Large-scale Industrial Processes Research Articles

Fast Locally Weighted PLS Modeling for Large-Scale Industrial Processes

Locally weighted partial least-squares (LW-PLS) is an efficient just-in-time (JIT) modeling method, which can handle process collinearity, nonlinearity, and time-varying characteristics. However, it is not suitable for modeling large-scale industrial processes due to its huge computational cost. To solve this issue, the present work proposes a novel fast LW-PLS (FLW-PLS) method that can handle large-scale process data well. FLW-PLS is designed based on the exact Euclidean locality-sensitive hashing algorithm. Unlike standard LW-PLS which implements a brute-force linear search of similar samples over a reference data set, FLW-PLS searches for similar samples in sublinear time. Thus, significant computational savings can be obtained by FLW-PLS. The effectiveness of the proposed FLW-PLS method was validated through the case studies of predicting the silicon content and phosphorus content in the ironmaking process. The application results have shown that when faced with large-scale data, FLW-PLS can significantly reduce the prediction time without significantly reducing the prediction accuracy in comparison to the conventional LW-PLS method.

Cyber-security of centralized, decentralized, and distributed control-detector architectures for nonlinear processes

Decentralized and distributed control systems provide an efficient solution to many challenges of controlling large-scale industrial processes. With the expansion in communication networks, vulnerability to cyber intrusions also increases. This work investigates the effect of different types of standard cyber-attacks on the operation of nonlinear processes under centralized, decentralized, and distributed model predictive control (MPC) systems. The robustness of the decentralized control architecture over distributed and centralized control architectures is analyzed. Moreover, a machine-learning-based detector is trained using sensor data to monitor the cyber security of the overall system. Specifically, detectors built using feed-forward neural networks are used to detect the presence of an attack or identify the subsystem being attacked. A nonlinear chemical process example is simulated to demonstrate the robustness of decentralized control architectures and the effectiveness of the neural-network detection scheme in maintaining the closed-loop stability of the system.

A Holistic Probabilistic Framework for Monitoring Nonstationary Dynamic Industrial Processes

Multivariate statistical process monitoring (MSPM) methods provide sensitive indicators of process conditions by harnessing the value of massive process data. Large-scale industrial processes are subject to wide-range time-varying operating conditions such that some variables inevitably exhibit nonstationary behavior, which poses significant challenges for the design of MSPM schemes. In this brief, a novel nonstationary probabilistic slow feature analysis algorithm is developed to comprehensively describe both nonstationary and stationary variations that underlie process measurements during routine operations. For efficient parameter estimation, the expectation-maximization algorithm is employed. By modeling nonstationarity and stationarity as the random walk and stable autoregressive processes, interpretable monitoring statistics are constructed to detect abnormality in nonstationary dynamics, stationary dynamics, and stationary steady conditions. This forms a holistic and pragmatic monitoring framework for industrial processes, which is beneficial for reducing false alarms and providing meaningful operational information for industrial practitioners. The efficacy of the proposed monitoring framework is validated via two case studies.

Double-Layer Distributed Monitoring Based on Sequential Correlation Information for Large-Scale Industrial Processes in Dynamic and Static States

Due to the complex static, dynamic, and large-scale characteristics for modern industrial processes, in this article, we propose a double-layer distributed monitoring approach based on multiblock slow feature analysis and multiblock independent component analysis. To this end, the processed dataset is divided into the static and dynamic blocks on the basis of the sequential information of each variable in the first layer. Considering the correlations between the variables in the large-scale processes, the sequential correlation matrices in two blocks are calculated, which serves as the second-layer block division rule. Then, the static and dynamic blocks are further divided into several static and dynamic subblocks in which the variables in each subblock are strongly correlated and in the same state. The slow feature analysis and independent component analysis monitoring models are, respectively, generated for the dynamic and static subblocks. Finally, the monitoring results in each subblock are integrated by Bayesian inference to get the final statistics. The average fault detection rate of the proposed method for the Tennessee Eastman process is 0.842, while those of the other traditional methods are lower than 0.75, which shows the advantages of the proposed method.

Squalene-Tetrahymanol Cyclase Expression Enables Sterol-Independent Growth of Saccharomyces cerevisiae.

Biosynthesis of sterols, which are considered essential components of virtually all eukaryotic membranes, requires molecular oxygen. Anaerobic growth of the yeast Saccharomyces cerevisiae therefore strictly depends on sterol supplementation of synthetic growth media. Neocallimastigomycota are a group of strictly anaerobic fungi which, instead of containing sterols, contain the pentacyclic triterpenoid "sterol surrogate" tetrahymanol, which is formed by cyclization of squalene. Here, we demonstrate that expression of the squalene-tetrahymanol cyclase gene TtTHC1 from the ciliate Tetrahymena thermophila enables synthesis of tetrahymanol by S. cerevisiae Moreover, expression of TtTHC1 enabled exponential growth of anaerobic S. cerevisiae cultures in sterol-free synthetic media. After deletion of the ERG1 gene from a TtTHC1-expressing S. cerevisiae strain, native sterol synthesis was abolished and sustained sterol-free growth was demonstrated under anaerobic as well as aerobic conditions. Anaerobic cultures of TtTHC1-expressing S. cerevisiae on sterol-free medium showed lower specific growth rates and biomass yields than ergosterol-supplemented cultures, while their ethanol yield was higher. This study demonstrated that acquisition of a functional squalene-tetrahymanol cyclase gene offers an immediate growth advantage to S. cerevisiae under anaerobic, sterol-limited conditions and provides the basis for a metabolic engineering strategy to eliminate the oxygen requirements associated with sterol synthesis in yeasts.IMPORTANCE The laboratory experiments described in this report simulate a proposed horizontal gene transfer event during the evolution of strictly anaerobic fungi. The demonstration that expression of a single heterologous gene sufficed to eliminate anaerobic sterol requirements in the model eukaryote Saccharomyces cerevisiae therefore contributes to our understanding of how sterol-independent eukaryotes evolved in anoxic environments. This report provides a proof of principle for a metabolic engineering strategy to eliminate sterol requirements in yeast strains that are applied in large-scale anaerobic industrial processes. The sterol-independent yeast strains described in this report provide a valuable platform for further studies on the physiological roles and impacts of sterols and sterol surrogates in eukaryotic cells.

Bayesian Network Based on an Adaptive Threshold Scheme for Fault Detection and Classification

Data-driven multivariate statistical analysis methods have been widely used in fault monitoring of large scale and complex industrial processes. Condition Gaussian Network (CGN) provides a way of p...

Increased efficiency of combined heat and power plants by utilizing waste heat for resorption chillers and their combination with hydrocarbon chillers

The use of waste heat has long been a topic to increase the efficiency of large-scale industrial processes, but it is becoming as well important in the commercial sector and in small and medium-sized industrial plants. For a sustainable and decentralized energy supply it is necessary to utilize all exergy flows much more than before. Particularly in decentralized combined heat and power (CHP) plants, the annual coefficient of performance is low due to seasonal temperature fluctuations, as the generated heat cannot be used or can only be used partly during the summer and transitional months. Interface technologies such as sorption technology can be used to further increase the utilization rate of such CHP plants and provide cooling capacity at the same time. This concept allows the utilization of previously unused waste heat to generate cooling in the temperature range from -6°C to 15°C, without the necessity of significant amounts of electric energy for the operation of additional compression refrigeration systems. In a pilot project funded by the German Federal Ministry of Economic Affairs and Energy, a thermally driven refrigeration plant with a small cooling capacity of maximum 25 kW using the resorption principle was evolved under these aspects and installed in a supermarket. Special focus is given to the flexibility of the plant technology as well as to the potentials of combinations of thermally and electrically driven chillers with natural refrigerants such as hydrocarbons and ammonia and the integration of thermal storage for temporal decoupling or discontinuous availability of heat and cooling load.

Distributed-ensemble stacked autoencoder model for non-linear process monitoring

Determining whether a fault occurs locally or globally is highly important for large-scale industrial processes involving multiple operating units. Moreover, the complex non-linearity among process variables is a prominent feature of modern industries. This paper proposes a distributed-ensemble stacked autoencoder (DE-SAE) model based on deep learning technology for monitoring non-linear, large-scale, multi-unit processes. First, the deep features of the variables involved in each operating unit are extracted with the stacked autoencoder (SAE) to represent the essential structure of the unit. Two statistics are separately constructed using the deep features and the reconstruction error for detecting the faults in local units. Subsequently, the deep representations of the variables from each operating unit are modeled with the SAE to extract the global information for global monitoring. The proposed DE-SAE model uses deep learning techniques to solve the complex non-linear relationships in industrial processes, while considering their local and global information. Therefore, the method can explain the monitoring results better. Experimental results obtained from the numerical simulation and Tennessee-Eastman process confirm the feasibility and superiority of this method.

Preparation and thermal properties of lauric acid/raw fly ash/carbon nanotubes composite as phase change material for thermal energy storage

In this study, a novel ternary system form-stable phase change material (FSPCM) composed of lauric acid (LA)/raw fly ash (RFA)/carbon nanotubes (CNT) was prepared via low cost, easy and industrially applicable fabrication process for low-temperature heat storage. Particularly, the unmodified RFA was directly acted as supporting material to prevent the leakage of the melted LA almost at no cost. A series of leakage experiments were performed to evaluate the package efficiency. The maximum mass fraction of LA absorbed in RFA and CNT was found to be 25 wt% without the LA leakage. Hence, the LA/RFA/CNT (25/75/5 wt%) composite was characterized as FSPCM. The chemical structures, microstructure thermal properties and thermal stability of the FSPCM was investigated by Fourier transformation infrared spectroscope (FTIR), scanning electronic microscope (SEM), differential scanning calorimetry (DSC) and thermal gravimetric analyzer (TGA). The SEM and FTIR results indicated that LA was adsorbed on the RFA’s surface porous or into the porous structure of CNT. And there was good chemical compatibility among LA, RFA and CNT. The DSC results demonstrated that the phase change temperatures and latent heats of LA/RFA/CNT FSPCM were 45.36 °C and 37.83 J/g for melting and 40.51 °C and 36.48 J/g for freezing, respectively. TGA analysis test revealed that the composite PCM had excellent thermal stability. Moreover, the heat transfer efficiency of LA/RFA/CNT FSPCM has been improved by the addition of RFA and CNT. In short, the LA/RFA/CNT FSPCM has a promising application prospect in low-temperature application due to feasible and in large scale industrial preparation, low-cost, simple and facile process.

Decarbonising energy: The developing international activity in hydrogen technologies and fuel cells.

Hydrogen technologies and fuel cells offer an alternative and improved solution for a decarbonised energy future. Fuel cells are electrochemical converters; transforming hydrogen (or energy sources containing hydrogen) and oxygen directly into electricity. The hydrogen fuel cell, invented in 1839, permits the generation of electrical energy with high efficiency through a non-combustion, electrochemical process and, importantly, without the emission of CO2 at its point of use. Hitherto, despite numerous efforts to exploit the obvious attractions of hydrogen technologies and hydrogen fuel cells, various challenges have been encountered, some of which are reviewed here. Now, however, given the exigent need to urgently seek low-carbon paths for humankind's energy future, numerous countries are advancing the deployment of hydrogen technologies and hydrogen fuel cells not only for transport, but also as a means of the storage of excess renewable energy from, for example, wind and solar farms. Furthermore, hydrogen is also being blended into the natural gas supplies used in domestic heating and targeted in the decarbonisation of critical, large-scale industrial processes such as steel making. We briefly review specific examples in countries such as Japan, South Korea and the People's Republic of China, as well as selected examples from Europe and North America in the utilization of hydrogen technologies and hydrogen fuel cells.

Playing with Fire? A Safe and Effective Deactivation of Raney Cobalt using Aqueous Sodium Nitrate

Sponge or skeletal metal catalysts (such as Raney-type hydrogenation catalysts) are ubiquitous and extensively used in large-scale industrial hydrogenation processes, including petrochemical refini...

Durable Light-Driven Three-Dimensional Smart Switchable Superwetting Nanotextile as a Green Scaled-Up Oil–Water Separation Technology

Stimuli-responsive polymer architectures are attracting a lot of interest, but it still remains a great challenge to develop effective industrial-scale strategies. A single-stage and cost-effective approach was applied to fabricate a three-dimensional (3D) smart responsive surface with fast and reversibly switchable wetting between superhydrophobicity and superhydrophilicity/underwater superoleophobicity properties induced by photo and heat stimuli. Commercially available PVDF and P25TiO2 as starting materials fabricated with a scaled-up electrospinning approach were applied to prepare 3D smart switchable PVDF-P25TiO2 nanotextile superwetted by both UV and solar light that is simply recovered by heat at a reasonable time. The superhydrophilic/underwater superoleophobic photo-induced nanotextile will act in “water-removing” mode in which water quickly passes through and the oil is blocked on the surface. An acceptable recycling, reusing, and superior antifouling and self-cleaning performance arising from a TiO2 photocatalytic effect makes it highly desired in a green scaled-up industry oily wastewater treatment technology. With these advantages, a large-scale industrial production process can be simply simulated by applying a conducting mesh-like collector substrate.

Modeling Large-Scale Industrial Processes by Multiple Deep Belief Networks With Lower-Pressure and Higher-Precision for Status Monitoring

Typically, fault detection using deep learning is performed based on the features extracted from only one well-trained deep model. However, our results show that large-scale data is complicated and originates from different schemas, which will cause great pressure on deep neural networks, furthermore, the quality of the extracted features will be affected, and the training complexity and time will also be increased. Conversely, deep models would feel comfortable to extract features from raw data that contain less complex relationships and the quality of extracted features are higher and more representative. Hence, variables from large-scale industrial processes in this study are reasonably divided into various schemas with simple relationships by mutual information. Then, the corresponding deep belief network (DBN) models are established under a lighter pressure state to sufficiently extract the abstract and high-order information from data in each schema. Experimental analysis shows that the training efficiency, the accuracy of extracted features and the monitoring performance based on the proposed model system are all better than using only one DBN. What’s more, a comparison with those of representative and state-of-the-art methods on numerical and Tennessee Eastman processes also demonstrates the high performance of the proposal called M-DBN.

Multi-Block Fault Detection for Plant-Wide Dynamic Processes Based on Fault Sensitive Slow Features and Support Vector Data Description

This study proposes a multi-block fault detection method based on fault-sensitive slow features for large-scale dynamic industrial processes. Firstly, slow feature analysis (SFA) can effectively extract the process dynamic information. However, the slowest changing features may not contain more fault information. Thus, through the analysis of T 2 statistic in SFA-based process monitoring model, a fault sensitivity coefficient is defined as a new slow feature sorting criterion to select the most sensitive slow features to fault in each variable direction. Then, considering the unknown characteristics of the fault in the real-time monitoring process, the monitoring model is established for each dimension of variables based on the multi-block strategy. Finally, the support vector data description is used as a fusing method to integrate the statistics calculated in each sub-block to obtain an intuitive detection result. The effectiveness and superiority of the proposed strategy are demonstrated by the experiments on Tennessee Eastman benchmark process and an actual blast furnace ironmaking process.

Multi-mode solid–gas thermochemical resorption heat transformer using NiCl2-SrCl2/NH3

Thermal energy storage is considered as an attractive technology for waste heat recovery and solar energy utilization. A multi-mode solid–gas thermochemical resorption heat transformer for thermal energy storage has been developed. The working principle of three different operating modes, including the direct thermal energy storage and release mode, thermal energy upgrade mode and combined cooling and heating effect mode, is addressed. Performance analysis and comparisons of thermochemical resorption heat transformer are carried out by employing NiCl2-SrCl2/NH3 under the different operating modes. Thermodynamic analysis showed that the lowest and highest heat storage efficiency occur in the thermal energy upgrade mode and the direct thermal energy storage and release mode, respectively. The highest heat storage efficiency obtained is 0.978. Among three operating modes, the thermal energy upgrade mode has the lowest thermal energy storage density, while the combined cooling and heating effect mode has the highest thermal energy storage density. The smallest and largest thermal energy storage density are 1320.66 kJ/kg and 2646.92 kJ/kg, respectively. The developed thermochemical sorption energy storage provides a compact high energy density heat storage approach and therefore can boost the application of thermal energy storage in large-scale industrial processes and renewable energy utilization.

Nonzero-Sum Game Reinforcement Learning for Performance Optimization in Large-Scale Industrial Processes.

This article presents a novel technique to achieve plant-wide performance optimization for large-scale unknown industrial processes by integrating the reinforcement learning method with the multiagent game theory. A main advantage of this technique is that plant-wide optimal performance is achieved by a distributed approach where multiple agents solve simplified local nonzero-sum optimization problems so that a global Nash equilibrium is reached. To this end, first, the plant-wide performance optimization problem is reformulated by decomposition into local optimization subproblems for each production index in a multiagent framework. Then, the nonzero-sum graphical game theory is utilized to compute the operational indices for each unit process with the purpose of reaching the global Nash equilibrium, resulting in production indices following their prescribed target values. The stability and the global Nash equilibrium of this multiagent graphical game solution are rigorously proved. The reinforcement learning methods are then developed for each agent to solve the nonzero-sum graphical game problem using data measurements available in the system in real time. The plant dynamics do not have to be known. Finally, the emulation results are given to show the effectiveness of the proposed automated decision algorithm by using measured data from a large mineral processing plant in Gansu Province, China.

Bifunctional Synergy in CO Hydrogenation to Methanol with Supported Cu

Future energy storage could be distributed at local plants and involve production of methanol from reaction of sustainably derived hydrogen with CO or CO2 from locally available carbon sources. Such decentralized production would benefit from milder operating conditions than found in the current large-scale industrial process. We propose that a route via CO hydrogenation deserves consideration for this purpose, as it will be free of water, which is unavoidable from CO2-containing gas and strongly inhibiting to the methanol synthesis at lower temperatures. On pure Cu the rate of methanol synthesis from CO is an order of magnitude lower than the rate from CO2, but active CO hydrogenation catalysts can emerge from a bifunctional mechanism in catalysts that combine copper with a basic oxide. Mechanistic studies are consistent with the bifunctional Cu/support synergy arising from a mechanism, where basic oxide sites activate CO as formates at the metal/oxide interface followed by metal assisted hydrogenation of the interfacial formates. Active catalysts for CO hydrogenation are strongly inhibited by CO2, which forms carbonates that block the basic oxide sites and thereby prevent the synergistic pathway from CO.

A distributed expectation maximization-principal component analysis monitoring scheme for the large-scale industrial process with incomplete information

Large-scale process monitoring has become a challenging issue due to the integration of sub-systems or subprocesses, leading to numerous variables with complex relationship and potential missing information in modern industrial processes. To avoid this, a distributed expectation maximization-principal component analysis scheme is proposed in this paper, where the process variables are first divided into several sub-blocks using two-layer process decomposition method, based on knowledge and generalized Dice’s coefficient. Then, the missing information of variables is estimated by expectation maximization algorithm in the principal component analysis framework, then the expectation maximization-principal component analysis method is applied for fault detection to each sub-block. Finally, the process monitoring and fault detection results are fused by Bayesian inference technique. Case studies on the Tennessee Eastman process is applied to show the effectiveness and performance of our proposed approach.

Prediction intervals based soft sensor development using fuzzy information granulation and an improved recurrent ELM

With the increasing complexity of large-scale industrial production processes, the number of variable factors is increasing. As a result, it is demanding to predict process key variables accurately. Currently, most of soft sensor models using support vector regression and artificial neural networks are based on point prediction. The soft measurement models using the technique of point prediction can only track or fit set values. It is difficult to deal with the problem of system uncertainty and to make reliability analysis using the point prediction based soft sensors. To address this problem, this paper proposes a development method of soft sensor using the technique of prediction intervals. Under this condition, the prediction intervals instead of the point prediction of the stable operation of the industrial process system are used. The interval boundaries of the trend change can be utilized to quantify and estimate the associated uncertainty. The proposed prediction intervals based soft sensor is based on fuzzy information granularity and improved recurrent extreme learning machine. First, the fuzzy information granularity is adopted to get the lower bound, trend and upper bound of the interval. Secondly, an improved recurrent extreme learning machine is built to further enhance the ability of prediction intervals. In the improved extreme learning machine model, a feedback layer is adopted to store the hidden layer output, calculate the data trend change and dynamically update the outputs of the feedback layer. Third, the comprehensive interval evaluation function is used to evaluate the rationality of the interval results. Through case studies using a University of California Irvine dataset and the purified Terephthalic acid solvent system, the provided prediction intervals method can directly generate the upper and lower bounds for process key variables with high accuracy.

About Making Lignin Great Again-Some Lessons From the Past.

Lignin, the second most abundant biopolymer on the planet, serves land-plants as bonding agent in juvenile cell tissues and as stiffening (modulus-building) agent in mature cell walls. The chemical structure analysis of cell wall lignins from two partially delignified wood species representing between 6 and 65% of total wood lignin has revealed that cell wall-bound lignins are virtually invariable in terms of inter-unit linkages, and resemble the native state. Variability is recognized as the result of isolation procedure. In native state, lignin has a low glass-to-rubber transition temperature and is part of a block copolymer with non-crystalline polysaccharides. This molecular architecture determines all of lignin's properties, foremost of all its failure to undergo interfacial failure by separation from (semi-) crystalline cellulose under a wide range of environmental conditions. This seemingly unexpected compatibility (on the nano-level) between a carbohydrate component and the highly aromatic lignin represents a lesson by nature that human technology is only now beginning to mimic. Since the isolation of lignin from lignocellulosic biomass (i.e., by pulping or biorefining) necessitates significant molecular alteration of lignin, isolated lignins are highly variable in structure and reflect the isolation method. While numerous procedures exist for converting isolated (carbon-rich) lignins into well-defined commodity chemicals by various liquefaction techniques (such as pyrolysis, hydrogenolysis, etc.), the use of lignin in man-made thermosetting and thermoplastic structural materials appears to offer greatest value. The well-recognized variabilities of isolated lignins can in large part be remedied by targeted chemical modification, and by adopting nature's principles of functionalization leading to inter-molecular compatibility. Lignins isolated from large-scale industrial delignification processes operating under invariable isolation conditions produce polymers of virtually invariable character. This makes lignin from pulp mills a potentially valuable biopolymeric resource. The restoration of molecular character resembling that in native plants is illustrated in this review via the demonstrated (and in part commercially-implemented) use of pulp lignins in bio-degradable (or compostable) polymeric materials.