance index and value functions. Then, the Q-function is introduced, and an off-policy Q-learning algorithm is combined with the Kronecker product without knowledge of system model parameters to design an optimal controller unbiased to detection noise. Finally, the effectiveness of the algorithm is verified by considering the injection molding process as an example. The experimental results validate that the designed controller demonstrates good control and extends the range of tolerable faults while maintaining good tracking performance.

Improved Smooth Watermarking Methods for Detecting Replay Attacks in Process Control Systems

Network attack detection is a crucial issue in cyber–physical systems. One classical type of network attack is the replay attack. Although the watermarking scheme has proven to be effective in detecting replay attacks, it may confront numerous challenges when applied to process control systems. The watermarking signal, which consists of independent and identically distributed Gaussian noise sequences, responds slowly to industrial process control systems. To address this issue, two novel smooth watermarking methods are proposed. The smooth watermarking signals exhibit low-frequency characteristics, enabling them to effectively handle large inertia processes. The proposed schemes are validated through two simulated examples. These examples indicate that the proposed methods can significantly enhance detection efficiency while minimizing additional control performance loss.

Optimal operational self-learning control for multi-time scale industrial processes with signal compensations

Since the multi-time scales and strong nonlinearity are objectively existed in the practical industrial processes, it is extremely challenging to design control policy for optimizing the industrial operation. Therefore, this paper focuses on the multi-time scale optimal operational control (OOC) problem with unmodeled dynamics. Under the framework of reinforcement learning (RL) technology, a self-learning model-free method by integrating zero-sum game and singular perturbation (SP) theories is proposed in this paper, such that composite controllers with signal compensation can be found under which the suboptimal operation of the whole process can be achieved using only measured data. The remarkable highlight in this paper is the novel design method of the composite controllers with signal compensations for unknown nonlinear systems with multi-rate operations. Finally, the effectiveness of the proposed method is verified by a practical example and a numerical example.

Modeling Battery Formation: Boosted SEI Growth, Multi-Species Reactions, and Irreversible Expansion

This work proposes a semi-empirical model for the SEI growth process during the early stages of lithium-ion battery formation cycling and aging. By combining a full-cell model which tracks half-cell equilibrium potentials, a zero-dimensional model of SEI growth kinetics, and a semi-empirical description of cell thickness expansion, the resulting model replicated experimental trends measured on a 2.5 Ah pouch cell, including the calculated first-cycle efficiency, measured cell thickness changes, and electrolyte reduction peaks during the first charge dQ/dV signal. This work also introduces an SEI growth boosting formalism that enables a unified description of SEI growth during both cycling and aging. This feature can enable future applications for modeling path-dependent aging over a cell’s life. The model further provides a homogenized representation of multiple SEI reactions enabling the study of both solvent and additive consumption during formation. This work bridges the gap between electrochemical descriptions of SEI growth and applications toward improving industrial battery manufacturing process control where battery formation is an essential but time-consuming final step. We envision that the formation model can be used to predict the impact of formation protocols and electrolyte systems on SEI passivation and resulting battery lifetime.

Attention-Guided Multitask Learning for Surface Defect Identification

Surface defect identification is an essential task in the industrial quality control process, in which visual checks are conducted on a manufactured product to ensure that it meets quality standards. Convolutional Neural Network (CNN) based surface defect identification method has proven to outperform traditional image processing techniques. However, the real-world surface defect datasets are limited in size due to the expensive data generation process and the rare occurrence of defects. To address this issue, this paper presents a method for exploiting auxiliary information beyond the primary labels to improve the generalization ability of surface defect identification tasks. Considering the correlation between pixel level segmentation masks, object level bounding boxes and global image level classification labels, we argue that jointly learning features of the related tasks can improve the performance of surface defect identification tasks. This paper proposes a framework named Defect-Aux-Net, based on multi-task learning with attention mechanisms that exploit the rich additional information from related tasks with the goal of simultaneously improving robustness and accuracy of the CNN based surface defect identification. We conducted a series of experiments with the proposed framework. The experimental results showed that the proposed method can significantly improve the performance of state-of-the-art models while achieving an overall accuracy of 97.1%, Dice score of 0.926 and mAP of 0.762 on defect classification, segmentation and detection tasks.

Application of engineering thinking for risk assessment in a Canadian elementary school

This paper outlines the application of a facility-level research methodology for hazard assessment and adoption of controls through a case study performed in an elementary school located in Alberta, Canada. Two classrooms with different educational activities planned were monitored for two weeks using multiple sensors to establish the impact of the activity on the accumulation of carbon dioxide in the space. Building operators were particularly concerned with the relative risks of choral singing as compared to traditional classroom activities during the COVID-19 pandemic, and little existing research supported decision-making in that area. The data collected in this study challenges the basis for public health controls in schools and demonstrates the feasibility of data collection and reporting outside the under-resourced public health departments. The classroom activities in this facility had little measurable impact on carbon dioxide levels. A challenge for public health officials during COVID-19 was in bridging a theoretical understanding of aerosol hazard transmission with the operationalization of that theory into effective risk controls for all facilities operating within a jurisdiction. The utility and effectiveness of risk controls should be re-evaluated routinely and supported by measurements and data as is feasible to collect, and this study demonstrates how researchers can bridge the gap between policy makers and regulated entities. Engineering thinking, and principles of industrial risk management and process control can be readily applied in any commercial building facility, and this study demonstrates some of the opportunities available to building operators and policy makers with the mass proliferation of affordable instruments with high utility for assessing building conditions.

Optimizing Three-Tank Liquid Level Control: Insights from Prairie Dog Optimization

The management of chemical process liquid levels poses a significant challenge in industrial process control, affecting the efficiency and stability of various sectors such as food processing, nuclear power generation, and pharmaceutical industries. While Proportional-Integral-Derivative (PID) control is a widely-used technique for maintaining liquid levels in tanks, its efficacy in optimizing complex and nonlinear systems has limitations. To overcome this, researchers are exploring the potential of metaheuristic algorithms, which offer robust optimization capabilities. This study introduces a novel approach to liquid level control using the Prairie Dog Optimization (PDO) algorithm, a metaheuristic algorithm inspired by prairie dog behavior. The primary objective is to design and implement a PID-controlled three-tank liquid level system that leverages PDO to regulate liquid levels effectively, ensuring enhanced stability and performance. The performance of the proposed system is evaluated using the ZLG criterion, a time domain metric-based objective function that quantifies the system's efficiency in maintaining desired liquid levels. Several analysis techniques are employed to understand the behavior of the system. Convergence curve analysis assesses the PDO-controlled system's convergence characteristics, providing insights into its efficiency and stability. Statistical analysis determines the algorithm's reliability and robustness across multiple runs. Stability analysis from both time and frequency response perspectives further validates the system's performance. A comprehensive comparison study with state-of-the-art metaheuristic algorithms, including AOA-HHO, CMA-ES, PSO, and ALC-PSODE, is conducted to benchmark the performance of PDO. The results highlight PDO's superior convergence, stability, and optimization capabilities, establishing its efficacy in real-world industrial applications. The research findings underscore the potential of PDO in PID control applications for three-tank liquid level systems. By outperforming benchmark algorithms, PDO demonstrates its value in industrial control scenarios, contributing to the advancement of metaheuristic-based control techniques and process optimization. This study opens avenues for engineers and practitioners to harness advanced control solutions, thereby enhancing industrial processes and automation.

In-line Raman spectroscopy for characterization of an industrial poultry raw material stream

In this work, we evaluated the feasibility of Raman spectroscopy as an in-line raw material characterization tool for industrial process control of the hydrolysis of poultry rest raw material. We established calibrations (N = 59) for fat, protein, ash (proxy for bone) and hydroxyproline (proxy for collagen) in ground poultry rest raw material. Calibrations were established in the laboratory using poultry samples with high compositional variation. Samples were measured using a wide area illumination Raman probe at varying working distance (6 cm, 9 cm, 12 cm) and probe tilt angle (0°, 30°) to mimic expected in-line variations in the measurement situation. These moderate variations did not significantly affect performance for any analytes. The obtained calibrations were tested in-line with continuous measurements of the ground poultry by-product stream at a commercial hydrolysis facility over the course of two days. Measurements were acquired under demanding conditions, e.g. large variations in working distance. Reasonable estimates of compositional trends were obtained. Validation samples (N = 19) were also reasonably well predicted, with RMSEPcorr = [0.14, 1.37, 2.36, 1.51]% for hydroxyproline, protein, fat and ash, respectively. However, there were indications that further calibration development and robustification of pre-processing would be advantageous, particularly with respect to hydroxyproline and protein models. It is the authors’ impression that with such efforts, potentially in combination with development of practical measurement setup, the use of Raman spectroscopy as a process control tool for the hydrolysis of poultry rest raw materials is within reach.

Learn-to-supervise: Causal reinforcement learning for high-level control in industrial processes

Possessing efficient supervisory control systems is crucial for maintaining the desired operational performance of complex industrial processes. Several challenges face the developers of these systems, such as requiring accurate physical models, dealing with the variability and uncertainty of process operating conditions and coordinating between local controllers to reach desired global performance. This paper proposes an intelligent supervisory control approach based on causal reinforcement learning (CRL) to effectively manipulate the controllers’ setpoints of the process in a way that optimizes its key performance indicators (KPIs), thereby improving the energy efficiency of the process. The approach adopts deep reinforcement learning (DRL) to develop an efficient control policy through interaction with a process simulation. The DRL training history is then exploited using interpretable machine learning and process mining to build a discrete event system (DES) model, in the form of a state-event graph. The DES model identifies causal relationships between events and provides interpretability to the control policy developed by the DRL method. The DES discovered is exploited as a Markov decision process to apply the Q-learning algorithm as a CRL supervisor. The supervisor incorporates causal knowledge into its training process, thus improving the DRL control policy developed and identifying the event paths that optimize the process’s KPIs. The proposed approach is validated using two heat recovery systems in a pulp & paper mill. It successfully achieves a control policy that reduces energy consumption by up to 15.6% for the first system and 5.02% for the second, compared to the expert’s baseline methods.