Advanced Process Control Strategies Research Articles

Energy efficiency technologies in cement and steel industry

In this paper, Advanced Process Control strategies aimed at energy efficiency achievement and improvement in cement and steel industry are proposed. A flexible and smart control structure constituted by several functional modules and blocks has been developed. The designed control strategy is based on Model Predictive Control techniques, formulated on linear models. Two industrial control solutions have been developed, oriented to energy efficiency and process control improvement in cement industry clinker rotary kilns (clinker production phase) and in steel industry billets reheating furnaces. Tailored customization procedures for the design of ad hoc control systems have been executed, based on the specific needs and specifications of the analysed processes. The installation of the developed controllers on cement and steel plants produced significant benefits in terms of process control which resulted in working closer to the imposed operating limits. With respect to the previous control systems, based on local controllers and/or operators manual conduction, more profitable configurations of the crucial process variables have been provided.

On‐Line Control of Glucose Concentration in High‐Yielding Mammalian Cell Cultures Enabled Through Oxygen Transfer Rate Measurements

Glucose control is vital to ensure consistent growth and protein production in mammalian cell cultures. The typical fed-batch glucose control strategy involving bolus glucose additions based on infrequent off-line daily samples results in cells experiencing significant glucose concentration fluctuations that can influence product quality and growth. This study proposes an on-line method to control and manipulate glucose utilizing readily available process measurements. The method generates a correlation between the cumulative oxygen transfer rate and the cumulative glucose consumed. This correlation generates an on-line prediction of glucose that has been successfully incorporated into a control algorithm manipulating the glucose feed-rate. This advanced process control (APC) strategy enables the glucose concentration to be maintained at an adjustable set-point and has been found to significantly reduce the deviation in glucose concentration in comparison to conventional operation. This method has been validated to produce various therapeutic proteins across cell lines with different glucose consumption demands and is successfully demonstrated on micro (15 mL), laboratory (7 L), and pilot (50 L) scale systems. This novel APC strategy is simple to implement and offers the potential to significantly enhance the glucose control strategy for scales spanning micro-scale systems through to full scale industrial bioreactors.

Aquaponic Integration and Automation – A Critical Evaluation

Aquaponics is technology developed from the aquaculture industry that integrates intensive farming of fish and utilizes plants (integrates hydroponics) in a continuous closed loop to clean the water for the fish. The plants clean the water of nitrate (waste form is initially ammonia) which has been converted into a form that is not toxic to fish by bacteria and is accessible to plants. Hydroponics technology is a technique used to grow plants and vegetables that does not incorporate soil, but nutrients that are dissolved in water and plants are either floated or treated with a nutrient film delivered to the roots by a variety of processes. These technologies are becoming more popular with communities and governments as the preservation of water becomes more of an issue in environments where water is becoming restricted in supply and has to be supplemented by mechanical means such as desalination, which comes at additional cost to consumers. The technologies have had a great amount of research interest to find parametres and ranges that the systems require to function successfully yet also remain productive with certain crop varieties. The successful diversification of crop varieties could increase the viability of commercial aquaponics, which could be achieved with the use of optimized advanced process control strategies.

Resolving critical dimension drift over time in plasma etching through virtual metrology based wafer-to-wafer control

As semiconductor devices are scaled down to sub-20 nm, process window of plasma etching gets extremely small so that process drift or shift becomes more significant. This study addresses one of typical process drift issues caused by consumable parts erosion over time and provides feasible solution by using virtual metrology (VM) based wafer-to-wafer control. Since erosion of a shower head has center-to-edge area dependency, critical dimensions (CDs) at the wafer center and edge area get reversed over time. That CD trend is successfully estimated on a wafer-to-wafer basis by a partial least square (PLS) model which combines variables from optical emission spectroscopy (OES), VI-probe and equipment state gauges. R2 of the PLS model reaches 0.89 and its prediction performance is confirmed in a mass production line. As a result, the model can be exploited as a VM for wafer-to-wafer control. With the VM, advanced process control (APC) strategy is implemented to solve the CD drift. Three σ of CD across wafer is improved from the range (1.3–2.9 nm) to the range (0.79–1.7 nm). Hopefully, results introduced in this paper will contribute to accelerating implementation of VM based APC strategy in semiconductor industry.

An EKF observer to estimate semi-autogenous grinding mill hold-ups

A non-linear observer model of a semi-autogenous grinding mill is developed. The observer model distinguishes between the volumetric hold-up of water, solids, and the grinding media in the mill. Solids refer to all ore small enough to discharge through the end-discharge grate, and grinding media refers to the rocks and steel balls. The rocks are all ore too large to discharge from the mill. The observer model uses the accumulation rate of solids and the mill's discharge rate as parameters. It is shown that with mill discharge flow-rate, discharge density, and volumetric hold-up measurements, the model states and parameters are linearly observable. Although instrumentation at the mill discharge is not yet included in industrial circuits because of space restrictions, this study motivates the benefits to be gained from including such instrumentation. An extended Kalman filter is applied in simulation to estimate the model states and parameters from data generated by a semi-autogenous mill simulation model from literature. Results indicate that if sufficiently accurate measurements are available, especially at the discharge of the mill, it is possible to reliably estimate grinding media, solids and water hold-ups within the mill. Such an observer can be used as part of an advanced process control strategy.

Single screw extrusion control: A comprehensive review and directions for improvements

Polymer extrusion is usually a complex process, particularly due to the coupled nature of process parameters, and hence highly prone to fluctuations. Although a number of different approaches have been attempted in research/industry over the last few decades for extrusion control, it is still experiencing some problems in achieving consistent product quality. Presently, most of the polymer processing extruders are equipped with PID controllers mainly for the control of the screw speed and barrel temperatures in their set limits. It seems that only both of these controllers are commonly used as the major aids of process control to achieve the required melt quality. Although, the quality of the melt output (i.e., a thermally homogeneous melt output which is constant in quantity and quality over the time) is the key variable in polymer extrusion, only a few control techniques are available which make control decisions by observing the actual melt flow quality. Therefore, the development of new control strategies which consider the actual melt quality, perhaps incorporating industrially popular nonlinear techniques such as artificial intelligence, should be highly valuable. In this work, a critical evaluation is made on the state-of-the-art of the previous control approaches in polymer extrusion in industry and research while identifying their limitations. Then, some of the possible directions for future research and also to develop an advanced process control strategy are presented by eliminating a few of the existing limitations.

Improved steady-state models for chlorine dioxide delignification sequences that include washer carryover effects

In previous studies, generalized steady-state models were proposed to approximate the chlorine dioxide demand needed for the delignification of softwood and hardwood pulps, where the kappa number entering the bleach plant can fluctuate. However, these expressions neglect the effect of dissolved solids with the stock that originate from incomplete pulp washing. In this study, the original elemental chlorine-free (ECF) models are modified to include the effects of carryover from brownstock or post-oxygen washing. The stoichiometric bleach consumption from carryover, based on its composition, was calculated from various literature sources. The majority of the bleach demand (about 70%) results from the dissolved lignin contained in the brownstock carryover, with the remainder resulting from the inorganic sulfur constituents (e.g., sulfide and thiosulfate). When the effect of brownstock carryover was taken into account, the modified models accurately predicted the amount of chlorine dioxide consumed for a given delignification level (about ±0.1% chlorine dioxide) vs. actual bleach plant data. The improved models can be used to gauge the level of washer carryover entering the bleach plant if this parameter is not regularly monitored by the mill. Additionally, these modified expressions could be integrated into advanced process control strategies for ECF bleaching where the washer carryover or dissolved lignin entrainment is measured with online sensors.

Experimental methods for screening parameters influencing the growth to product yield (Y<sub>(x/CH4)</sub>) of a biological methane production (BMP) process performed with <em>Methanothermobacter marburgensis</em>

New generation biofuels are a suitable approach to produce energy carriers in an almost CO2 neutral way. A promising reaction is the conversion of carbon dioxide (CO2) and molecular hydrogen (H2) to methane (CH4) and water (H2O). In this contribution, this so-called Sabatier reaction was performed biologically by using hydrogenotrophic and autotrophic methanogenic microorganisms from the archaea life domain. For the development of a biological methane production (BMP) process, one key parameter is the ratio of biomass production rate (rx) to methane evolution rate (MER) reflected in the growth to product yield (Y(x/CH4)) because it represents both a physiological and a scalable entity for the bioprocesses development as it quantify the selectivity of reaction with respect to the carbon. Y(x/CH4) needs also to be held constant in order to establish an adaptable media composition for developing a scalable feeding strategy. Identification of parameters and quantification of their impact on Y(x/CH4) is a necessary prerequisite for obtaining a growth kinetic model and developing advanced process control strategies especially for dynamic operation modes. In this work, process conditions and parameters impacting Y(x/CH4) were investigated by using a combination of multivariate and univariate chemostat cultures, as well as dynamic experiments. The proposed combination of methods is a novel modular approach for the development of BMP processes. It allowed determining the effects of multiple process factors on physiology and methane productivity of Methanothermobacter marburgensis. In fact, quantitative analysis of basal medium, sulphide and ammonium dilution rates, as well as the ammonium concentration revealed that all these variables vary rx without affecting MER. Hence Y(x/CH4) can be used to identify limiting or

Informative Property of the Data Set in a Single-input Single-output (SISO) Closed-loop System with a Switching Controller

Closed-loop identification is important and necessary to various model-based advanced process control strategies, whose performance depends greatly on the informative property of the data set. Switching control is an important method in process control. Therefore, this paper studies the informative property of a data set in a single-input single-output (SISO) closed-loop system with a switching controller. It is proved that this data set is informative if the controller switches among at least two modes (i.e., feedback laws). Our result does not require any assumption on the way of switch and removes the constraints on the switching manner required in some classical literature. Finally, simulation case studies based on a continuous stirred-tank reactor (CSTR) process are given to validate the results.

PLC-based Automatic Control of Food Production Line

Standard programmable logic controller (PLC) based proportional-integral-derivative (PID) was originally employed as part of the food production line control strategy, but after observing the response of those process due to measured disturbances during normal operation, it was evident that PID control could not meet the desired cooking specifications. It was decided to pursue advanced process control strategies as a means to meet the food production specifications. The problem of this research study was to design and analyze the performance of a PLC-based model state feedback controller implementation for an industrial food production line, and to determine its viability in comparison to commercially available PC-based model predictive controller implementations applied to food production line.

Post-Optimality Approach to Prevent Cycling in Linear MPC Target Calculation

Model predictive control (MPC) is an advanced process control strategy which is widely applied in many industries and it is often implemented in two levels; steady-state target calculation and dynamic optimization. Target calculation level computes the optimal steady-state targets of the process and sends them to dynamic optimization for implementation. The existence of uncertainty or variation in the parameters of the target calculation layer can lead to unstable or cyclic targets which can significatively affect the overall performance of the controller. Many methods have been proposed to deal with model uncertainty using robust optimization. In this study, a new approach using post-optimality analysis is proposed to study the effect of uncertainty or variation in problem parameters on the optimal solution of linear target calculation. This approach can compute the stability limits, for simultaneous variations in objective function coefficients or process limitations, before the optimal target or basis are changed. It aims to improve the closed-loop robustness of MPC by assessing the required accuracy of parameter estimates and variation limits before the calculated target start cycling. The proposed approach offers a design-phase test for targets stability.

Final Comment by the Authors

Model predictive control (MPC) is an advanced process control strategy which is widely applied in many industries and it is often implemented in two levels; steady-state target calculation and dynamic optimization. Target calculation level computes the optimal steady-state targets of the process and sends them to dynamic optimization for implementation. The existence of uncertainty or variation in the parameters of the target calculation layer can lead to unstable or cyclic targets which can significatively affect the overall performance of the controller. Many methods have been proposed to deal with model uncertainty using robust optimization. In this study, a new approach using post-optimality analysis is proposed to study the effect of uncertainty or variation in problem parameters on the optimal solution of linear target calculation. This approach can compute the stability limits, for simultaneous variations in objective function coefficients or process limitations, before the optimal target or basis are changed. It aims to improve the closed-loop robustness of MPC by assessing the required accuracy of parameter estimates and variation limits before the calculated target start cycling. The proposed approach offers a design-phase test for targets stability.

Neural Network Modeling for Advanced Process Control Using Production Data

The fabrication of integrated circuits involves many unit processes, some linear and some non-linear, and each with multiple inputs and outputs. These complexities suggest that benefits could be derived from the development and implementation of advanced process control tools and strategies. Empirical process models are one of these tools. In this research, sequential neural network models are developed to characterize critical steps in a fabrication process. The data used was collected from an industrial process. The data comes from experiments related to the processes under investigation, but not systematically designed to generate data for modeling. The models performed well, with an average prediction error of 3.3%, demonstrating the flexibility of the sequential neural network modeling process. Additionally, the models are used in a sensitivity analysis to study the output response to the various inputs. The methodologies presented are currently being ported to a similar manufacturing process with a larger database. Future work includes using the models for process optimization and as part of a model-based supervisory control system.

Extremely Fast Catalyst Temperature Pulsing: Design of a Prototype Reactor

Abstract Abstract This paper discusses a novel principle of advanced process control strategy: the extremely fast and local pulsing of temperature. This strategy leads to some interesting potential applications, but there are no devices implementing it available yet. One such device currently under construction by the authors is introduced in this paper. It operates by converting electrical energy into heat by forcing a very high current through a very thin resistive element, which also acts as the catalyst for heterogeneous reactions. A design procedure for the key parameters is developed and a simulation of heat distribution in the design under construction is presented. The simulation shows that it should be possible to get local temperature peaks of 500 K which exist for only about 20 μs.

Performance Assessment of Model Pedictive Control for Variability and Constraint Tuning

Multivariate controller performance assessment (MVPA) has been developed over the last several years, but its application in advanced model predictive control (MPC) has been limited mainly due to issues associated with comparability of the variance control objective and that of MPC applications. MPC has been proven as one of the most effective advanced process control (APC) strategies to deal with multivariable constrained control problems with an ultimate objective toward economic optimization. Any attempt to evaluate MPC performance should therefore consider constraints and economic performance. In this work, we show that the variance based performance assessment may be transferred to performance assessment of MPC applications. The MPC economic performance can be evaluated by solving benefit potentials through either variability reduction of quality output variables or tuning of constraints. Algorithms for MPC performance assessment and constraint/variance tuning guidelines are developed through linear matrix inequalities (LMIs) using routine operating process data plus the process steady-state gain matrix. The proposed approach for MPC economic performance evaluation is illustrated and verified via a simulation example of an MPC application as well as a pilot-scale experiment.

POST-OPTIMALITY ANALYSIS OF STEADY-STATE LINEAR TARGET CALCULATION IN MODEL PREDICTIVE CONTROL

Model predictive control (MPC) is an advanced process control strategy that is usually separated into two levels; steady-state target calculation and dynamic optimization. The existence of uncertainty in model parameters of target calculation can significantally affect the overall performance of the controller. Methods have been proposed to deal with model uncertainty using robust optimization. In this study, a new approach using post-optimality analysis is proposed to study the effect of uncertainty or variation in model parameters on the optimal solution of linear target calculation. This approach can compute the stability limits, for simultaneous variation in objective function coefficients or process limitations, before the optimal target or basis are changed.

ASSESSMENT OF ECONOMIC PERFORMANCE OF MODEL PREDICTIVE CONTROL THROUGH VARIANCE/CONSTRAINT TUNING

Multivariate controller performance assessment (MVPA) has been developed over the last several years, but its application in advanced model predictive control (MPC) has been very limited mainly due to issues associated with comparability of variance control objective and that of MPC. MPC has been proven as one of the most effective advanced process control (APC) strategies to deal with multivariable constrained control problems with an ultimate objective towards economic optimization. Any attempt to evaluate MPC performance should therefore consider constraints and economic performance. This work is to establish a link between variance control and MPC in terms of economic performance. We show that the variance based performance assessment may be transfered to economic assessment of MPC. Algorithms for economic performance assessment and tuning are developed through linear matrix inequalities using routine operating process data. The proposed algorithms are illustrated via an industrial MPC application example.

CAPTURING AND SUSTAINING THE BENEFITS OF ADVANCED PROCESS CONTROL

The paper describes the application of an advanced process control (APC) strategy to a complex Fluid Catalytic Cracking, two Vacuum Distillation Units and a Crude Distillation Unit located at Galp Energia Sines Refinery (Portugal). The paper describes the strategy and methodology adopted in the projects, specifically focusing on the best practices related to the implementation and maintenance of large multivariable controllers. The technologies, tools and procedures adopted to keep the applications at their top performance are described in detail. The benefits are shown based on data of rigorous post-audits.

Data requirements and communication issues for advanced process control

Data streams and communication issues are the most critical areas for successful advanced process control (APC) programs. These areas are vital for both APC run-to-run controllers and fault detection and classification (FDC) systems used for high volume manufacturing applications in the semiconductor industry. All APC systems rely on data streams to make their process changes, to keep the process on target and in control, and to otherwise signal a need for engineering involvement to make similar corrective actions. The access to, communication of, and reliability and integrity of these data streams are essential to all APC programs. APC run-to-run controllers use the data to make changes in the process. FDC systems focus on predicting pending equipment- or process- related problems or detecting them quickly when they occur. The inability to access the needed data stream can prohibit the use of APC run-to-run controllers or FDC systems on critical process operations. Worse yet, the use of unreliable or corrupted data can cause undesirable consequences. In order to better capitalize on the improvements demonstrated with APC run-to-run controllers and FDC systems, end users have often had to create their own communication and data processing methods. The first decade of the 21st century will place increased demands on process and metrology equipment manufacturers, APC software and hardware suppliers, and APC programmers. Improvements in these areas through the use of industry standards and best known methods could greatly accelerate the APC field. Wafer-to-wafer and within wafer process control could be essential for 300 mm wafer processing; large flat panel processing will also need these improvements. We will discuss examples that Advanced Micro Devices experienced in Fast atom beam-25 within the past year. The case studies relate to complex, but necessary, methods to get the data we need for a FDC system and the role of metrology data on APC run-to-run controllers. Data and communication requirements for the next three to five years will also be discussed. The increased demands on the current process and metrology systems will increase as we begin to use new and alternative technologies to support more advanced APC strategies.

State Profile Estimation for a Nonisothermal Adsorption Process by a Nonlinear Distributed Parameter Observer

A major dynamical characteristic of pressure swing adsorption plants are nonlinear wave propagation phenomena travelling forwards and backwards within the fixed beds due to the change in flow direction. The control of product purity by an advanced process control strategy requires knowledge of the spatial concentration profiles. In this paper, the reconstruction of the spatial profiles of a single fixed bed of an oxygen production plant with a nonlinear distributed parameter observer is examined as a contribution to the design and the realization of a model based process control strategy. Assuming a binary mixture, the distributed parameter model describing the nonisothermal adsorption process comprises six space and time dependent states: temperature, pressure, oxygen mole fraction in the gaseous phase, the adsorbed amounts of oxygen and nitrogen, and the molar flux. The location of sensors is examined and discussed with respect to observer convergence and a nonlinear distributed parameter observer is designed for the adsorption process injecting correction terms in selected balance equations. The placement of the injected correction terms as well as their tuning parameters are of major importance concerning the convergence properties of the observer. The observer is implemented with the simulation environment DIVA and examined with respect to its convergence behavior.