E-News February 2017 Edition

Control Engineering is Vitally Important

Control engineering plays a large part in providing safe and reliable operation of any process or industrial system. It is the actions of the control loop that determine the actuator changes and thereby the stability of the system, the stress in components, and behavior in the presence of disturbances, or even faults. A control system is part of an electro-mechanical system or chemical process which cannot be seen, and it is usually a very small proportion of the plant costs, yet it has a disproportionate effect on the performance and reliability of the system.

The design of control loops for industrial applications is therefore importance even if this is not obvious to engineers out with the controls community. For example, the plant may be built with oversize pumps that result in large changes in flow for small changes in valve position, and the layout/arrangement of plant vessels, piping, and processes can be quite poor leading to a dynamical system that is very difficult to control. However, a control system can often make up for such deficiencies providing good performance in a bad situation. This is not true in reverse. If a control system is poor it is very unlikely that changing the electro-mechanical devices or types of process tank will fix the problem.

Despite the great virtues of feedback control there remain challenges to the control loop designer mainly due to nonlinearities, uncertainties and complexity. Major advances have been made over the last two decades in providing practical solutions to model based nonlinear control problems. Nonlinearities can cause very poor plant responses in certain operating conditions and severely limit the performance and quality of product that is achievable. The time to tune a nonlinear system can be excessive, which is costly in the time involved and the loss of on-specification product. They can result in unpredictable and unsafe behavior. Early control design methods focused on linear systems behavior, and these methods are adequate for many industrial control loops that can be detuned when nonlinearities are troublesome. More advanced control methods exploit model knowledge to provide higher quality and performance. Most of the advanced ACTC courses provided at company premises are now directed at nonlinear control problems.

Mike Grimble


Introduction to Process Control Training Course

ISC is pleased to announce that registration is now open for the Introduction to Process Control Training Course being held on 14th March 2017 in Glasgow City Centre.

This basic level course is aimed at technicians, operators and new graduates working in a process control environment. The course combines computer based learning together with some lecture based learning to reinforce the ideas through real-life examples.

All presented topics will be supported by practical engineering examples and will arm trainees with efficient approach to tackle real-life problems. Lectures and Hands-On sessions will provide a methodology and step-by-step guide of using all presented algorithms in their engineering practice.


New Robust Control Course

A new course on robust control is being developed with Professor Mario Garcia-Sanz of Case Western Reserve University. Professor Garcia-Sanz is a leading world expert on the Quantitative Feedback Theory (QFT) robust control design approach. He has developed specialist software to make the job of using this method much simpler and straightforward.

The QFT approach involves design procedures, which are similar to classical frequency domain design. They are therefore very easy to use by those trained in classical design methods, and they provide a truly robust control solution. One of the great advantages is that the order of the controller is specified by the designer and is just generally much lower order than given by optimal or predictive control methods.

The ACTC training activity can provide a bespoke course, which combine QFT with other robust design methods, such as H? robust control. In such cases the course could be 3days long. Alternatively, only the QFT design method can be presented in say a 1.5day course. The hands-on application examples include multivariable problems and there is a demonstration of the different features using the QFT design package.

If you would like further details please contact Dr Meghan McGookin.

American Control Conference in Seattle

ISC is pleased that it will have Professor Mike Grimble attending the American Control Conference in Seattle to present a paper entitled 'Total Engine Optimization and Control for SI Engines Using Linear Parameter-Varying Models'. This reflects the cooperation with General Motors on the use of advanced controls in powertrain and particularly considers the use of hierarchical control. That is, using predictive control at both supervisory and lower levels. Professor Mike J Grimble will always be happy to discuss this work on new control algorithms at the event or arrange a meeting if required.


Recent ISC Training Courses

Ford, Detroit Training course
Mike and Dr Pawel Majecki had successfully delivered another two training courses to Ford, Detroit back in October 2016. Both courses were spread over 4 days where the morning sessions focused on beginner's level (i.e. 'Control Fundamentals') and the afternoon sessions on advanced level (i.e. 'Introduction to Estimation and Robust Control'). This particular training structure works really well with Ford's engineers as they still have half a day to carry out their mandate routine.

EDF, Barnwood
This 4-day training course in control engineering practices for the nuclear industry is another exceptional example of ISC's bespoke courses capability and was designed to address various interests within a multidisciplinary Control & Instrumentation group of engineers at EDF. The first run of the course was successfully delivered by Dr. Andy Clegg, Petros Savvidis and other tutors by EDF in September 2016. It covered both the theoretical as well as the practical aspects of the key control systems of the plant. Learning was reinforced through interactive emulators developed in LabVIEW and case studies in which the delegates were encouraged to work as groups to provide solutions to real world examples. The course was very well received and attendees reported gaining valuable information, usable in their everyday tasks.

Control Fundamentals, Glasgow
Our 3-days Control Fundamental held in Glasgow had received very positive feedback and the delegates would recommend our training course to engineers who require a basic understanding to control engineering. Our next Control Fundamentals is scheduled from 6th to 8th June 2017.

Introduction to Estimation and Kalman Filtering
The agenda of this one-day training course has been altered to suit the delegates' requirement compared to our standard training course. ISC Ltd is happy to host a training course that based on our clients' needs. Please get in touch with us if you require more information.

Introduction to Pharmaceutical Process Control, ADDoPT Program, Glasgow
ISC Ltd has successfully delivered another 2-days training course in Glasgow as part of the ADDoPT Program with Perceptive Engineering. This training course will be delivered at various locations throughout UK to suit the other ADDoPT members.

Control Fundamentals and System Identification, OHB, Germany
This 3-days training course is tailored to OHB requirements and delivered by our managing director, Dr Andy Clegg in Oberpfaffenhofen. The selected topics were based on their ongoing projects and knowledge transfers within the group. The training course is very well-received and group discussions were encouraged throughout the course.

If you would like further details on any of these courses please contact Dr Meghan McGookin.


The Secrets of Model Based Control

The most popular controller in industry is of course the PID controller and most of the tuning methods do not require a model of the system be known. This is in fact a great advantage of the basic PID algorithm, since models can be expensive to construct and require a level of expertise, which may not be available. A good question is therefore why is model based control becoming so popular.

Most of the model based advanced control algorithms involve optimal control or optimization in some way and this includes many of the predictive control laws. The use of a cost function to judge performance provides a useful benchmark for comparison, and it provides tuning inputs, which have a physical meaning. However, despite all the advantages of using optimal methods there are not many occasions when the cost-function truly actually reflects the needs of the actual system. The cost-function is therefore used as the way to incorporate design information, rather than a physically motivated criterion to be optimized. This is not therefore the reason model based control has been so popular.

Models are of course required for modelling physical disturbances such as wind speed variations in wind turbines and wave disturbances in marine systems. The prediction equations required in predictive control do of course require reasonable models to provide the predictive capability. Thus, although models are essential for representing the system behaviour adequately, the benefits of model based control are still not obvious.

The main benefit experienced is rather indirect and not so obvious. In fact, it is the understanding that is gained through building models of the system, and the understanding of behaviour and performance, which is achieved. The very act of building models and producing a simulation adds to the engineers understanding of how to design a control a system. It reveals where the major disturbances, interactions, nonlinearities and problems are likely to arise. It is this rather hidden property of model-based control, which has made it so successful and why when high performance is essential that this is the approach which is recommended.


Economic Model Predictive Control

The subject of predictive control has now become so popular that specialist areas are emerging dealing with particular aspects of problems in model predictive control (MPC) systems. One of the most interesting is the subject of economic model predictive control. In this case, the emphasis is on optimisation for economic benefits. In traditional model predictive control for process plants, the objective is often to maintain the process variables around desired steady-state values. Economic MPC aims to optimize process operations taking account of the time-varying nature of the problem. The optimal operation might not be at an equilibrium point, in which case the use of linearized models is inappropriate. Even if the equilibrium is optimal, the economic benefits might stem from the transient performance.

Whilst it is true that most predictive controls involve the minimisation of a cost function, it is often the case that the cost function does not really have a physical significance. There are some problems where the energy consumed in the system can be described in terms of a relatively simple cost-function. This applies for some autopilot problems in marine systems for example. However, in the majority of cases the cost-function is only indirectly related to what is to really to be optimised. The work in economic model predictive control is more closely aligned with the economic needs of the process.

Another category of predictive control problems is that of robust predictive control. A well-designed predictive control system will often have a natural robustness but some electromechanical systems pose real problems due to uncertainties and nonlinearities. Robust predictive control methods are tailored for this type of application, but it remains one of the most challenging areas, which will probably be the subject of research for years to come.


Automotive Control

Most of the major automotive manufacturers are investigating the use of nonlinear model predictive control for automotive systems. There are thought to be many advantages of model based control methods, including speeding up the design process for new engines or automotive systems. The main difficultly to be addressed is the intensive computations that are often required and the need to have good tuning procedures. There is great potential for some of the more recent nonlinear model based predictive control methods.

To generate an algorithm which is suitable for automotive applications it should be possible to use very general powertrain control models and general optimal control cost-functions which can be tailored to different automotive application areas. The Linear Parameter Varying (LPV) models enable very general nonlinear engine models to be approximated. In fact black box nonlinear input subsystems can also be included with LPV, state-dependant or quasi-LPV output subsystems. These can include known and unknown disturbances and stochastic or deterministic reference signals.

Similarly for the optimal control cost indices dynamic cost-function weights can be used which can also have a very general representation using for example LPV or nonlinear operator models. The family of nonlinear predictive controllers available range from being very basic and simple to those that are more advanced but are more computationally intensive. The full range of options have been made available in a simple to use Matlab / Simulink design facility, developed by ISC, which has been used on research projects or ACTC training courses for most of the major US automotive manufacturers.


Dr David Fraser Joins ISC

In 2016, ISC appointed Dr David Fraser as Non-Executive Chair. David has a background in control engineering and simulation and a senior career in programme management and business leadership in a range of industries. In his own business, David works with leadership teams in organisations assisting them in being fully effective and enterprising, drawing on both his industry background and his expertise in transformational leadership. He is the author of two books in this area. Commenting after the Company's recent AGM, David said: "It's great to see ISC delivering well for its customers and, at the same time, making an appropriate return. I look forward to supporting the team with the projects the Company undertakes to solve challenging issues for its customers, and in developing its capability for the future. I have been particularly impressed with the team's ability to model physical systems and arrive quickly at a deep understanding of clients' control problems."


Books on Mathematical Foundations

Two interesting texts have been received from SIAM (Society for Industrial and Applied Mathematics) on mathematical fundamentals and importance in engineering. The first is by Rene Carmona and is titled 'Lectures on BSDEs, Stochastic Control, and Stochastic Differential Gains with Financial Applications'. The second is by Thomas Trogdon and Sheehan Olver entitled 'Reeman-Hilbery Problems, the Numerical Solution and the computation of Nonlinear Special Functions'. It is recognized that these topics are rather specialist but there may be some research engineers or engineering students that will find them helpful in specialist research activities.


Book Review: Power System Dynamics and Control, by: Harry G Kwatny and Karen Miu-Miller, Published By: Birkhauser, Cost: 56

There is a real need for text on the construction of models for power systems. This monograph covers static and dynamic network stability and an introduction to power system optimal control with reliability constraints. Classical control problems such as voltage regulation and load-frequency control are covered and the coordination of economic dispatch with load-frequency control.

The text begins by covering the basics of electricity and magnetism, and it includes electric circuits and devices. Basic ideas in the behavior of synchronous generators and power flows are described. This provides the starting point for the main power systems modelling and control chapters. The real value of the book is really in chapter 5 onwards. This is likely to be the most valuable material for a control engineer working in industry. It is very nicely written and well-illustrated using figures and diagrams.

This is a basic text on the subject but it is a welcome addition to the bookshelf and the hard copy cost is quite reasonable.

Book Nerd.


Book Review: Digital Control Applications Illustrated with Matlab, by: Hemchandra Madhusudan and Shertukde , Published By: CRC Press, Cost: 84

This is a valuable text for engineers that use Matlab and for researchers in advanced control. It includes a nice introduction to digital control followed by some basic ideas in system modelling and simulation and it then goes on to several methods of compensator design. For example, generating a controller from a continuous system using a discrete equivalent, or generating a discrete controller directly. The digital PID controller is an example of the latter. Advanced control design methods are covered and state estimation. One of the most useful chapters is probably that on the implementation issues in digital control.

The book is very nicely illustrated and provides appendices with Matlab code together with numerous examples and simulation results. The book is probably very good for engineers in industry, since it is laid out in a very clear manner. There is a frequent use of diagrams to illustrate points, and with only the essential mathematics involved.

The book is available in both eBook and hardback forms. The costs are 58.80 and 84.00, respectively. One can also rent this book and there is an e-book possibility as well.

Book Nerd.


Book Review: Stability Theory for Dynamic Equations on Time Scales, by: Martynyuk, Anatoly A., Published By: Birkhauser, Cost: 55.99

This text published by Birkhauser is relatively short being in a monograph form and presents three approaches for stability analysis and the solution of dynamic equations. The book is more aimed at mathematical control theorists, but it could be important in flight and spacecraft dynamics and other application areas.

The first chapter of the text involves the elements of the timescales analysis and the second chapter deals with dynamic integral inequalities. It will not be surprising that the Lyapunovs stability theory is considered in the third chapter and this is followed by a chapter entitled comparison method. The final chapter involves applications but they are in rather academic areas and hence this text is more useful for researchers and advanced courses in mathematical control.

Book Nerd.


Book Review: Fuel Cells: Dynamic Modeling and Control with Power Electronics Applications, Second Edition, by: Bei Gou, Woonki Na, Bill Diong , Published By: CRC Press, Cost: 114

This is quite a comprehensive text by CRC Press (Taylor & Francis Group) on fuel cell technology which is becoming massively important for a whole range of power systems and other applications. This book provides a state of the art overview ranging from the fundamentals of fuel cells to exciting future applications. The modelling is of course very important and an early chapter covers linear and nonlinear models for fuel cell dynamics. This is then followed by linear and nonlinear control methods for fuel cells. It represents a challenging area for control engineers.

There is then a chapter on Simulink implementation followed by the application of fuel cells in vehicles. Many in the automotive industry will find this text valuable for this chapter alone.

The application of fuel cells in power systems is then described on the following chapter covers the application in renewable energy systems. Optimisation is discussed and power electronics applications for fuel cells. Various applications issues are then discussed in later chapters.

This book is a second addition and it should be valuable, both as a reference text and for students and researchers in universities. It has an engineering emphasis and is liberally illustrated with helpful diagrams and results. It is certainly important for anyone involved with the development or application of fuel cell technology.

Book Nerd.