Robust process control

Specialized equipment for industrial measurements and automatic control have been developed (18) (see Process control). In general, the pH of an industrial process need not be controlled with great accuracy. Consequendy, frequent standardization of the cell assembly may be uimecessary. On the other hand, the ambient conditions, eg, temperature and humidity, under which the industrial control measurements are made, may be such that the pH meter must be much more robust than those intended for laboratory use. To avoid costiy downtime for repairs, pH instmments may be constmcted of modular units, permitting rapid removal and replacement of a defective subssembly. 

M. Morari and E. Zafiriou, Robust Process Control, Prentice-HaH, Inc., Englewood Chffs, N.J., 1989. 

Successful management of a central HTS operation requires a discipline of process control that extends beyond the automation systems. The progression of HTS projects must be carefully guided and tracked through several milestones designed to ensure that a steady and predictable stream of robust assays are available for transfer to the automation platforms and that sufficient bulk reagents and consumable supplies will arrive in time to support the HTS campaigns. 

In the field of in-process analysis, analytical NMR applications also constitute a growth area - and also in relation to additives. This stems from the fact that the method makes it possible to use chemical analytical data in polymer quality control. Robust tools for hostile chemical plant environments are now available. The field of process analytical chemistry has been pushed to the forefront of the partnership between industry and academia. 

Despite all these changes in hardware, the basic concepts of control system structure and control algorithms (types of controllers) remain essentially the same as they were thirty years ago. It is now easier to implement control structures wc just reprogram a computer. But the process control engineers joh is the same come up with a control system that will give good, stable, robust control. 

Chapter 16 covers the analysis of multivariable processes stability, robustness, performance. Chapter 17 presents a practical procedure for designing conventional multiloop SISO controllers (the diagonal control structure) and briefly discusses some of the full-blown multivariable controller structures that have been developed in recent years. 

If the uncertainty has a known structure, structured singular values are used. In the book by Morari and Zafiriou (Robust Process Control, 1989, Prentice-Hall) this topic is discussed in detail. 

P.M. Frank and X. Ding. Survey of robust residual generation and evaluation methods in observer-based fault detection systems. J. Process Control, 7(6) 403-424, 1997. 

In some manufacturing process analysis applications the analyte requires sample preparation (dilution, derivatization, etc.) to afford a suitable analytical method. Derivatization, emission enhancement, and other extrinsic fluorescent approaches described previously are examples of such methods. On-line methods, in particular those requiring chemical reaction, are often reserved for unique cases where other PAT techniques (e.g., UV-vis, NIR, etc.) are insufficient (e.g., very low concentrations) and real-time process control is imperative. That is, there are several complexities to address with these types of on-line solutions to realize a robust process analysis method such as post reaction cleanup, filtering of reaction byproducts, etc. Nevertheless, real-time sample preparation is achieved via an on-line sample conditioning system. These systems can also address harsh process stream conditions (flow, pressure, temperature, etc.) that are either not appropriate for the desired measurement accuracy or precision or the mechanical limitations of the inline insertion probe or flow cell. This section summarizes some of the common LIF monitoring applications across various sectors. 

Robustness DOEs process control under nominal operation... 

A comprehensive framework of robust feedback control of combustion instabilities in propulsion systems has been established. The model appears to be the most complete of its kind to date, and accommodates various unique phenomena commonly observed in practical combustion devices. Several important aspects of distributed control process (including time delay, plant disturbance, sensor noise, model uncertainty, and performance specification) are treated systematically, with emphasis placed on the optimization of control robustness and system performance. In addition, a robust observer is established to estimate the instantaneous plant dynamics and consequently to determine control gains. Implementation of the controller in a generic dump combustor has been successfully demonstrated. 

Multiplexed diode-laser sensors were applied for measurement and control of gas temperature and species concentrations in a large-scale (50-kilowatt) forced-vortex combustor at NAWC to prove the viability of the techniques and the robustness of the equipment for realistic combustion and process-control applications [11]. The scheme employed was similar to that for measurements and control in the forced combustor and for fast extractive sampling of exhaust gases above a flat-flame burner at Stanford University (described previously). 

Synchrotron beamlines are a complex hybrid of hardware and software. Although current designs have achieved a level of robustness inconceivable a decade ago, tight process control is essential. For example, at SGX-CAT the position of the beam is controlled to within 0.5 jxradians (0.000028°). This tolerance corresponds to keeping the X-ray beam centroid within a 25 xm diameter circle at a location 50 m from the undulator source. This performance, reflecting the combined capabilities of the synchrotron and the beamline, is impressive to say the least. 

Even once a method is standardized, erroneous results can still be generated. As a result, it is critical to have robust quality control procedures in place. Here, careful attention should be paid to identify opportunity for in-process control measures such as internal standards, calibration, control plates, replicates and so on as opposed to post-processing data review steps. Inline QC approaches allow sources of error to be identified and remedied much more rapidly and help limit costly re-tests, or the possibility of erroneous data leaving the laboratory. 

In pharmaceutical technology, quality assurance of the pharmaceutical formulation is important. When a pharmaceutical formulation is produced, on-line quality monitoring and control has to be performed in order to check the quality of the outgoing products. Methodology to perform this task is Statistical Process Control (SPC) and is not included in this book. Good text books in the area of SPC exists [6-9]. In this book the focus is on off-line quality control, e.g. how to make products that are intrinsic robust against process variations. 

In recent years, Raman spectroscopy has undergone a major transformation from a specialist laboratory technique to a practical analytical tool. This change was driven on several parallel fronts by dramatic advances in laser instrumentation, detectors, spectrometers, and optical filter technology. This resulted in the advent of a new generation of compact and robust Raman instruments with improved sensitivity and flexibility. These devices could be operated for the first time by non-specialists outside the laboratory environment. Indeed, Raman spectroscopy is now found in the chemical and pharmaceutical industries for process control and has very recently been introduced into hospitals. Handheld instruments are used in forensic and other security applications and battery-operated versions for field use are found in environmental and geological studies. 

A bioprocess system has been monitored using a multi-analyzer system with the multivariate data used to model the process.27 The fed-batch E. coli bioprocess was monitored using an electronic nose, NIR, HPLC and quadrupole mass spectrometer in addition to the standard univariate probes such as a pH, temperature and dissolved oxygen electrode. The output of the various analyzers was used to develop a multivariate statistical process control (SPC) model for use on-line. The robustness and suitability of multivariate SPC were demonstrated with a tryptophan fermentation. 

---