Multivariate Statistical Process Monitoring and Control: Recent Developments and Applications to Chemical Industry

Multivariate statistical process monitoring and control (MSPM&C) methods for chemical process monitoring with statistical projection techniques such as principal component analysis (PCA) and partial least squares (PLS) are surveyed in this paper. The four-step procedure of performing MSPM&C for chemical process, modeling of processes, detecting abnormal events or faults, identifying the variable(s) responsible for the faults and diagnosing the source cause for the abnormal behavior, is analyzed. Several main research directions of MSPM&C reported in the literature are discussed, such as multi-way principal component analysis (MPCA) for batch process, statistical monitoring and control for nonlinear process, dynamic PCA and dynamic PLS, and on-line quality control by inferential models. Industrial applications of MSPM&C to several typical chemical processes, such as chemical reactor, distillation column, polymerization process, petroleum refinery units, are summarized. Finally, some concluding remarks and future considerations are made.

A HybridManufacturing ProcessMonitoringMethod Using Stacked Gated Recurrent Unit and Random Forest

This study proposed a new real-time manufacturing process monitoring method to monitor and detect process shifts in manufacturing operations.Since real-time production process monitoring is critical in today’s smart manufacturing.The more robust the monitoring model,the more reliable a process is to be under control.In the past,many researchers have developed real-time monitoring methods to detect process shifts early.However,thesemethods have limitations in detecting process shifts as quickly as possible and handling various data volumes and varieties.In this paper,a robust monitoring model combining Gated Recurrent Unit(GRU)and Random Forest(RF)with Real-Time Contrast(RTC)called GRU-RF-RTC was proposed to detect process shifts rapidly.The effectiveness of the proposed GRU-RF-RTC model is first evaluated using multivariate normal and nonnormal distribution datasets.Then,to prove the applicability of the proposed model in a realmanufacturing setting,the model was evaluated using real-world normal and non-normal problems.The results demonstrate that the proposed GRU-RF-RTC outperforms other methods in detecting process shifts quickly with the lowest average out-of-control run length(ARL1)in all synthesis and real-world problems under normal and non-normal cases.The experiment results on real-world problems highlight the significance of the proposed GRU-RF-RTC model in modern manufacturing process monitoring applications.The result reveals that the proposed method improves the shift detection capability by 42.14%in normal and 43.64%in gamma distribution problems.

Integration Between Enterprise Process Monitoring and Controlling System and Enterprise Application

The relationships and the features of integration between Enterprise ProcessMonitoring and Controlling System (EPMCS) and Enterprise Process Related Applications (EPRA) wereanalyzed. An integration architecture centered on EPMCS was presented, in which there were fourlayers to connect from EPMCS to EPRA: EPMCS, application integration layer, transport layer andEPRA, and there were four layers used to etstablish integration: presentation layer, function layer,data layer and system layer. The frameworks to connect EPMCS and EPRA were designed, thatEnterprise-Independent Model (EIM), Enterprise-Specific Model (ESM) and meta-model to describe thesetwo models were defined. The method to integrate data based on XML was designed to exchange datafrom EPMCS to EPRA according to the mapping between EIM and ESM. The approches are suitable forintegrating EPMCS and systems in Product Data Management (PDM), project management and enterprisebusiness management.

Experimental Study of Monitoring and Controlling of Composite Cure Process in Autoclave Featured with Fiber Optic Sensor

With the aid of the latest fiber optic sensing technology parameters in the cure process of ther- mosetting resin-matrix composite, such as temperature, viscosity,void and residual stress, can be monitored entirely and efficiently.In this paper, experiment results of viscosity measurement in composite cure process in autoclave using fiber optic sensors are presented. Based on the sensed information, a computer program is utilized to control the cure process. With this technology, the cure process becomes more apparent and controllable, which will greatly improve the cured products and reduce the cost.

Some Group Runs Based Multivariate Control Charts for Monitoring the Process Mean Vector

In this article, we propose two control charts namely, the “Multivariate Group Runs’ (MV-GR-M)” and the “Multivariate Modified Group Runs’ (MV-MGR-M)” control charts, based on the multivariate normal processes, for monitoring the process mean vector. Methods to obtain the design parameters and operations of these control charts are discussed. Performances of the proposed charts are compared with some existing control charts. It is verified that, the proposed charts give a significant reduction in the out-of-control “Average Time to Signal” (ATS) in the zero state, as well in the steady state compared to the Hotelling’s T2 and the synthetic T2 control charts.

Nonparametric Control Scheme for Monitoring Phase Ⅱ Nonlinear Profiles with Varied Argument Values

Profile monitoring is used to check the stability of the quality of a product over time when the product quality is best represented by a function at each time point.However,most previous monitoring approaches have not considered that the argument values may vary from profile to profile,which is common in practice.A novel nonparametric control scheme based on profile error is proposed for monitoring nonlinear profiles with varied argument values.The proposed scheme uses the metrics of profile error as the statistics to construct the control charts.More details about the design of this nonparametric scheme are also discussed.The monitoring performance of the combined control scheme is compared with that of alternative nonparametric methods via simulation.Simulation studies show that the combined scheme is effective in detecting parameter error and is sensitive to small shifts in the process.In addition,due to the properties of the charting statistics,the out-of-control signal can provide diagnostic information for the users.Finally,the implementation steps of the proposed monitoring scheme are given and applied for monitoring the blade manufacturing process.With the application in blade manufacturing of aircraft engines,the proposed nonparametric control scheme is effective,interpretable,and easy to apply.

Multimodal process monitoring based on transition-constrained Gaussian mixture model

Reliable process monitoring is important for ensuring process safety and product quality.A production process is generally characterized bymultiple operation modes,and monitoring thesemultimodal processes is challenging.Most multimodal monitoring methods rely on the assumption that the modes are independent of each other,which may not be appropriate for practical application.This study proposes a transition-constrained Gaussian mixture model method for efficient multimodal process monitoring.This technique can reduce falsely and frequently occurring mode transitions by considering the time series information in the mode identification of historical and online data.This process enables the identified modes to reflect the stability of actual working conditions,improve mode identification accuracy,and enhance monitoring reliability in cases of mode overlap.Case studies on a numerical simulation example and simulation of the penicillin fermentation process are provided to verify the effectiveness of the proposed approach inmultimodal process monitoring with mode overlap.

Control performance monitoring and advanced control towards energy conservation

High performance control of an interactive process such as iron and steel plant relies on ability to honor safety and operational constraints;reduce the standard deviations of variables that need to be controlled(e.g.product quantity,quality );de-bottlenecking the process;and,maximize profitability or lower cost(e.g.energy savings, improve hot metal content).These objectives may be prioritized in this order,but can vary and are very difficult to achieve optimally through conventional control.A multivariable predictive controller solution,along with its extensive inferential sensor and built-in optimizer,provides online closed loop control and optimization for many interactive metal and mining processes to lower the energy cost,increase throughput,and optimize product quality and yield. Control loop performance is also a key factor to improve iron and steel plant automation and operation result; Honeywell CPM offers vender-independent product which provides monitoring,tuning,modeling of control loop and sustainable loop performance analysis and maintenance solution towards operation stability and energy saving.

Industry 4.0 Application in Manufacturing for Real-Time Monitoring and Control

Modern manufacturing aims to reduce downtime and track process anomalies to make profitable business decisions.This ideology is strengthened by Industry 4.0,which aims to continuously monitor high-value manufacturing assets.This article builds upon the Industry 4.0 concept to improve the efficiency of manufacturing systems.The major contribution is a framework for continuous monitoring and feedback-based control in the friction stir welding(FSW)process.It consists of a CNC manufacturing machine,sensors,edge,cloud systems,and deep neural networks,all working cohesively in real time.The edge device,located near the FSW machine,consists of a neural network that receives sensory information and predicts weld quality in real time.It addresses time-critical manufacturing decisions.Cloud receives the sensory data if weld quality is poor,and a second neural network predicts the new set of welding parameters that are sent as feedback to the welding machine.Several experiments are conducted for training the neural networks.The framework successfully tracks process quality and improves the welding by controlling it in real time.The system enables faster monitoring and control achieved in less than 1 s.The framework is validated through several experiments.

An Integrated Approach for Process Control Valves Diagnosis Using Fuzzy Logic

Control valves are widely used in industry to control fluid flow in several applications. In nuclear power systems they are crucial for the safe operation of plants. Therefore, the necessity of improvements in monitoring and diagnosis methods started to be of extreme relevance, establishing as main goal of the reliability and readiness of the system components. The main focus of this work is to study the development of a model of non-intrusive monitoring and diagnosis applied to process control valves using artificial intelligence by fuzzy logic technique, contributing to the development of predictive methodologies identifying faults in incipient state. Specially in nuclear power plants, the predictive maintenance contributes to the security factor in order to diagnose in advance the occurrence of a possible failure, preventing severs situations. The control valve analyzed belongs to a steam plant which simulates the secondary circuit of a PWR—Pressurized Water Reactor. The maintenance programs are being implemented based on the ability to diagnose modes of degradation and to take measures to prevent incipient failures, improving plant reliability and reducing maintenance costs. The approach described in this paper represents an alternative departure from the conventional qualitative techniques of system analysis. The methodology used in this project is based on signatures analysis, considering the pressure (psi) in the actuator and the stem displacement (mm) of the valve. Once the measurements baseline of the control valve is taken, it is possible to detect long-term deviations during valve lifetime, detecting in advance valve failures. This study makes use of MATLAB language through the “fuzzy logic toolbox” which uses the method of inference “Mamdani”, acting by fuzzy conjunction, through Triangular Norms (t-norm) and Triangular Conorms (t-conorm). The main goal is to obtain more detailed information contained in the measured data, correlating them to failure situations in the incipient stage.

New Method for Multivariate Statistical Process Monitoring

A new method using discriminant analysis and control charts is proposed for monitoring multivariate process operations more reliably.Fisher discriminant analysis （FDA） is used to derive a feature discriminant direction （FDD） between each normal and fault operations,and each FDD thus decided constructs the feature space of each fault operation.Individuals control charts （XmR charts） are used to monitor multivariate processes using the process data projected onto feature spaces.Upper control limit （UCL） and lower control limit （LCL） on each feature space from normal process operation are calculated for XmR charts,and are used to distinguish fault from normal.A variation trend on an XmR chart reveals the type of relevant fault operation.Applications to Tennessee Eastman simulation processes show that this proposed method can result in better monitoring performance than principal component analysis （PCA）-based methods and can better identify step type faults on XmR charts.

Enhancing the Effectiveness of Trimethylchlorosilane Purification Process Monitoring with Variational Autoencoder

In modern industry,process monitoring plays a significant role in improving the quality of process conduct.With the higher dimensional of the industrial data,the monitoring methods based on the latent variables have been widely applied in order to decrease the wasting of the industrial database.Nevertheless,these latent variables do not usually follow the Gaussian distribution and thus perform unsuitable when applying some statistics indices,especially the T^(2) on them.Variational AutoEncoders(VAE),an unsupervised deep learning algorithm using the hierarchy study method,has the ability to make the latent variables follow the Gaussian distribution.The partial least squares(PLS)are used to obtain the information between the dependent variables and independent variables.In this paper,we will integrate these two methods and make a comparison with other methods.The superiority of this proposed method will be verified by the simulation and the Trimethylchlorosilane purification process in terms of the multivariate control charts.

A Review for Model Plant Mismatch Measures in Process Monitoring

Model is usually necessary for the design of a control loop. Due to simplification and unknown dynamics, model plant mismatch is inevitable in the control loop. In process monitoring, detection of mismatch and evaluation of its influences are demanded. In this paper several mismatch measures are presented based on different model descriptions. They are categorized into different groups from different perspectives and their potential in detection and diagnosis is evaluated. Two case studies on mixing process and distillation process demonstrate the efficacy of the framework of mismatch monitoring.

Fuzzy Control Model for Structural Health Monitoring of Civil Infrastructure Systems

This paper presents a Fuzzy Control Model for SHM （Structural Health Monitoring） of civil infrastructure systems. Two important considerations of this model are （a） effective control of structural mechanism to prevent damage of civil infrastructure systems, and （b） energy-efficient data transmissions. Fuzzy Logic is incorporated into the model to provide （a） capability for handling imprecision and non-statistical uncertainty associated with structural monitoring, and （b） framework for effective control of the mechanism of civil infrastructure systems. Moreover, wireless smart sensors are deployed in the model to measure dynamic response of civil infrastructure systems to structural excitation. The operation of these wireless smart sensors is characterized as discounted SMDP （Semi-Markov Decision Process） consisting of two states, namely： sensing/processing and transmitting/receiving. The objective of the SMDP-based measurement scheme is to choose policy that offers optimal energy-efficient transmission of measured value of vibration-based dynamic response. Depending on the net magnitude of measured dynamic responses to excitation signals, data may （or may not） be transmitted to the Fuzzy control segment for appropriate control of the mechanism of civil infrastructure systems. The efficacy of this model is tested via numerical analysis, which is implemented in MATLAB software. It is shown that this model can provide energy-efficient structural health monitoring and effective control of civil infrastructure systems.

A blast furnace fault monitoring algorithm with low false alarm rate:Ensemble of greedy dynamic principal component analysis-Gaussian mixture model

The large blast furnace is essential equipment in the process of iron and steel manufacturing. Due to the complex operation process and frequent fluctuations of variables, conventional monitoring methods often bring false alarms. To address the above problem, an ensemble of greedy dynamic principal component analysis-Gaussian mixture model(EGDPCA-GMM) is proposed in this paper. First, PCA-GMM is introduced to deal with the collinearity and the non-Gaussian distribution of blast furnace data.Second, in order to explain the dynamics of data, the greedy algorithm is used to determine the extended variables and their corresponding time lags, so as to avoid introducing unnecessary noise. Then the bagging ensemble is adopted to cooperate with greedy extension to eliminate the randomness brought by the greedy algorithm and further reduce the false alarm rate(FAR) of monitoring results. Finally, the algorithm is applied to the blast furnace of a large iron and steel group in South China to verify performance.Compared with the basic algorithms, the proposed method achieves lowest FAR, while keeping missed alarm rate(MAR) remain stable.

Fuzzy reasoning model using fuzzy Petri Nets for the monitoring of robotic assembly

This paper presented a fuzzy Petri net model to deal with the monitoring of robotic assembly. Based on the fuzzy Petri net model, an efficient composite reasoning mode was proposed to perform fuzzy reasoning automatiealy. It can determine whether there exists an antecedent-consequence relationship between two contact states. Furthermore, various types of sensor signals can be converted to the same form of real values between zero and one, and the contradiction among large number, high degree of truth and importance of input conditions can be resolved very well by introducing the weight factors and priorities for sensor signals. Finally, a peg- in-the-hole example was given to illustrate the reasonability and feasibility of the proposed model.

Data-Driven Process Monitoring and Fault Tolerant Control in Wind Energy Conversion System with Hydraulic Pitch System

Wind energy is one of the widely applied renewable energies in the world. Wind turbine as the main wind energy converter at present has very complex technical system containing a huge number of components,actuators and sensors. However, despite of the hardware redundancy, sensor faults have often affected the wind turbine normal operation and thus caused energy generation loss. In this paper, aiming at the wind turbine hydraulic pitch system, data-driven design of process monitoring(PM) and diagnosis has been realized in the wind turbine benchmark. Fault tolerant control(FTC) strategies focused on sensor faults have also been presented here, where with the implementation of soft sensor the sensor fault can be handled and the performance of the system is improved. The performance of this method is demonstrated with the wind turbine benchmark provided by Math Works.

Continuous monitoring of residual water content in boiling waterhydrocarbon emulsions during thermomechanical dehydration

Significant waste resources are generated in the form of water-oil emulsions.These emulsions cannot be effectively destroyed on an industrial scale by traditional methods that rely on the settling of the aqueous phase,and therefore,they accumulate in large quantities.Thermomechanical dehydration,based on the evaporation of the water phase,presents a promising process for recycling such waste.However,within the framework of thermomechanical dehydration,the issue of optimizing energy costs for heating raw materials and controlling the water content in the product arises.Standard methods of determining water content under the boiling conditions of highly stable water-hydrocarbon emulsions are characterized by low efficiency,as they require constant sampling and the involvement of additional equipment and personnel.Consequently,this presents a challenge in predicting and creating an automated thermomechanical dehydration process.Therefore,dynamic curves depicting changes in the water content of these emulsions,depending on the temperature of the boiling liquid,have been obtained.It is proposed to determine the rate of temperature increase(dT/dt)of the boiling emulsion for continuous,real-time monitoring of the residual water content and for recording the moment of complete dehydration.Achieving a boiling emulsion temperature of 130-170℃(or higher)and/or the rate of temperature increase from 3.0 to 5.5(or above)indicates the complete dehydration of the emulsion.The proposed method can be implemented in any industrial or laboratory-scale unit for thermomechanical dehydration without significant capital costs.It is based on the use of simple devices consisting of temperature sensors and a computing unit for determining the temperature and rate of heating.

Iterative Learning Model Predictive Control for a Class of Continuous/Batch Processes

An iterative learning model predictive control （ILMPC） technique is applied to a class of continuous/batch processes. Such processes are characterized by the operations of batch processes generating periodic strong disturbances to the continuous processes and traditional regulatory controllers are unable to eliminate these periodic disturbances. ILMPC integrates the feature of iterative learning control （ILC） handling repetitive signal and the flexibility of model predictive control （MPC）. By on-line monitoring the operation status of batch processes, an event-driven iterative learning algorithm for batch repetitive disturbances is initiated and the soft constraints are adjusted timely as the feasible region is away from the desired operating zone. The results of an industrial application show that the proposed ILMPC method is effective for a class of continuous/batch processes.

VPLS Based Quality and Cost Control for Tennessee Eastman Process

Product quality and operation cost control obtain increasing emphases in modern chemical system engineering. To improve the fault detection power of the partial least square (PLS) method for quality control, a new QRPV statistic is proposed in terms of the VP (variable importance in projection) indices of monitored process variables, which is significantly advanced over and different from the conventional Q statistic. QRPV is calculated only by the residuals of the remarkable process variables (RPVs). Therefore, it is the dominant relation between quality and RPV not all process variables (as in the case of the conventional PLS) that is monitored by this new VP-PLS (VPLS) method. The combination of QRPV and T2 statistics is applied to the quality and cost control of the Tennessee Eastman (TE) process, and weak faults can be detected as quickly as possible. Consequently, the product quality of TE process is guaranteed and operation costs are reduced.