Advanced

Model of Hydro Power Plant—New Algorithm for Turbine Governors

Slättorp, Karl and Tuszynski, Kristian (2005) In MSc Theses
Department of Automatic Control
Abstract
The process of hydro plant is on first sight easy to understand and well documented. Development of new control strategies could be accordingly based on that knowledge and trial and error approach at the commissioning. This traditional approach shows to be time consuming and expensive. Plant models available are often obsolete as they are simplified for the narrow linear range of the working area and not directly executable. Intuitive knowledge of the original control developers is practically not available. Real plants are run under stringent economical demands and thus not available for testing. The main goal of this Master Thesis was to test modern strategies for control development using a 'test bench' for test and verification. The... (More)
The process of hydro plant is on first sight easy to understand and well documented. Development of new control strategies could be accordingly based on that knowledge and trial and error approach at the commissioning. This traditional approach shows to be time consuming and expensive. Plant models available are often obsolete as they are simplified for the narrow linear range of the working area and not directly executable. Intuitive knowledge of the original control developers is practically not available. Real plants are run under stringent economical demands and thus not available for testing. The main goal of this Master Thesis was to test modern strategies for control development using a 'test bench' for test and verification. The 'test bench' was developed as a Hydro Plant library in the modelling language Modelica, under the simulation shell Dymola (www.dynasim.se). Applicability of that test bench was thereafter tested through evaluation of a gain scheduling approach based on an online system identification using a PRBS. The results were good, proving that the Hydro Plant library is an efficient development tool. The model of an entire hydro plant allows tests of new control laws, verification of new designs of plant control components and last but not least making commissioning more efficient. (Less)
Please use this url to cite or link to this publication:
author
Slättorp, Karl and Tuszynski, Kristian
supervisor
organization
year
type
H3 - Professional qualifications (4 Years - )
subject
publication/series
MSc Theses
report number
TFRT-5744
ISSN
0280-5316
language
English
id
8847903
date added to LUP
2016-03-18 10:01:20
date last changed
2016-03-18 10:01:20
@misc{8847903,
  abstract     = {The process of hydro plant is on first sight easy to understand and well documented. Development of new control strategies could be accordingly based on that knowledge and trial and error approach at the commissioning. This traditional approach shows to be time consuming and expensive. Plant models available are often obsolete as they are simplified for the narrow linear range of the working area and not directly executable. Intuitive knowledge of the original control developers is practically not available. Real plants are run under stringent economical demands and thus not available for testing. The main goal of this Master Thesis was to test modern strategies for control development using a 'test bench' for test and verification. The 'test bench' was developed as a Hydro Plant library in the modelling language Modelica, under the simulation shell Dymola (www.dynasim.se). Applicability of that test bench was thereafter tested through evaluation of a gain scheduling approach based on an online system identification using a PRBS. The results were good, proving that the Hydro Plant library is an efficient development tool. The model of an entire hydro plant allows tests of new control laws, verification of new designs of plant control components and last but not least making commissioning more efficient.},
  author       = {Slättorp, Karl and Tuszynski, Kristian},
  issn         = {0280-5316},
  language     = {eng},
  note         = {Student Paper},
  series       = {MSc Theses},
  title        = {Model of Hydro Power Plant—New Algorithm for Turbine Governors},
  year         = {2005},
}