Modeling Of Industrial Processes Research Articles

The value of recycling for low-carbon energy systems - A case study of Germany's energy transition

To achieve climate neutrality, synergies between circular economy and reduction of greenhouse gas emissions must be strengthened. Previously unused emission reduction potentials of resource efficiency are to be exploited. Since all potentially possible emission reduction measures are linked by interactions, the evaluation of a single measure in terms of cost efficiency, effectiveness, and compliance with climate protection targets is very complex and requires a model-based analysis that takes the entire energy system into account.This work advances an energy system model for Germany so that through comprehensive modeling of industrial processes and implementation of recycling options, the impact of recycling measures in the context of national greenhouse gas mitigation strategies can be analyzed.The scenario evaluation shows that different recycling strategies have large effects on the German energy system. Without recycling energy demand in 2050 will increase by more than 300 TWh and cost of transformation will rise by 85% compared to a reference scenario, with today's recycling rates. On the other hand, if maximum recycling rates can be achieved, costs of transformation can be reduced by 26% until 2050. Recycling is an essential and cost-efficient greenhouse gas reduction strategy for future low-carbon energy system designs.

Chance constrained dynamic optimization approach for single machine scheduling involving flexible maintenance, production, and uncertainty

Chance constraints are suitable for industrial process modeling under uncertain conditions, where constraints cannot be strictly satisfied or do not need to be fully satisfied. In this paper, a single machine scheduling problem involving flexible maintenance, production, and uncertainty is modeled as a chance constrained dynamic optimization problem (CCDOP). A novel method is proposed for transforming the CCDOP into an equivalent deterministic dynamic optimization problem (DOP) with fixed state jump times. Furthermore, by using the idea of l1 penalty function and a smooth approximation technique, the resulting deterministic DOP becomes a smoothing penalty problem, which is a non-convex nonlinear parameter optimization problem (NNPOP) with simple bounds on the variables. To solve the NNPOP, a gradient-based stochastic search algorithm (GSSA) is developed based on a gradient-based adaptive search algorithm (GASA) and a novel stochastic search algorithm (NSSA). The convergence analysis result shows that the GSSA is a globally convergent algorithm. Finally, two numerical examples are used to illustrate the effectiveness of the proposed method. Numerical results show that the GSSA has excellent convergence behavior with robust computation feature, providing better results compared with the other typical methods.

A penalty function-based random search algorithm for optimal control of switched systems with stochastic constraints and its application in automobile test-driving with gear shifts

Practical industrial process is usually a dynamic process including uncertainty. Stochastic constraints can be used for industrial process modeling, when system sate and/or control input constraints cannot be strictly satisfied. Thus, optimal control of switched systems with stochastic constraints can be available to address practical industrial process problems with different modes. In general, obtaining an analytical solution of the optimal control problem is usually very difficult due to the discrete nature of the switching law and the complexity of stochastic constraints. To obtain a numerical solution, this problem is formulated as a constrained nonlinear parameter selection problem (CNPSP) based on a relaxation transformation (RT) technique, an adaptive sample approximation (ASA) method, a smooth approximation (SA) technique, and a control parameterization (CP) method. Following that, a penalty function-based random search (PFRS) algorithm is designed for solving the CNPSP based on a novel search rule-based penalty function (NSRPF) method and a novel random search (NRS) algorithm. The convergence results show that the proposed method is globally convergent. Finally, an optimal control problem in automobile test-driving with gear shifts (ATGS) is further extended to illustrate the effectiveness of the proposed method by taking into account some stochastic constraints. Numerical results show that compared with other typical methods, the proposed method is less conservative and can obtain a stable and robust performance when considering the small perturbations in initial system state. In addition, to balance the computation amount and the numerical solution accuracy, a tolerance setting method is also provided by the numerical analysis technique.

Two-stage energy scheduling optimization model for complex industrial process and its industrial verification

As complex industries move towards small batches and customization, the uncertainty of the operation time of intermittent electrical devices increases, leading to an increasing fluctuation range of the complex industrial electrical load. To ensure the regular production process, the fossil power plant increases the installed capacity of generators, which can sharply raise carbon emissions. Renewable energy generation has been introduced to reduce carbon emissions and achieve power generation sustainability. However, the instability of renewable energy generation increases the fluctuation range of transmission voltage, leading to highly unsafe electricity transmission. Energy Storage Systems (ESSs) solves the instability problem of renewable energy generation. Thus, this study proposes a two-stage energy scheduling optimization model for complex industrial processes. The first stage proposes a scheduling optimization model for intermittent electrical devices with high electricity consumption. The second stage proposes a scheduling optimization model for the ESS to optimize the transmission capacity proportion of photovoltaic (PV) power plants, ESS, and fossil power plants. The results show that the proposed two-stage scheduling optimization model for the complex industrial process can reduce electricity cost by 7.1%–9.1%, and carbon emissions by 384 tons of standard coal in a year.

Semi-supervised echo state network with temporal–spatial graph regularization for dynamic soft sensor modeling of industrial processes

Echo state network (ESN) has been successfully applied to industrial soft sensor field because of its strong nonlinear and dynamic modeling capability. Nevertheless, the traditional ESN is intrinsically a supervised learning technique, which only depends on labeled samples, but omits a large number of unlabeled samples. In order to eliminate this limitation, this work proposes a semi-supervised ESN method assisted by a temporal–spatial graph regularization (TSG-SSESN) for constructing soft sensor model with all the available samples. Firstly, the traditional supervised ESN is enhanced to construct the semi-supervised ESN (SSESN) model by integrating both unlabeled and labeled samples in the reservoir computing procedure. The SSESN computes the reservoir states under high sampling rate for better process dynamic information mining. Furthermore, the SSESN’s output optimization objective is modified by applying the local adjacency graph of all training samples as a regularization term. Especially, in view of the dynamic data characteristic, a temporal–spatial graph is constructed by considering both the temporal relationship and the spatial distances. The applications to a debutanizer column process and a wastewater treatment plant demonstrate that the TSG-SSESN model can build much smoother model and has better generalization capability than the basic ESN models in terms of soft sensor prediction results.

Novel Transformer Based on Gated Convolutional Neural Network for Dynamic Soft Sensor Modeling of Industrial Processes

Industrial process data are usually time-series data collected by sensors, which have the characteristics of high nonlinearity, dynamics, and noises. Many existing soft sensor modeling methods usually focus on dominant variables and auxiliary variables at a single time point while ignoring the timing characteristics of industrial process data. Meanwhile, the soft-sensing methods considering timing characteristics based on the deep learning are usually faced with gradient vanishing and the difficulty in parallel computing. Therefore, a novel Gated Convolutional neural network-based Transformer (GCT) is proposed for dynamic soft sensor modeling of industrial processes. The GCT encodes short-term patterns of the time series data and filters important features adaptively through an improved gated convolutional neural network (CNN). Then, the multihead attention mechanism is applied to modeling the correlation between any two moments. Finally, the prediction results are obtained through a linear neural network layer with the highway connection. In this article, the experiments in the dynamic soft sensor modeling of polypropylene and purified terephthalic acid industrial processes show that the proposed method achieves state-of-the-art comparing with the back propagation neural network, the extreme learning machine, the long short-term memory (LSTM) and the LSTM based on the CNN.

Optimal surrogate building using SVR for an industrial grinding process

ABSTRACT Transient states modeling of industrial grinding process with significant accuracy is extremely essential to run these energy intensive processes in optimal conditions ensuring sustainability. Traditional modeling using physics-based approach not only demands extensive process knowledge but also results in time-expensive models difficult to use during iterative processes like optimization. Proposing Support Vector Regression (SVR) as an alternative data driven tool, such a surrogate building task has been performed under an optimization framework. Minimizing square root of mean square error (RMSE) between ground truth and model predictions, optimal hyper-parameter combination is achieved using a novel genetic algorithm-based formulation indicating a paradigm shift as opposed to the usual practice of determining them heuristically. The RMSE of the grinding model obtained using the proposed formulation is reported as 0.00496. When compared with another model obtained using conventional approach, prediction plots indicate the effectiveness of the novel algorithm. Comparison with coefficient of correlation leads similar conclusion, where the least RMSE model has 99.88% and the conventional model has 71.40% correlation value. Such dynamic surrogates with optimal hyper-parameter settings can be extremely useful for control and optimization of grinding processes and can be easily extendable to design of experiment-based response surface modeling.

Rethinking the Value of Just-in-Time Learning in the Era of Industrial Big Data

Just-in-time learning (JITL) has become a widely used industrial process modeling tool. With the advent of the industrial big data era, rich data information has brought new opportunities to JITL. Specifically, the completeness of data samples in the era of big data provides an important premise and support for the JITL method, prompting us to rethink the application value of JITL in the context of industrial big data. At the same time, the huge amount of data causes certain difficulties in data searching, which is a key issue for JITL. In this article, a parallel computing strategy is adopted to divide the entire computational searching task into several subtasks, and assign the subtasks to parallel computing nodes to complete parallel searching. In this way, not only the full advantage of big data information is effectively utilized, but also the search capability and efficiency under big data are improved. In addition, in order to improve the real-time nature of JITL, a model library management (MLM) strategy is adopted, and the query similar samples are used to operate with existing similar models. And by selectively adding new data, the database management (DBM) strategy is also developed, which not only alleviates the problem of information redundancy, but also reduces the search pressure caused by the increasingly large database. Obviously, MLM and DBM are particularly important as JITL auxiliary tools under industrial big data. Combining parallel computing, a parallel JITL (P-JITL) framework is proposed. As an example, the variational Bayesian factor regression model is transformed into the parallel Bayesian-JITL method for big process data modeling, which is further extended to a nonlinear form. To evaluate the feasibility and efficiency of the developed methods, a real industrial case is demonstrated.

Enhanced cuckoo search algorithm for industrial winding process modeling

Enhanced cuckoo search algorithm for industrial winding process modeling

SYNTHESIS OF A NEURAL NETWORK MODEL OF INDUSTRIAL CONSTRUCTION PROCESSES USING AN INDICATOR SYSTEM

Context. The problem of a neural network model synthesis for industrial processes with the definition of an optimal topology characterized by a high level of logical transparency and acceptable accuracy is considered. The object of research is the process of neural network modeling of industrial processes using an indicator system to simplify and select the topology of neuromodels..
 Objective of the work is consists in synthesis a neural network model of industrial processes with a high level of logical transparency and acceptable accuracy based on the use of the system.
 Method. A method is proposed to use artificial neural networks of feedforward propagation for modeling industrial processes. After evaluating the overall level of complexity of the modeling problem based on the indicator system, it was decided to build a neuromodel based on historical data. Using the characteristics of the input data of the problem, the most optimal structure of the neural network was calculated for further modeling of the system. A high level of logical transparency of neuromodels significantly expands their practical use and reduces the resource intensity of industrial processes.
 Results. Neuromodels of industrial processes are obtained based on historical data. The use of an indicator system made it possible to significantly increase the level of logical transparency of models, while maintaining a high level of accuracy. Constructed neuromodels reduce the resource intensity of industrial processes by increasing the level of preliminary modeling.
 Conclusions. The conducted experiments confirmed the operability of the proposed mathematical software and allow us to recommend it for use in practice in modeling industrial processes. Prospects for further research may lie in the neuroevolutionary synthesis of more complex topologies of artificial neural networks for performing multi-criteria optimization.

A novel dynamic nonlinear partial least squares based on the cascade structure

AbstractFor the collinearity problem of input variables in actual industrial process modeling, a novel dynamic nonlinear partial least squares (PLS) approach is presented to solve this problem. In the proposed method, a novel cascade structure which is composed of an autoregressive exogenous model and a radial basis function neural network is introduced as the inner model of the conventional PLS method, so that the newly established PLS method has both dynamic and nonlinear characteristics. In this improvement, an external model for extracting the latent variables of input and output is established through the original PLS model, which eliminates the collinearity between original dependent variables, and then the aforementioned cascade framework is employed to describe the dynamic and nonlinear relationships between the input and output score variables. Furthermore, in order to estimate the unknown parameters of the established dynamic nonlinear inner model, a stochastic gradient algorithm and a forgetting factor stochastic gradient algorithm are presented. Additionally, a simple cross validation method is used to determine the orders of the inner model. Finally, a numerical simulation example and an application example are given to demonstrate the effectiveness of the proposed approach.

Industrial Process Modeling and Monitoring Based on Jointly Specific and Shared Dictionary Learning

With the development of Industrial Cyber-Physical Systems, data of industrial processes is collected and be used for data-driven process monitoring. But due to the variation of working conditions, the data samples are always featured by characteristics and commonalities. Generally, in industrial process data, characteristics are sparse and commonalities are strongly correlated. The presence of characteristics hampers the feature extract of industrial data, thus further bringing difficulties to process monitoring. To solve this problem, this paper proposes a Jointly Specific and Shared Dictionary Learning (JSSDL) method. Specifically, we first build a specific dictionary and a shared dictionary to reconstruct characteristics and commonalities feature respectively. Then, considering the sparsity of characteristics and the strong correlation of commonalities, we add sparsity constraint to specific dictionary and low-rank constraint to shared dictionary. After that, an iterative optimization algorithm is proposed to optimize the two dictionaries. When online data samples arrive, we use the two dictionaries to reconstruct the data and determine the data is normal or faulty according to the reconstruction error. To verify the superiority of the proposed JSSDL method in process monitoring, we do extensive experiments. The experimental results demonstrate that the proposed method can achieve satisfactory monitoring results when compared with several state-of-the-art methods.

A dynamic sample augmentation Kriging metamodeling method for industrial residue hydrogenation process

AbstractResidue hydrogenation process (RHP) plays an important role in efficient utilization of heavy oil resources. The high‐fidelity model of RHP is so complex that its optimization cost is expensive and even intractable. Furthermore, in the actual industrial processes, due to the low sampling frequencies of some sensors, only a few sampling points can be obtained with some missing values. The insufficient samples can directly affect the performance of the final model. Therefore, a new adaptive dynamic sampling (ADS) method for Kriging metamodeling is proposed. Based on a small number of valid industrial sample points, the proposed method adaptively obtains key information sampling points and selectively adds key information sampling points to update the model. First, the difference maximization strategy and the outlier distance strategy are proposed to obtain new sampling points with both global and regionalized information. Then, the new key points are used to iteratively update the Kriging predictor to obtain satisfactory accuracy. Finally, the calculation result of benchmark cases and two actual experiments validate the effectiveness of the proposed method for constructing surrogate models of complex industrial processes.

Machine Learning for Modeling and Control of Industrial Clarifier Process

In sugar production, model parameter estimation and controller tuning of the nonlinear clarification process are major concerns. Because the sugar industry’s clarification process is difficult and nonlinear, obtaining the exact model using identification methods is critical. For regulating the clarification process and identifying the model parameters, this work presents a state transition algorithm (STA). First, the model parameters for the clarifier are estimated using the normal system identification process. The STA is then utilized to improve the accuracy of the system parameters that have been identified. Metaheuristic algorithms such as Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and State Transition Algorithm are used to evaluate the most accurate model generated by the algorithms. By capturing the principal dynamic features of the process, the clarifier model produced from State Transition Algorithm (STA) acts more like the actual clarifier process. According to the findings, the controllers provided in this paper may be used to achieve greater performance than the standard controller design during the control of any nonlinear procedure, and STA is extremely helpful in modeling a nonlinear process.

Guest Editorial: Thematic Section on Applications of Emerging Computing Technologies in Smart Manufacturing and Industry 4.0

The papers in this special section focus on applications of emerging computing technologies in smart manufacturing and Industry 4.0. Both of these paradigms are transforming factories into highly complex IT systems, generating massive amounts of data. The modeling and optimization of smart industrial processes, to support decision making, has consequently become a new application domain, with its own unique and peculiar issues, for some of the most prominent emerging technologies across the entire computing stack, from hardware and systems design to applicationlevel software. The massive data collection which forms the foundation of digital industries is enabled by increasingly complex and heterogeneous cyber physical systems (CPSs). Owing to that, the design of software for CPSs is receiving renewed interest from research and industry in relation to challenges such as managing variability and dependencies with an unprecedented degree of customization.

Production Prediction Modeling of Industrial Processes for Energy Saving: A Novel Cnn-Based Transformer Integrating Boruta Algorithm

Production Prediction Modeling of Industrial Processes for Energy Saving: A Novel Cnn-Based Transformer Integrating Boruta Algorithm

Gated Broad Learning System Based on Deep Cascaded for Soft Sensor Modeling of Industrial Process

With the advancement of computer and sensor technology, soft sensors have been more and more extensively used in industrial processes. Soft sensors based on deep learning often need to redesign the structure and retrain the model when the prediction results are poor, which consumes a lot of time. Therefore, a deep cascade-gated broad learning system with fast update capability is proposed for industrial process soft sensor modeling. Being inspired by deep learning, the hidden layer features extracted by the autoencoder (AE) are used in the feature nodes of the broad learning system (BLS) to obtain the deep-BLS (D-BLS), which can circumvent the problem of insufficient feature extraction caused by stochastically generated weights in the feature nodes of BLS. On this basis, each feature node is integrated and sent to the enhancement nodes through the gated neurons. The enhancement nodes are cascaded to construct the deep cascaded-gated BLS (DC-GBLS), which can improve the prediction effect of the model while enhancing the utilization rate of the hidden layer features. Finally, a fast update method is developed for the model when the accuracy is insufficient. The validity and superiority of proposed model are demonstrated by two industrial processes.

Feature-based modeling for industrial processes in the context of digital twins: A case study of HVOF process

Industrial process modeling is currently undergoing a fundamental transformation, leading towards interconnected close-loop twins of models, i.e., the parametrically-controlled real-world physics model, and its corresponding digitalized virtual system model. Between these models, a highly time-sensitive mutual validation mechanism and a coherent and consistent control strategy are applied, which demand complex virtual-and-real world model-based information mapping and synchronization. Thus, this research proposes a semantic conceptual framework for industrial process modeling in the context of digital twins. Based on a hierarchical structure of digital twins, this framework modularizes the modeling process in terms of the semantic information modules of physics in the real-world phenomena, and clarifies inter-module associations and near-real-time data transmission coupled with virtual analysis in CAE environment so that the time-sensitive phenomenon information objects distributed on virtually-separated sub-level physics models can be supported for representing the real-world process comprehensively. Advanced feature concept is adopted to construct the digital models as the basic compositions of any virtual industrial process. The related feature definitions are extended in this work so that the common characteristics in the concept of digital twins could be generically and concisely represented. To validate the modeling method, the digital model of a prototyped High-velocity Oxygen-fuel (HVOF) coating process system was constructed as the case study. The data from the nozzle trajectory, the flame, and the in-flight particle behavior, and the transient thermal performance of the coating layer and substrate were synchronously incorporated for simulating the transient phenomena of the substrate component temperature and coating thickness distribution on the substrate surface. The final simulation result validates that the feature-based digital model is able to comprehensively reflect the real-world scenario on the virtual side. To illustrate how the developed model provides meaningful feedback to the real-world control, a general digital twin setup of the model is given at the end of the case study.

Microfluidics: A Novel Approach for Dehydration Protein Droplets.

The equation of state of colloids plays an important role in the modelling and comprehension of industrial processes, defining the working conditions of processes such as drying, filtration, and mixing. The determination of the equation is based on the solvent equilibration, by dialysis, between the colloidal suspension and a reservoir with a known osmotic pressure. In this paper, we propose a novel microfluidic approach to determine the equation of state of a lysozyme solution. Monodispersed droplets of lysozyme were generated in the bulk of a continuous 1-decanol phase using a flow-focusing microfluidic geometry. In this multiphasic system and in the working operation conditions, the droplets can be considered to act as a permeable membrane system. A water mass transfer flow occurs by molecule continuous diffusion in the surrounding 1-decanol phase until a thermodynamic equilibrium is reached in a few seconds to minutes, in contrast with the standard osmotic pressure measurements. By changing the water saturation of the continuous phase, the equation of state of lysozyme in solution was determined through the relation of the osmotic pressure between protein molecules and the volume fraction of protein inside the droplets. The obtained equation shows good agreement with other standard approaches reported in the literature.

Modeling of Industrial and Technological Processes in Complex Systems Based on Neuro-Fuzzy Petri Nets

For effective functioning within a complex system, industrial and technological processes are supported by appropriate information-analytical processes that ensure the collection and analysis of information, as well as modeling and making control decisions for the industrial and technological process. A variety of neuro-fuzzy Petri nets with temporal fuzzy neurons is proposed. An example of con-structing a model of an industrial and technological process and the corresponding information-analytical processes is considered. The developed specialized software for modeling industrial and technological processes and the implementation of information-analytical processes is considered.