Projects / Programmes
January 1, 2009
- December 31, 2014
| Code |
Science |
Field |
Subfield |
| 2.06.00 |
Engineering sciences and technologies |
Systems and cybernetics |
|
| Code |
Science |
Field |
| T125 |
Technological sciences |
Automation, robotics, control engineering |
| Code |
Science |
Field |
| 2.02 |
Engineering and Technology |
Electrical engineering, Electronic engineering, Information engineering |
control technology, automatic control, cybernetics, control systems, mathematical modelling, systems engineering, software engineering, industrial informatics, fault detection, electronics, system dynamic
Researchers (33)
Organisations (1)
| no. |
Code |
Research organisation |
City |
Registration number |
No. of publicationsNo. of publications |
| 1. |
0106 |
Jožef Stefan Institute |
Ljubljana |
5051606000 |
90,753 |
Abstract
The operation of modern systems and processes is largely based on embedded control technology, which integrates knowledge from control and systems theory, electronics, computer and information sciences, systems engineering and also knowledge from target application domains.
The key objective of the underlying research programme is to contribute new methods and technical solutions for applications, which are expected to bring benefits in terms of better quality, efficiency and reliability as well as improved design, commissioning and use of control systems.
The programme is composed of four major subprojects:
A.COMPLEX SYSTEMS CONTROL
Most processes can be controlled by relatively simple and standard control methods. Unfortunately there are also processes, which are rather complex regarding their dynamics and uncertain behaviour. The focus within this subproject will be on research and application of nonlinear, predictive, self-tuning and adaptive control, able to cope with these problems. The applications on chemical processes, power systems and - in particular - wastewater treatment plants are foreseen.
B.FAULT DETECTION AND ISOLATION
On-line condition monitoring of products and processes is getting essential in modern production systems. Research in this field will concentrate on design of diagnostic systems robust to modelling errors. Particular endeavour will be paid to the application of advanced nonlinear signal processing techniques for extraction of features that proved useful in a range of industrial problems. The algorithms will be evaluated on laboratory test rigs and in the frame of demonstration applications in industry.
C.COMPUTER INTEGRATED PRODUCTION
One of the main issues in industry today is appropriate integration of the control processes on the physical (equipment) level, production level and business level. This is a hot topic, which is receiving a great deal of attention right now. Our work will concentrate on problems related to mastering the life-cycle of (integrated) control systems, to production decision support systems, to production scheduling algorithms, and to non-technical aspects of control and IT systems introduction. Research will be carried out in the close co-operation with several industrial partners in Slovenia.
D.DEVELOPMENT OF ADVANCED IMPLEMENTATION TECHNOLOGY
Control systems implementation is heavily based on mastering the appropriate implementation technology. Our focus will be on development of special-purpose modules for industrial programmable logic controllers (PLC's), on tools for more efficient implementation of advanced control algorithms, tools for better production control and on automatic code generation for PLC's.
Significance for science
The international relevance of the programme relies on available expertise of the team members and content that address problems in the domains of highest priority in the EU FP7 (Information Society Technology, Embedded Systems, New Production Technologies) and European Technology Platforms (Manufuture, Fuel Cells, Embedded Systems). A list of assorted notable scientific contributions is as follows: • Modelling of dynamic systems with evolving Gaussian-process models is a contribution to the area of on-line systems identification beyond the current state of the art. This kind of online identification enables modelling of dynamic systems with variable properties from input and output data that can be found in everyday practice of biological systems, power engineering, economics, etc. The prediction of such nonlinear models can be characterized with a measure of confidence in the predicted value. • Stochastic estimation methods such as particle filtering achieved such popularity that they are now considered as the standard tool for prognostics in PHM systems. We have addressed the computational limitations of these algorithms by using the technique known as the Rao-Blackwellization, which enables the separation of the problem to the linear and nonlinear part. The linear part of the problem can then be solved analytically, which greatly reduces the computational load. • Several improvements of cascade scheme of magnetic plasma control for the Iter fusion tokamak reactor were presented. An explicit MPC controller for vertical stabilisation was designed. It was also shown that the performance may be improved by using an MPC controller for the ITER plasma current and shape control, and that its real-time implementation is feasible. • An original methodology for the diagnosis of rotatinoal drives based on stochastic signal processing and entropy indices represents a significant novelty in the area of diagnosis under nonstationary conditions. Apart from the state-of-the-art solutions, the new ones do not need extra sensors and detaild a priori information about the machine. • A new object model of equipment and procedural control of batch processes, with dynamically defined and potentially overlapping unit classes, enables a substantial increase of the reuse of recipes and complete elimination of the repetition of information in the recipes. Experience of using the new approach shows that the number of recipes is reduced by more than a half compared with the actual approach. • A new methodology for determining the operating strategies for biochemical, wastewater treatment plants is proposed. A multi-criteria evaluation methodology is based on a model analysis under an uncertainty that can present multiple steady states. The motivation for this approach is given by a case study using an anaerobic digestion model adapted for multiple co-substrate. • New methods for tuning PI and PID controller parameters for integrating processes and underdamped processes, based on Magnitude optimum criterion, have been developed and tested in practice. A new tuning method for multivariable PID controllers, based on output decoupling and magnitude optimum criterion, has been also developed and tested. • A novel condition indicator for PEM fuel cells is constructed from estimated fuel cell impedance. The impedance values are treated as dependent complex random variables and the copula function are employed for aggregation of these values into the condition indicator. The approach provides an online condition indicator, which in an universal and unambiguous manner describes the diagnostics information of the PEM fuel cell. • In line with emerging smart-manufacturing initiatives a novel concept for production analysis and optimization was introduced that enables a systematic derivation of the actionable knowledge from the historical production data.
Significance for the country
The objectives of our group are compliant with the mission: "to foster excellence of applied research and transfer of research achievements into the practical applications, hence serving to the progress in Slovenia". Our focus is therefore mainly applied research, where a substantial emphasis is largely derived from the needs of our economy. The research has been carried out in accordance with the wellknown MODE 2 research paradigm*, characterized by (i) research in the context of application, (ii) interlacement of basic and applied research and (iii) the development with strong accent on design. That is why applied and technological results are prevailing over pure research results. The impact achieved by the underlying programme is both direct and indirect. An example of the direct impact is transfer of research results into realization of an intelligent valve drive for Danfoss Trata Ltd. Here an innovative solution has been adopted to ensure longer and more stable valve operation and energy saving in district heating systems. The valve has already become bestseller at Danfoss (till 2014 around 90,000 have been sold, over the next three years another 120,000 pieces are planned to be sold in the total amount of approximately € 13 Mio). Another example is the development of a diagnostic system for fully automatic final quality control of electric motors in the company Domel Ltd. With the introduction of these systems on seven production lines the final control became fully automatic and consequently the scrap was dramatically reduced. According to the claims of the representatives of Domel Ltd. the use of these systems provide a fundamental contribution to high endquality of their products and is one of the essential elements to increase the company's competitiveness in the global market. During the last few years, the company has become a major European supplier of electric motors for vacuum cleaner units (more than 60% market share), it annually produces more than 3 million motors and had in 2013 on annual turnover of almost € 100 Mio and net profit of € 2 Mio. Indirect impact can be seen through the results of the Competence Centre for Advanced Control Technologies, where our group played leading role in research and development activities. Twelve companies involved in the competence centre, during the period 2010-2013 increased sales in the market by 33%, reaching a total value of sales of € 473.4 Mio. With an increase of employees by 2%, the companies’ GVA per employee increased by 8%, with an average GVA of € 54,027. The share of exports in sales has also increased by 2 % to the total export value of € 381.6 Mio. Therefore, we believe that the impact of our research on our economy and society is obvious and substantial. * See , for example: Gibbons and coauthors, " The new production of knowledge " , Sage Publications , London , 1995 , or "Measuring Excellence in Research Egineering", Royal Academy of Engineering, London, 2000
Most important scientific results
Annual report
2009,
2010,
2011,
2012,
2013,
final report,
complete report on dLib.si
Most important socioeconomically and culturally relevant results
Annual report
2009,
2010,
2011,
2012,
2013,
final report,
complete report on dLib.si
