Design of controllers

October 28, 2005

This press release is also available in Spanish.

In her PhD thesis at the Public University of Navarre, Industrial Engineer Marta Barreras Carracedo put forward a new method of designing controllers based on QFT (Quantitative Feedback Theory) and which facilitates its real implantation in the government of real physical processes. The thesis itself validates the method putting into practice two concrete cases, an industrial electric furnace used for the drying of large pieces of composite and a heat exchanger in a solar water-heating plant.

The PhD is entitled: "Multivariable QFT Robust Control by means of non-diagonal sequential methods. Its application to the government of thermal processes.

Development of sophisticated industrial processes

Over the last decades, the rapid advance of technology and the appearance of increasingly more powerful computers have favoured the development of industrial processes that are more and more sophisticated. These processes include, in general, an ever higher level of automation the aim of which is the enhancement of characteristics and performance. Nevertheless, this brings with it a growing complexity in installations that have accelerated demand for new, advanced control strategies capable of covering new needs.

In this respect, Control Engineering plays a fundamental role within modern industrial processes. Marta Barreras's PhD was aimed at developing a new advanced strategy for the design of controllers of multiple sensor systems and actuators, of a robust nature and with a capacity for internal decoupling and rejection of external perturbations. Moreover, the system has to be sufficiently simple in order to facilitate its implantation into real industrial processes.

To this end, the research work used the Quantitative Feedback Theory, given its capacity to deal with the uncertainty of the system and the existing external perturbations. This theory is based on the use of non-diagonal elements of the controller to deal simultaneously with problems such as coupling reduction and rejection of external perturbations.

Validation with two real processes

After a theoretical analysis, the results were validated in two distinct stages. First through the resolution of a theoretical example and, secondly, by means of its application in two real processes.

Marta Barreras chose an industrial electric furnace used for the baking of large pieces of composite and a heat exchanger located in a solar water-heating plant.

On the one hand, the high multiplicity of actuators and sensors present in the industrial electric furnace means that its operation is of a high multivariable character. Thus, on governing automatically its thermal and dynamic behaviour, there arise significant problems of coupling between existing control loops.

On the other hand, the interaction existing between two loops of a heat exchanger, as with the existence of uncertainty in the model and presence of external perturbations, make this application another good example for valilidating the multivariable QFT strategy.

Marta Barreras points out that, amongst the advantages of the technique applied in these two cases is the act that it manages to enhance the general response of the system to be controlled. Specifically, the rejection of external perturbations was augmented and we managed to reduce the effect of coupling between the loops of the system.
-end-


Elhuyar Fundazioa

Related Cases Articles from Brightsurf:

Multidisciplinary approaches to solving cold cases
Forensic DNA analysis enables new and increasingly sophisticated Technology for solving cold cases.

TB cases and deaths predicted to spike due to COVID-19
Study estimates at least 110,000 additional deaths from TB in China, India and South Africa unless health services maintained and strengthened.

Almost half of all postpartum psychosis are isolated cases
A new research result from iPSYCH shows that 40% of the women who suffer a psychosis after giving birth -- known as postpartum psychosis - do not subsequently become ill again.

54.8% of COVID-19 cases imported to Brazil by March 5 came from Italy
In contrast with China and other countries where the disease spread slowly, in Brazil more than 300 people started the epidemic.

Artificial intelligence could help predict future diabetes cases
A type of artificial intelligence called machine learning can help predict which patients will develop diabetes, according to an ENDO 2020 abstract that will be published in a special supplemental section of the Journal of the Endocrine Society.

Maggot analysis goes molecular for forensic cases
Maggots on a dead body or wound can help pinpoint when a person or animal died, or when maltreatment began in elder, child care or animal neglect cases.

Many lyme disease cases go unreported; A new model could help change that
A new model based on 17 years of data can help predict the trajectory of where Lyme disease will spread.

Cases of poisoning: Liquids containing cannabidiols for e-cigarettes can be manipulated
The health risks of e-cigarettes have come into focus after the deaths of several 'vapers' due to lung injury in the USA recently.

Study looks at distribution of new cases of diabetes, density of specialists
Researchers from Brigham and Women's Hospital analyzed national data on the prevalence of diabetes and the number of internal medicine specialists in each U.S. state.

Age distribution of new obesity-associated cancer cases
This observational study examines changes in the age distribution of new obesity-associated cancer cases and nonobesity-associated cancer cases from 2000 to 2016 by sex and race/ethnicity.

Read More: Cases News and Cases Current Events
Brightsurf.com is a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for sites to earn advertising fees by advertising and linking to Amazon.com.