Type of research: applied Duration from: 04/30/91. to 12/31/95. Papers on project (total): 122
Institution name: Fakultet elektrotehnike,strojarstva i brodogradnje, Split (23) Department/Institute: Department of electrical engineering Department of telecommunications and information systems Address: Ruđera Boškovića bb. City: 21000 - Split, Croatia
Communication
Phone: 385 (0)21-56-37-77
Fax: 385 (0)21-56-38-77
E-mail: rozic@fesb.hr
Summary: Within the framework of the project "Forecasting in
Engineering", the application of predictive methods is being investigated
at the following levels: 1. Information source modelling, signal
processing and coding, 2. Information and communication systems modelling,
3. The application of the predictive methods in planning and control. At
the level 1., we investigate the predictive methods suitable for source
modelling (e.g. natural and artificial languages), speech, image and data
source modelling. These models are the basis for the application of
predictive coding methods in data compression (speech, image and video,
data, time series and sequences) and the basis of general adaptive signal
processing methods, applied in signal restauration or filtering,
equalization and echo cancellation. At level 2. we investigate the
predictive methods suitable for modelling of technical systems (or
subsystems) which can be considered as closed entities. These models are
the basis for the simulation of these systems. They provide the prediction
of the system performances in different working conditions. The typical
example is modelling of communication channels, protocols, local area
networks, process and production chains, power and hydrology systems etc.
At level 3. we investigate the predictive methods suitable for the
modelling of so-called complex systems, that cannot be considered as
closed systems because of their interaction with the environment. These
systems are exposed to the outer disturbances of generally stochastic
type. Such complex systems are generally statistically undetermined, are
time variant with the unknown structure. Identified and adapted models are
the basis for forecasting, planning and decision making. The typical
example is traffic planning, telecommunication systems development
planning, forecasting of new sevices demand and introduction, forecasting
of the products demand and selling, and even the planning in socioeconomic
systems.
Keywords: Modelling, prediction methods, adaptive algorithms, data compression, redundant coding, speech, audio, video, image, graphics, enciphering, communication networks, protocols, linear stochastic systems, nonlinear systems, fractals, vector quantization, time series, identification, forecasting, expert systems, decision making.
Research goals: The objective of the project is investigation of the
application oriented forecasting methods in the field of engineering,
particulary in the field of information and communication systems. It is
also applied in the field of socioeconomics as long as it interferes with
engineering. Namely, because of the fast development in engineering
technology, there is a need for a comprehensive methodology which resolves
problems depth as well as in breadth in a multidisciplinary way, i.e. at
all levels. Based on the studying of the published research results, we
believe that forecasting methodology based on the source, signal and
system modelling incombination with the optimization algorithms, offers
required solution. This kind of approach makes the educational process
together with the knowledge implementation process in system development,
exploatation and maintenance, easier and more efficient. The contribution
of our research is therefore expected in the following areas: - the
publication of scientific papers, books and textbooks related to the
field of forecasting methods and their applications, - taking
part in meetings, conference and seminars organization, - education of
undergraduate and graduate students and young researchers, -
development of original software products and their promotion,
education of engineers employed in bussiness and industry.
COOPERATION - INSTITUTIONS
Name of institution
: Arizona State University, Dept. of
Electrical Engineering Type of institution: University/Faculty Type of cooperation: Joint project City: 85287-7206 - Tempe, Arizona, USA
Name of institution
: Fakultet elektrotehnike i računalstva,
Zavod za telekomunikacije Type of institution: University/Faculty Type of cooperation: Joint publishing of scientific papers City: 10000 - Zagreb, Croatia
Name of institution
: Fakulteta za strojništvo Type of institution: University/Faculty Type of cooperation: Joint project City: 61000 - Ljubljana, Slovenija
Name of institution
: Institut Jozef Stefan Type of institution: State institute Type of cooperation: Systematic exchange of information City: 61000 - Ljubljana, Slovenija
Name of institution
: Hrvatska pošta i telekomunikacije HPT-TKC
Split Type of institution: Institution whose primary activities ar Type of cooperation: Joint project City: 21000 - Split, Croatia
Name of institution
: TERENA Type of institution: International organization Type of cooperation: Joint project City: ox11 0qs - Oxon, United Kingdom Other information about the project.