Optimal Process Control Research Articles

Enhanced anaerobic digestion model no.1 for high solids fermentation: Integrating homoacetogenesis and chain elongation.

The production of volatile fatty acids (VFA) through high-solids anaerobic fermentation of organic waste offers a promising route for resource recovery. This study used a batch-mode anaerobic leach bed reactor (LBR) with leachate circulation to ferment the organic fraction of municipal solid waste, producing high concentrations of butyric acid, along with notable amounts of lactic and caproic acids. These results provide valuable insights and underscore the need for process optimization within conventional fermentation systems. To better model the LBR's complex dynamics, the Anaerobic Digestion Model No.1 (ADM1) was modified and extended to account for slow hydrolysis in dry fermentation conditions, thermodynamic constraints imposed by Gibbs Free Energy and hydrogen partial pressures, as well as lactic acid production and chain elongation pathways including homoacetogenesis and caproic acid formation. These enhancements provided deeper insights into high-solids anaerobic fermentation, advancing strategies for improved process control and system optimization.

Closed-Loop Manufacturing with AI-Enabled Digital Twin Systems

This research provides an in-depth analysis of recent literature on Closed-Loop Manufacturing with AI-Enabled Digital Twin Systems, focusing on the integration of Artificial Intelligence (AI) and Digital Twin technologies within modern manufacturing environments. Through a systematic literature review of studies sourced from leading academic databases such as Scopus, Web of Science, IEEE Xplore, Science Direct, Springer Link, and Google Scholar, this research examines how these advanced technologies are being used to optimize production processes, improve operational efficiency, and reduce costs. The review synthesizes key findings related to real-time data collection, predictive maintenance, and quality control, highlighting the role of AI in enabling self-regulating and self-improving manufacturing workflows. The application of AI and Digital Twin systems in closed-loop manufacturing facilitates enhanced decision-making through continuous feedback loops. These systems allow manufacturers to simulate, predict, and monitor production processes in real-time, enabling proactive maintenance, process optimization, and better-quality control. Moreover, the integration of AI allows for the dynamic adjustment of manufacturing parameters, reducing waste and improving resource utilization. This research identifies and highlights the potential of AI and Digital Twin technologies in driving sustainability and flexibility in manufacturing operations. However, the research also identifies several challenges in implementing AI-enabled Digital Twin systems, including data integration issues, high initial investment costs, and cybersecurity risks. The shortage of skilled professionals capable of managing these advanced systems further hinders widespread adoption. Based on the synthesis of current literature, the research concludes with recommendations for overcoming these obstacles, offering insights into the future of closed-loop manufacturing systems and their potential to transform industrial production processes.

Simultaneous measurement of particle size distribution and mixing ratio based on Monte Carlo ultrasonic attenuation model

Abstract Mixed particle systems are commonly employed in industrial processes, where the characterization of particle size parameters and mixing ratio can frequently serve as key indicators in process control and production optimization. A Monte Carlo (MC) model was developed to numerically predict and study the ultrasonic attenuation spectrum characteristics in the polymethyl methacrylate (PMMA)-glass aqueous suspension, and together with the particle swarm optimization (PSO) algorithm, to handle the inverse problem in solving the particle size, distribution width, and mixing ratio. The results of the numerical simulations indicate that there exists a linear relationship between the attenuation coefficient and the mixing ratio, with the particle size exerting a significant influence. Furthermore, the multi-parameter simultaneous inversion also yielded calculation deviations of less than 1%, 3%, and 6% for the mixing ratio, characteristic diameter, and distribution width, respectively, in comparison to their given values. Afterward, a series of experiments were conducted to quantify the particle size and mixing ratio through the analysis of ultrasonic spectra. In spherical PMMA-glass aqueous suspensions, the measurement error for the mixing ratio and particle size parameters are found to be less than 7% and 10%, respectively, when compared to the image method and the given values. Nevertheless, the measurement errors are slightly increased in a non-spherical mixed particle system, where the volume median diameter and mixing ratio are still less than 10%. The MC modeling and PSO algorithm offer the potential to characterize particle size and mixing ratio for mixed particle systems in industrial applications.

Optimal Control of Nonlinear Competitive Epidemic Spreading Processes With Memory in Heterogeneous Networks

Optimal Control of Nonlinear Competitive Epidemic Spreading Processes With Memory in Heterogeneous Networks

Nonlinear optimal control of rope hook recovery process

Nonlinear optimal control of rope hook recovery process

Determination of the Bentonite Content in Molding Sands Using AI-Enhanced Electrical Impedance Spectroscopy.

Molding sand mixtures in the foundry industry are typically composed of fresh and reclaimed sands, water, and additives such as bentonite. Optimizing the control of these mixtures and the recycling of used sand after casting requires an efficient in-line monitoring method, which is currently unavailable. This study explores the potential of an AI-enhanced electrical impedance spectroscopy (EIS) system as a solution. To establish a fundamental dataset, we characterized various sand mixtures containing quartz sand, bentonite, and deionized water using EIS in the frequency range from 20 Hz to 1 MHz under laboratory conditions and also measured the water content and density of samples. Principal component analysis was applied to the EIS data to extract relevant features as input data for machine learning models. These features, combined with water content and density, were used to train regression models based on fully connected neural networks to estimate the bentonite content in the mixtures. This led to a high prediction accuracy (R2 = 0.94). These results demonstrate that AI-enhanced EIS has promising potential for the in-line monitoring of bulk material in the foundry industry, paving the way for optimized process control and efficient sand recycling.

Sensitivity Analysis on Policy‐Augmented Graphical Hybrid Models With Shapley Value Estimation

ABSTRACTDriven by the critical challenges in biomanufacturing, including high complexity and high uncertainty, we propose a comprehensive and computationally efficient sensitivity analysis framework for general nonlinear policy‐augmented knowledge graphical (pKG) hybrid models that characterize the risk‐ and science‐based understandings of underlying stochastic decision process mechanisms. The criticality of each input (i.e., random factors, policy parameters, and model parameters) is measured by applying Shapley value (SV) sensitivity analysis to pKG (called SV‐pKG), accounting for process causal interdependences. To quickly assess the SV for heavily instrumented bioprocesses, we approximate their dynamics with linear Gaussian pKG models and improve the SV estimation efficiency by utilizing the linear Gaussian properties. In addition, we propose an effective permutation sampling method with TFWW transformation and variance reduction techniques, namely the quasi‐Monte Carlo and antithetic sampling methods, to further improve the sampling efficiency and estimation accuracy of SV for both general nonlinear and linear Gaussian pKG models. Our proposed framework can benefit efficient interpretation and support stable optimal process control in biomanufacturing.

Safe Transfer-Reinforcement-Learning-Based Optimal Control of Nonlinear Systems.

Traditional reinforcement learning (RL) methods for optimal control of nonlinear processes often face challenges, such as high computational demands, long training times, and difficulties in ensuring the safety of closed-loop systems during training. To address these issues, this work proposes a safe transfer RL (TRL) framework. The TRL algorithm leverages knowledge from pretrained source tasks to accelerate learning in a new, related target task, significantly reducing both learning time and computational resources required for optimizing control policies. To ensure safety during knowledge transfer and training, data collection and optimization of the control policy are performed within a control invariant set (CIS) throughout the learning process. Furthermore, we theoretically analyze the errors between the approximate and optimal control policies by accounting for the differences between source and target tasks. Finally, the proposed TRL method is applied to the case studies of chemical processes to demonstrate its effectiveness in solving the optimal control problem with improved computational efficiency and guaranteed safety.

Metabolomics reveal changes of flavonoids during processing of “nine-processed” tangerine peel (Jiuzhi Chenpi)

“Nine-processed” tangerine peel (Jiuzhi Chenpi) is a special snack in Guangdong, China, which is made by traditional and complex techniques. However, it is not clear how the flavonoid composition changes during its complex processing. In this study, we employed UPLC-ESI-Q Trap-MS/MS to comprehensively evaluate the alterations in flavonoids metabolites at four key processing stages: raw materials (CP1), soaking (CP2), curing (CP3), and aging (CP4). The results showed that total of 165 flavonoids metabolites were detected. Naringin, narirutin, hesperidin, neohesperidin and 3,5,6,7,8,3′,4′-heptamethoxyflavone were the top five compounds in terms of the relative content (>5.00%), and their relative content did not change significantly during processing. From CP1 to CP4, the relative content of 38 compounds, including 3′,4′,5,6,7-Pentamethoxyflavanone, apigenin 6,8-C-diglucoside, and diosmin, decreased continuously during processing. The relative contents of astragalin, naringenin-7-O-glucoside and apigenin-7-O-β-D-glucoside were increased by 2.06 times, 2.12 times and 3.34 times, respectively. Isorhamnetin-O-hexoside-O-pentoside (1.98%), 2,4,4′-trihydroxychalcone (0.43%), daidzein-hexoside-xyloside (0.03%), acacetin-7-O-galactoside (0.03%), chrysin-O-glucoside (0.01%), glycitein (0.01%), and prunetin (0.01%) were only detected in CP4. Correlation analysis showed that total flavonoids and naringin were significantly positively correlated with antioxidant activity. These findings offer valuable insights into the metabolites of “nine-processed” tangerine peel during processing and provide a basis for process optimization and quality control.

Blister-actuated laser-induced forward transfer (BA-LIFT): Understanding blister dynamics for enhanced process control

Blister-Actuated Laser-Induced Forward Transfer (BA-LIFT) stands as an innovative laser-based printing method designed to circumvent conventional nozzle-based systems and eliminate direct interactions between the laser pulse and the material to be transferred. This technique involves the interposition of a polyimide layer, strategically placed between the laser and the material. This work is aimed at understanding the critical role played by blister dynamics in the transference process, which is induced by the laser.Building upon a prior investigation where variations in laser pulse energy were explored, the current research involves the capture of high-speed imagery portraying the evolution and expansion of the blister, executed at distinct focal positions. The primary objective is to assess the blister’s influence on the overall process.During the blister’s expansion, our observations unveiled the presence of rapid oscillations, a consistent phenomenon observed regardless of the focal position. Consequently, we propose two polynomial fits, aimed at accurately replicating the blister’s dynamic behavior. These polynomial fits serve as valuable inputs for subsequent numerical simulations in the context of BA-LIFT, allowing for enhanced process control and performance optimization.

Digital twinning of vertical centrifugal casting

This paper explores the transformative potential of digital twin technology in vertical centrifugal casting (VCC), a cornerstone manufacturing process for high-integrity cylindrical components. By integrating real-time data, physical models, and machine learning algorithms, digital twins unlock a new paradigm of process optimization, predictive maintenance, and quality control. Integration of digital twinning enhances the performance in different domains of work like enhancing the quality of research, production etc. Howbeit, digital twinning is nascent in the domain of manufacturing specially in the casting sub domain. A physical set up of VCC is integrated with Internet of Things (IoT) and data acquisition system to stream the collected data to the cloud-based server. Transformation of Internet of Things (IoT) enabled VCC integrated with different sensors into the digital twin helps in quality prognosis for future applications. The data is further fetched from the cloud and interconnection is established between the digital twin. Real time monitoring, controlling and operating can be done easily with the help of a digital twin to predict the quality and tentative defect locations. Further amplifying these benefits, emerging technologies like Virtual Reality (VR), Augmented Reality (AR), and the Metaverse hold immense promise for revolutionising VCC training, collaboration, and visualisation.

DEVELOPMENT OF AUTOMATED CONTROL SYSTEM AND REGISTRATION OF METAL IN CONTINUOUS CASTING

Context. Modern industrial enterprises face challenges that require introduction of latest technologies to improve efficiency and competitiveness. In metallurgy, one of key stages is continuous casting, where quality of products and economic performance of enterprise depend on accuracy and efficiency of process control. Products made using continuous casting technology are widely used in various industries due to their high mechanical properties, structural uniformity and cost-effectiveness. The development of automated metal management and registration system is becoming not only relevant, but also necessary to ensure stable and efficient production. The problem of improving quality of metal products has always been one of most important tasks in steel industry. Imperfect technological processes, human error and equipment malfunctions can lead to defects in finished metal products. This, in turn, affects final characteristics of products, their durability and reliability. To date, available sources have not yet found complete solution to this problem. Therefore, it is necessary to formulate problem and develop algorithm for operation of automated system for controlling and registering metal in continuous casting. Objective. The goal of work is to develop automated metal management and registration system to improve quality of metal products. Method. To achieve this goal, parametric model was proposed, which is formalized on basis of set theory. The model takes into account key parameters of continuous casting process: material characteristics, structural features of crystallizer, casting modes, metal level in crystallizer, and position of shot stopper. Results. The problem was formulated and key parameters were determined, which are taken into account in system’s algorithm, which made it possible to develop system for controlling parameters of continuous casting to solve problem of improving quality of metal products. Conclusions. To improve quality of metal products and stability of casting process, parametric model was created that is comprehensive, allows optimization of key parameters and ensures accuracy of process control by integrating not only modes of product formation, but also takes into account specific properties of source material (chemical composition of material grade, etc.) and design features of casting plant. Algorithm for automated control system has been developed that takes into account relationships between certain key parameters and ensures optimal control of casting process. Based on proposed complex parametric model and algorithm, automated control and metal registration system was created. The focus of work is on quality and efficiency of metal management and registration in continuous casting, based on modern methods of computer science and engineering. A comprehensive experimental comparison of developed system with commercial analogs in real production conditions was carried out, which allowed us to objectively assess its efficiency and reliability.

Towards High-Quality Investment Casting of Ti-6Al-4V with Novel Calcium Zirconate Crucibles and Optimized Process Control

The investment casting of titanium and its alloys relies on a high resistance of the crucibles and shell molds in terms of temperature and reactivity. The availability of ceramic crucibles that offer sufficient resistance to the titanium melt enables vacuum induction melting (VIM). CaZrO3 prepared from a mixture of CaO and ZrO2 as a raw material for refractory ceramics shows a high corrosion resistance against metallic melts even under very high temperatures up to 1800 °C. Crucibles and shell molds of CaZrO3 were successfully produced and used in subsequent casting trials. This study is focused on the refractory crucibles suitable for casting Ti-6Al-4V (Ti-64) using a tilt casting machine. In order to evaluate the crucible reaction and, therefore, the quality of the castings, chemical analyses, investigations of the microstructures and hardness measurements were carried out. Careful control of the melting duration is mandatory to avoid crucible reactions that otherwise result in contamination of the cast with oxygen and zirconium. This was achieved by modified coil geometries. Under optimized casting conditions, the oxygen and zirconium impurity limits of ASTM B367-09 for titanium castings were met. Based on the correlations found, optimized casting parameters with regard to material quantity, coil geometry and heating power could be determined in order to provide guidance for a high-quality casting process with VIM.

Machining performance analysis of wire-EDM for machining of titanium alloy using multi-objective optimization and machine learning approach

This study investigates the multi-objective optimization of process parameters in wire electrical discharge machining (WEDM) using titanium alloy grade 5 and zinc-coated brass microwire. The novelty lies in the comprehensive investigation of the effects of varying pulse on time (Ton), wire tension (WT), and wire feed rate (WF) on output responses, including surface roughness (SR), surface crack density (SCD), corner diameter (CD), and kerf width (KW). The Box-Behnken design, a sophisticated response surface methodology (RSM) technique, is employed to efficiently explore the complex relationships and interdependencies between these input and output parameters. The study's novel approach incorporates machine learning algorithms, including decision tree, light gradient boosting machine (LightGBM) and random forest, to predict the characteristics of the laser-cut components based on the input parameters. The contour plots generated provide valuable insights into the underlying patterns and associations between input and output features, aiding in the understanding of the WEDM process. The results reveal that the optimized values for SR, SCD, CD, and KW are obtained at specific machining conditions: Ton of 10 μs, WF of 3.79 m/min, and WT of 8 gf. Among the machine learning algorithms employed, random forest outperforms the others in predicting KW, SR, SCD and CD, exhibiting significantly lower mean squared error (MSE) and mean absolute error (MAE) values. This study contributes to the field of WEDM by providing a comprehensive analysis, incorporating advanced experimental design techniques and machine learning algorithms, which can aid in process optimization and quality control in manufacturing applications.

Machine learning for particle size prediction in iron ore grinding process

The mining industry is finally embracing the new opportunities brought by Industry 4.0, and one opportunity to be explored is the use of machine learning and artificial intelligence tools. To meet the continuous global demand for iron ore and improve the performance of facilities, process optimization is essential and may revolutionize the industry with data-based decision-making and increased efficiency and profitability. In iron ore beneficiation plants, the potential gains from machine learning tools are significant, and tend to increase in the milling process, where the grinding phase incurs the highest energy costs among mineral processing operations. This article presents the application of machine learning techniques for the prediction of the main product quality parameter of a milling plant. CRISP-DM, a well-established approach for data science projects, was applied, and the results demonstrated the effectiveness of the model, achieving satisfactory predictive performance. The deployment reduces reliance on laboratory testing and provides energy savings through proactive process control optimization.

Steady-state and dynamic simulation of gas phase polyethylene process

Gas-phase polyethylene (PE) processes are among the most important methods for PE production. A deeper understanding of the process characteristics and dynamic behavior, such as properties of PE and reactor stability, holds substantial interest for both academic researchers and industries. In this study, both steady-state and dynamic models for a gas-phase polyethylene process are established as a simulation platform, which can be used to study a variety of operation tasks for commercial solution polyethylene processes, such as new product development, process control and real-time optimization. The copolymerization kinetic parameters are fitted by industrial data. Additionally, a multi-reactor series model is developed to characterize the temperature distribution within the fluidized bed reactor. The accuracy in predicting melt index and density of the polymer, and the dynamic behavior of the developed models are verified by real plant data. Moreover, the dynamic simulation platform is applied to compare four practical control schemes for reactor temperature by a series of simulation experiments, since temperature control is important in industrial production. The results reveal that all four schemes effectively track the setpoint temperature. However, only the demineralized water temperature cascade control demonstrates excellent performance in handling disturbances from both the recycle gas subsystem and the heat exchange subsystem.

Ai-enhanced thermal modeling for integrated process-product-system optimization in zero-defect manufacturing chains

This paper proposes a novel approach for realising zero-defect manufacturing by integrating AI-based thermal modelling to perform multi-stage process, product, and system optimisation in complex manufacturing chains. It integrates advanced thermal simulation and sensor data feed in real-time as a multistage thermal prediction and management system via machine learning algorithms. This framework seeks to predict thermal behaviour using a combination of convolutional neural networks (CNNs), long short-term memory (LSTM) networks, and graph neural networks (GNNs) for encoding, predicting, and learning the spatial, temporal, and interactive thermal behaviour, respectively. It further embeds finite element analysis (FEA) simulations for high-fidelity thermal predictions using data fusion through Kalman filters. This helps obtain the optimal estimates of thermal states from sensor measurements involving different types of sensors and the characteristics of signals. A multistage multimodal optimization framework involves genetic algorithms (GA) for global thermal parameter optimisation, reinforcement learning (RL) for multi-stage dynamic process control optimisation, and multi-agent systems (MAS) for coordinated multi-stage multi-objective balance embedded in a digital twin architecture. Evaluation results show that the effectiveness of the proposed system in improving the overall production efficiency is 33%, reducing defects is 47%, and reducing energy utilisation is 22%, when compared to the current de facto approaches. There is also a 38% improvement in predictive capability in preventing, detecting, and predicting of cross-stage process faults.

Specific of mechanical alloying of solid-liquid binary system Fe-Ga and effect of different process control agents

Mechanical alloying allows obtaining nonequilibrium structures in various systems, often possessing unique properties, including magnetic ones. Considering the unusual structural features of the magnetostrictive Fe-Ga alloy, this approach may be promising for this system. In this work, extensive experimental studies were carried out aimed at studying the features of mechanical alloying of Fe-Ga. The object of the study was the system Fe-20 wt% Ga in which disordered solid solution α-Fe(Ga) is formed. It was shown that high-intensity milling is an effective tool for mechanical alloying of solid-liquid binary system Fe-Ga, but a serious problem is a low powder recovery, less than 50 %. To solve this problem, various process control agents were tested. Their influence on powder recovery, process kinetics, particle size, carbon contamination, and magnetic properties was studied using a large set of techniques such as XRD, SEM, EDS, VSM, LIBS, and others. It has been shown that, based on a combination of factors, the optimal process control agent for this system is ethanol in an amount of 1 wt%

Optimization of Quality Process Control and Preventive Maintenance Strategy: A Case Study

Optimization of Quality Process Control and Preventive Maintenance Strategy: A Case Study

Dynamic Programming, Neuro-Dynamic Programming, Rollout Method and Model Predictive Control to Optimal Control of a Fermentation Process

This work presents the use of dynamic programming (DP), neuro-dynamic programming (NDP), rollout algorithm (RA) and model predictive control (MPC) for optimal control of batch cultivation of the yeast Kluyveromyces marxianus var. lactis MC5. DP is a widespread method for solving problems related to optimization and optimal process control. To reduce the “curse of dimensionality”, NDP has been implemented as an alternative. In NDP, a neural network is used to solve the dimensionality problem. A simpler NDP method, called RA, is used to approximate the optimal cost through the cost of a relatively good suboptimal policy, called the baseline policy. RA is a suboptimal method for deterministic and stochastic problems that can be solved by DP. In this paper we also present off-line MPC technique for tracking of constrained fermentation systems and it overcomes the problem by off-line optimizations prior to implementation. MPC is used to provide perturbation feedback and it is developed theoretical on base a controller as an illustration how we can avoid disturbances in the process optimisation. The developed control algorithm-combined NDP and MPC ensures maximum biomass production at the end of the process and feedback during disturbances and process stability and shows that robust stability can be ensured.