88 / 94
88
Department of Electrical
Engineering, IT and
Cybernetics
2015 – 2017
Improved Flood
Management of Lake
Toke
Supervisor:
Bernt Lie
Co-supervisor:
Nils-Olav Skeie
Kristian Dønheim Kvam
MASTER THESIS
Background
Operation of hydro power plants according to
concession requirements under flooding is
challenging. Previouswork (Lake Toke) indicated
the importance of handling data in a rigorous
way, whilealsousingefficient algorithms tosolve
the constrained optimization problem of finding
optimal flood-gate openings. This thesis deals
with storing data in a database, refitting parts
of the model, integrating the database approach
in the controller, and making it possible with
simultaneous off-line what-if simulations.
•
The main steps in this work are:
•
Development and implementation of a database
for storing measurements and on-line computed
controller results,
•
Refitting/modifying model of Lake Toke to logged
data, including data from a major flood in the fall
of 2015,
•
Integrate the database approach to data storage (as
opposed to temporary storage in MATLAB arrays)
with advanced flood control algorithm.
•
Develop approaches to off-line what-if simulation
studies using the database information, which can
be carried out while the system is in on-line control
where the controller is connected to the same
database.
•
If there is time, refine the advanced flood controller,
and develop a GUI for running the combined on-line
control and off-line what-if routines.
Figure 2 - Diagram of the system
Figure 1- Levels in Lake Toke during 2011. Due to B.
Furenes, Skagerak Kraft.
Figure 1- Levels in Lake Toke during 2011. Due to B.
Furenes, Skagerak Kraft