Soft Sensor Technology Research Articles

Enhancing Fermentation Process Monitoring through Data-Driven Modeling and Synthetic Time Series Generation.

Soft sensors based on deep learning regression models are promising approaches to predict real-time fermentation process quality measurements. However, experimental datasets are generally sparse and may contain outliers or corrupted data. This leads to insufficient model prediction performance. Therefore, datasets with a fully distributed solution space are required that enable effective exploration during model training. In this study, the robustness and predictive capability of the underlying model of a soft sensor was improved by generating synthetic datasets for training. The monitoring of intensified ethanol fermentation is used as a case study. Variational autoencoders were employed to create synthetic datasets, which were then combined with original datasets (experimental) to train neural network regression models. These models were tested on original versus augmented datasets to assess prediction improvements. Using the augmented datasets, the soft sensor predictive capability improved by 34%, and variability was reduced by 82%, based on R2 scores. The proposed method offers significant time and cost savings for dataset generation for the deep learning modeling of ethanol fermentation and can be easily adapted to other fermentation processes. This work contributes to the advancement of soft sensor technology, providing practical solutions for enhancing reliability and robustness in large-scale production.

Dual temporal attention mechanism-based convolutional LSTM model for industrial dynamic soft sensor

Abstract Deep learning is an appropriate methodology for modeling complex industrial data in the field of soft sensors, owing to its powerful feature representation capability. Given the nonlinear and dynamic nature of the process industry, the key challenge for soft sensor technology is to effectively mine dynamic information from long sequences and accurately extract features of relevance to quality. A dual temporal attention mechanism-based convolutional long short-term memory network (DTA-ConvLSTM) under an encoder-decoder framework is proposed as a soft sensor model to acquire quality-relevant dynamic features from serial data. Considering different influences of process variables for prediction at multiple time steps and various locations, ConvLSTM and temporal self-attention mechanism are utilized as the encoder to adaptively fuse spatiotemporal features and capture long-term dynamic properties of process in order to capture the trends of industrial variables. Furthermore, a quality-driven temporal attention mechanism is employed throughout the decoding process to dynamically select relevant features to more accurately track quality changes. The encoder-decoder model meticulously analyses the interactions between process and quality variables by incorporating dual-sequence dynamic information to improve the prediction performance. The validity and superiority of the DTA-ConvLSTM model was validated on two industrial case studies of the debutanizer column and sulfur recovery unit. Compared to the traditional LSTM model, the proposed model demonstrated a substantial improvement with the accuracy R2 up to 97.3% and 94.9% and the root mean square error reducing to 0.122 and 0.022.

Soft sensor modeling based on masked convolutional transformer block deep residual shrinkage network

BackgroundData-driven soft sensor technology is currently an important means for industrial data prediction, addressing the challenge of predicting key quality variables in industrial production processes. Convolutional neural networks (CNN) are widely applied in soft sensor modeling due to their excellent nonlinear modeling and feature extraction capabilities. However, CNN also faces several issues, such as poor robustness to interference and difficulties in extracting features from complex process data. MethodsThis paper introduces a novel CNN model called the Masked Convolutional Transformer Block Deep Residual Shrinkage Network (MCTB-DRSN). Firstly, the Masked Convolutional Transformer Block (MCTB) is utilized to extract features from different positions, thereby enabling the network model to focus more on important information. Secondly, the Global Response Normalization (GRN) layer is incorporated into the Deep Residual Shrinkage Network (DRSN) module, which enhances feature competition among channels. Significant FindingsThis method can provide effective monitoring for chemical production process. Compared with the traditional soft sensor method on the debutanizer column dataset, the results show that the prediction accuracy of this model is significantly improved.

Soft Sensor Technology for the Determination of Mechanical Seal Friction Power Performance

Mechanical seals ensure the internal sealing of centrifugal pumps from the surrounding environment. They are one of the most critical components in a centrifugal pump. For this reason, the condition of mechanical seals should be monitored during operation. Mechanical seal friction power is an important component of mechanical losses in centrifugal pumps and is used as an indicator of wear and therefore seal condition. The soft sensor described in this paper is based on temperature measurements at the seal and can be used for determining the frictional power performance. A major factor in determining frictional power performance is the heat transfer between the mechanical seal and the medium inside the pump. For calculating the heat transfer, the stationary temperature fields in the rings of the mechanical seal are described by transmission efficiencies. The root mean squared error was determined for steady-state operating conditions to assess the quality of the soft sensor calculation. The frictional power performance can be determined by recording the temperature at the mechanical seal mating ring and the medium. The algorithm detects when the steady-state operating conditions change but does not map the dynamic changes between the stationary operating conditions.

Design of a soft sensor based on silver-coated polyamide threads and stress-strain modeling via Gaussian processes

The demand for reliable and efficient soft sensors has grown exponentially with the evolution of wearable devices and smart textiles. However, existing soft sensors often face challenges related to hysteresis, noise, and accuracy, hindering their seamless integration into practical applications. Our research presents a pioneering stress–strain soft sensor model based on silver-coated polyamide threads, augmented with additional silicone and graphite coatings for enhanced properties. Specifically, the silicone coating proves instrumental in elevating the sensor’s gauge factor and reducing noise, resulting in heightened accuracy. The extensive data analysis reveals the presence of hysteresis and non-linearities; however, our data exhibits remarkable robustness, as indicated by the high Spearman correlation coefficient values. In the context of system identification, a comparative analysis between traditional regression methods and Gaussian Process (GPs) Regression demonstrates the superior performance of GPs: this technique outperforms conventional regression techniques, obtaining 8.75 ± 4.06% Root Mean Square Error (RMSE) compared to the 12.70 ± 7.04% error observed in traditional methods. This research not only advances the field of soft sensor technology by developing an accurate, affordable and adaptable device, but also offers valuable insights into highly effective system identification techniques tailored for wearable devices.

Prediction of intensified ethanol fermentation of sugarcane using a deep learning soft sensor and process analytical technology

AbstractBACKGROUNDIntensified ethanol fermentation produces higher ethanol concentrations while reducing water and energy requirements. Nevertheless, the inhibitory and detrimental effect of the cellular stress barriers in this process further complicates the nonlinear dynamic relationship between the variables that directly reflect the fermentation quality. These key variables are hard to measure in real time and therefore cannot be directly controlled.RESULTSThis work presents the development of a soft sensor that predicts in real time the ethanol and substrate concentrations of an intensified fermentation. The soft sensor uses feedforward neural networks (FNNs) with easily measurable process analytical technology (PAT) tools. The application of advanced PAT tools such as redox potential and capacitance, in addition to temperature and pH are explored as input variables. The complex kinetic relationship between the studied variables was captured with FNN architectures with a single hidden layer and between 95 and 175 hidden neurons for the different cases studied. Acceptable predictions are achieved for the concentration of ethanol (RMSE = 9.5 and R2 = 0.97) and substrate (RMSE = 17.02 and R2 = 0.92).CONCLUSIONSThe results confirm that the proposed soft sensor can accurately predict the ethanol and substrate concentrations. Collectively, capacitance, redox potential, temperature and pH provide a powerful platform of PAT tools that can directly infer key variables showing the fermentation quality in real time. This study provides a significant step towards the systematic development of a reliable soft sensor with integration of advanced PAT tools. © 2023 Society of Chemical Industry.

An online transfer kernel recursive algorithm for soft sensor modeling with variable working conditions

Soft sensor technology has found widespread application in the real-time detection of challenging variables like product quality and key process parameters. However, changes in working conditions can lead to alterations in data distribution, resulting in model inaccuracies. Furthermore, data from new working conditions arrive sequentially and online. Due to technical difficulties, time-consuming measurements, and high costs, there is often a lack of sufficient new data for modeling in the initial phase. Moreover, industrial processes often exhibit dynamic, nonlinear, variable coupling and other factors, posing significant challenges for model establishment. To address these challenges, this study proposed an online transfer kernel recursive algorithm for soft sensor modeling under changing working conditions. The algorithm consists of an offline phase and an online phase. In the offline phase, the algorithm considers the dynamic features of the process and variable coupling, resulting in the establishment of a dynamic inner model. The original data is then projected onto a latent variable space. In the online phase, the proposed method incorporates the concept of online transfer learning and takes into account the nonlinearity of the process. This allows for the establishment of a kernel online recursive method in the latent variable space, integrating parameter transfer and sample transfer. Experimental results conducted on multiple industrial datasets validate the effectiveness of the proposed approach.

A virtual sample generation method based on manifold learning and a generative adversarial network for soft sensor models with limited data

BackgroundIn the modern industrial process, data-driven-based soft sensor technology plays a key role in intelligent measurement. However, due to the strong dependence of the prediction accuracy on the training data, it can be challenging to apply this technology to particular industrial processes for which obtaining enough training samples is difficult. MethodsTo solve the problem of insufficient training data for soft sensor models, this paper proposes the Potential of Heat-diffusion for Affinity-based Trajectory Embedding and generative adversarial network based virtual sample generation (PGAN-VSG) method, which combines Potential of Heat-diffusion for Affinity-based Trajectory Embedding (PHATE) based feature extraction, virtual feature generation (VFG) techniques, and PHATE based generative adversarial network (PGAN) to generate virtual samples. The methodology of the work consists of several phases. First, the PHATE is deployed to embed the original data into a low-dimensional space to obtain the original feature set, which preserves the local and global structure of the original data. Second, a number of candidate features are randomly generated in the neighborhood space of each original feature, and the features located in the sparse feature region are selected as virtual features. Furthermore, the PGAN with L1 distance loss is used to map the virtual features to the corresponding high-dimensional space to obtain virtual samples. To validate the performance of the proposed PGAN-VSG method, a multilayer perceptron (MLP) model was employed as a soft sensor model to showcase the effectiveness of the proposed approach. This was achieved through both a numerical example and an industrial case study involving the penicillin fermentation process. Significant findingsBy filling the sample scarcity region while ensuring the distribution of generated samples, the virtual sample generation (VSG) method proposed in this paper improves the prediction accuracy of the soft sensor model while also outperforming the classical VSG method.

Bioreceptor-inspired soft sensor arrays: recent progress towards advancing digital healthcare

Recent advances in soft sensor technology have pushed digital healthcare toward life-changing solutions. Data reliability and robustness can be realised by building sensor arrays that collect comprehensive biological parameter data from several points on the underlying organs simultaneously, a principle that is inspired by bioreceptors. The rapid growth of soft lithography and printing, three-dimensional (3D) printing, and weaving/knitting technologies has facilitated the low-cost development of soft sensors in the array format. Advances in data acquisition, processing, and visualisation techniques have helped with the collection of meaningful data using arrays and their presentation to users on personal devices through wireless communication interfaces. Local- or cloud-based data storage helps with the collection of adequate data from sensor arrays over time to facilitate reliable prognoses based on historical data. Emerging energy harvesting technologies have led to the development of techniques to power sensor arrays sustainably. This review presents developmental building blocks in wearable and artificial organ-based soft sensor arrays, including bioreceptor-inspired sensing mechanisms, fabrication methods, digital data-acquisition techniques, methods to present the results to users, power systems, and target diseases/conditions for treatment or monitoring. Finally, we summarise the challenges associated with the development of single and multimodal array sensors for advanced digital healthcare and suggest possible solutions to overcome them.

A Multistep Sequence-to-Sequence Model With Attention LSTM Neural Networks for Industrial Soft Sensor Application

Soft sensor technology is widely used in the industries to handle highly nonlinear, dynamic, time-dependent sequence data of industrial processes for predicting the key variables associated with auxiliary process variables. Many existing soft sensor algorithms based on deep-learning are able to build complex nonlinear models but ignore the dynamic characteristics of processes. The long short-term memory (LSTM) neural network is exploited to solve the modeling issue which is related to strong time-varying features. In this paper, a novel multi-step sequence-to-sequence model based on attention LSTM (MA-LSTM) neural networks is proposed to improve the soft sensor modeling performance of industrial processes with strong dynamics and nonlinearity. The LSTM-based encoder-decoder architecture with the attention mechanism is applied to extract inherent characteristics relevant to the quality variables and capture both the long-term and short-term dependences of sequence data. Instead of the traditional sequence-to-point quality prediction architecture, the improved architecture is designed to predict multi-step quality variables, and the intermediate results are fed into an additional one-dimensional weighted convolution module to obtain accurate prediction results. The superiority of the proposed framework is demonstrated through a debutanizer column case and a sulfur recovery process.

Dynamic transfer soft sensor for concept drift adaptation

Data-driven soft sensor technology has been widely used in process monitoring, quality prediction, etc. However, there are dynamic time-varying and concept drift problems in industrial processes, making the accurate modeling of the soft sensor is still a challenging task. To solve the above problems, this paper proposes a concept drift adaptive dynamic partial least squares method. The method maps high-dimensional process data into a low-dimensional latent variable subspace. Dynamic information is introduced by establishing the dynamic regression relationship between process latent variables and quality latent variables. At the same time, transfer learning is used to align second-order statistics between latent variables of data from different distributions to solve the concept drift problem. The experiments on multiple industrial datasets show that the method proposed in this paper can effectively reduce prediction errors and improve the generalization ability of the model.

Profitability related industrial-scale batch processes monitoring via deep learning based soft sensor development

Data-driven soft sensor technology has been widely developed to estimate quality-related variables, while following difficulties still limit its application in batch processes, such as different initial conditions, uneven-length of batches, and the extraction of within-batch multiphase features. To address these problems, a quality prediction and monitoring framework is proposed. Variables related to quality-related variables are first selected, and a data stacked strategy is proposed to transform three-dimensional batch data into time-lagged sequences that can be fed into soft sensor models. Aiming to extract the multiphase features, a novel differential recurrent neural networks is established by embedding differential operations into long short-term memory neural networks. Moreover, to ensure profitability, prediction residuals are employed for quality monitoring. Case study on a simulation dataset and an industrial-scale penicillin fermentation process demonstrates the effectiveness of the proposed method and its applicability to batch process monitoring and control in both academic research and industrial operation.

A semi-supervised soft sensor method based on vine copula regression and tri-training algorithm for complex chemical processes

Soft sensor technology is an important solution for timely prediction of some difficult-to-measure variables in the chemical process. For the data-driven soft sensor models, the modeling process needs a certain amount of labeled samples. In some actual processes, the number of unlabeled samples is very large, but the labeling process of unlabeled samples is very time-consuming and laborious. When soft sensor modeling is performed in the case of sparse labeled samples, the generalization ability of the model is bound to be poor. To address this problem, a semi-supervised soft sensor method based on vine copula regression and tri-training algorithm (tri-training VCR) is proposed in this paper. Semi-supervised learning is to enhance the performance of supervised learning models by using unlabeled samples. The tri-training algorithm is used to expand the labeled samples through the cooperation of three vine copula regression models with different structures and a new method to judge the confidence degree of pseudo-labeled samples, so as to improve the prediction accuracy of the soft sensor model. A numerical example and an industrial example are used to demonstrate the effectiveness of the proposed method.

Semi-supervised LSTM with historical feature fusion attention for temporal sequence dynamic modeling in industrial processes

In modern industrial processes, the data-driven soft sensor technology has been widely used for the prediction of key quality variables. Due to the important of dynamics and nonlinearity in industrial process data, deep learning models like long short-term memory (LSTM) network are well suited for temporal sequence dynamic modeling due to their excellent long-term memory function and feature extraction capability. Furthermore, industrial processes generate a large amount of process data with irregular sampling frequencies. However, traditional LSTM cannot fully utilize the process data with irregular sampling frequency and the guidance value of historical data samples for feature learning. To address these issues, a novel semi-supervised LSTM with history feature fusion attention (HFFA-SSLSTM) model is proposed in this paper. First, the semi-supervised learning strategy is implemented in LSTM to fully utilize the unlabeled data and mine the temporal sequence features of labeled samples and unlabeled samples with irregular sampling frequencies. Then, a novel historical feature fusion attention (HFFA) mechanism is developed, which utilizes historical hidden features to learn attention scores for obtaining weighted historical information-related features. Finally, the extracted features are combined to form the soft sensor model to perform time series prediction tasks for key quality variables in industrial processes. The experimental results on the actual industrial hydrocracking data set demonstrate the effectiveness of the proposed HFFA-SSLSTM model and its possibility of applicating in real industrial processes.

Horizontal Data Augmentation Strategy for Industrial Quality Prediction.

In recent years, neural network-based soft sensor technology has been widely used in industrial production processes and has excellent optimization, monitoring, and quality prediction performance. This paper proposes a horizontal data augmentation strategy to provide highly available data for subsequent prediction models, called the combined autoencoder data augmentation (CADA) strategy. This paper has developed a CADA-based convolutional neural network (CADA-CNN) soft sensor model and applied it to the process of industrial debutanizer and industrial steam volume. In terms of method validation, this paper compares the output data of the proposed CADA by the Spearman correlation coefficient to verify the strategy’s feasibility. Then, the output data of the CADA strategy is fed into the artificial neural network (NN), support vector regression (SVR), and convolutional neural network (CNN) for comparison experiments. The final experimental results show that our proposed CADA-CNN model has lower prediction error and better prediction error distribution.

Improving the predictive response using ensemble empirical mode decomposition based soft sensors with auto encoder deep neural network

In industries, several key quality variables used in complex processes are immeasurable online and are not reliable because of the factors like complex environmental criteria, limited techniques for testing and high cost. Soft sensor technology has become known to solve these complexities. In industrial process, the key factors like redundancy, noise and dynamic features of data affect the accuracy of soft sensors. Thus, a predictive control approach is required which has to integrate improved methods used to detect the control signal that uses direction of structural motion. This paper proposes an innovative Ensemble Empirical Mode Decomposition Based Auto Encoder Deep Neural Network (EEMD-AEDNN) which combines the advantages of Ensemble Empirical Mode Decomposition and neural network bringing problems of mode-mixing from EEMD and false modes from neural network under control. Moreover, dynamic characteristics are captured which are distributed over time improving the modelling effects. The advantage is that noise and redundancy from actual data are removed and information loss is minimized. Furthermore, data are sequential introducing historical data for dynamic modelling. This goal is consistently achieved and thus the proposed model outperforms few standard approaches which are considered like Ensemble Empirical Mode Decomposition based Long Short-Term Memory neural network (EEMD-LSTM), Wavelet Neural Network with Random Time (WNNRT) and Ensemble Empirical Mode Decomposition-General Regression Neural Network (EEMD-GRNN). It is found that the proposed EEMD-AEDNN method achieves 94.22% of accuracy, 84.68% of RMSE, 75.34% of RAE and 54.42% of MAE in 86.5 ms.

Deep Semi-Supervised Just-in-Time Learning Based Soft Sensor for Mooney Viscosity Estimation in Industrial Rubber Mixing Process.

Soft sensor technology has become an effective tool to enable real-time estimations of key quality variables in industrial rubber-mixing processes, which facilitates efficient monitoring and a control of rubber manufacturing. However, it remains a challenging issue to develop high-performance soft sensors due to improper feature selection/extraction and insufficiency of labeled data. Thus, a deep semi-supervised just-in-time learning-based Gaussian process regression (DSSJITGPR) is developed for Mooney viscosity estimation. It integrates just-in-time learning, semi-supervised learning, and deep learning into a unified modeling framework. In the offline stage, the latent feature information behind the historical process data is extracted through a stacked autoencoder. Then, an evolutionary pseudo-labeling estimation approach is applied to extend the labeled modeling database, where high-confidence pseudo-labeled data are obtained by solving an explicit pseudo-labeling optimization problem. In the online stage, when the query sample arrives, a semi-supervised JITGPR model is built from the enlarged modeling database to achieve Mooney viscosity estimation. Compared with traditional Mooney-viscosity soft sensor methods, DSSJITGPR shows significant advantages in extracting latent features and handling label scarcity, thus delivering superior prediction performance. The effectiveness and superiority of DSSJITGPR has been verified through the Mooney viscosity prediction results from an industrial rubber-mixing process.

SVAE-WGAN-Based Soft Sensor Data Supplement Method for Process Industry

Challenges of process industry, which is characterized as hugeness of process variables in complexity of industrial environment, can be tackled effectively by the use of soft sensor technology. However, how to supplement the dataset with effective data supplement method under harsh industrial environment is a key issue for the enhancement of prediction accuracy in soft-sensing model. Aimed at this problem, a SVAE-WGAN based soft sensor data supplement method is proposed for process industry. Firstly, deep features are extracted with the stacking of the variational autoencoder (SVAE). Secondly, a generation model is constructed with the combination of stacked variational autoencoder (SVAE) and Wasserstein generative adversarial network (WGAN). Thirdly, the proposed model is optimized with training of dataset in industrial process. Finally, the proposed model is evaluated with abundant experimental tests in terms of MSE, RMSE and MAE. It is shown in the results that the proposed SVAE-WGAN generation network is significantly better than that of the traditional VAE, GAN and WGAN generation network in case of industrial steam volume dataset. Specially, the proposed method is more effective than the latest reference VA-WGAN generation network in terms of RMSE, which is enhanced about 9.08% at most. Moreover, the prediction precision of soft sensors could be improved via the supplement of the training samples.

System Control Scheme and Design Idea of Soft Sensing Model for Heavy Metal Ion Concentration

This paper combines ion exchange technology and soft sensor technology to design and improve a series of integrated automatic control and module distributed combination of new ion exchange system. At the same time, a visual water management and control system for industrial heavy metal wastewater treatment is developed based on the industrial IoT platform of eForceControl. Through information collection and data integration of heavy metal pollutants with different characteristics in the process flow, a variety of ion exchange materials in the distribution module are optimized. Combined with big data analysis and ion sensing, it realizes online monitoring and treatment of water pollution factors, optimizes production capacity management of industrial system, and is committed to solving problems such as low efficiency of traditional heavy metal wastewater treatment, single data analysis mode and chaotic operation and control.

Sensitivities of Rheological Properties of Magnetoactive Foam for Soft Sensor Technology.

Magnetoactive (MA) foam, with its tunable mechanical properties and magnetostriction, has the potential to be used for the development of soft sensor technology. However, researchers have found that its mechanical properties and magnetostriction are morphologically dependent, thereby limiting its capabilities for dexterous manipulation. Thus, in this work, MA foam was developed with additional capabilities for controlling its magnetostriction, normal force, storage modulus, shear stress and torque by manipulating the concentration of carbonyl iron particles (CIPs) and the magnetic field with regard to morphological changes. MA foams were prepared with three weight percentages of CIPs, namely, 35 wt.%, 55 wt.% and 75 wt.%, and three different modes, namely, zero shear, constant shear and various shears. The results showed that the MA foam with 75 wt.% of CIPs enhanced the normal force sensitivity and positive magnetostriction sensitivity by up to 97% and 85%, respectively. Moreover, the sensitivities of the storage modulus, torque and shear stress were 8.97 Pa/mT, 0.021 µN/mT, and 0.0096 Pa/mT, respectively. Meanwhile, the magnetic dipolar interaction between the CIPs was capable of changing the property of MA foam from a positive to a negative magnetostriction under various shear strains with a low loss of energy. Therefore, it is believed that this kind of highly sensitive MA foam can potentially be implemented in future soft sensor systems.