Non-Linear Adaptive Control of Hydropower Generator Cooling System
No Thumbnail Available
Date
2018-11
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa University
Abstract
This thesis describes the use of nonlinear adaptive control to design the cooling system of
hydropower generators. The increasing load demands are posing serious threats to reliable
operation of power systems. These days, in the light of the attention given to reduction of carbon
dioxide emissions and efficient usage of raw materials, high-efficiency electric generators are in
the lime light over the world. As all losses generate heat it is important to have a cooling system
that can maintain a sufficiently low operating temperature within all parts of the generator.
The objective of this research was to model, design and simulate a reliable and efficient cooling
system for hydropower generators using nonlinear adaptive control. As the load on a generator
varies with time, the generators heat dissipation also varies. This variation in the heat causes the
temperature of the air surrounding the generator in the closed generator room to vary. In this
thesis this variation of temperature is studied and heat transfer analysis method is used to
determine the heat loss of the generators. Following that appropriate size of butterfly valve is
chosen which is driven by a DC servomotor and a tubular heat exchanger is designed using
dimensionless number analysis method which includes determining the Reynolds number,
Prandtl number, and Nusselt number. Then the required amount of cooling water is supplied
through the tubular heat exchanger according to the load on the generator.
In this research the cooling system is designed to keep the generator temperature below 50
C in
other words keeping the temperature of the surrounding air inside the generator room below 27
0
0
C. Matlab Simulink software is applied for performance analysis of the designed cooling
system. The simulation results show that the designed cooling system keeps the surrounding air
at a temperature of 27
C with a settling time of 0.6 seconds and without any overshoot when it
exceeds this set point (27
0
C). For this study we choose the Koka Hydropower Plant and the
required data for the design is collected from there. The butterfly valve opening angle for
adjusting the valve at the desired position in order to allow the required flow rate is related with
the temperature of the generator room in a non-linear way specifically having an exponential
relation. Thus an ANN controller is used which accounts for this process nonlinearity so that
increasing model robustness and reducing the generator stoppage time due to cooling system
failure. This results in increasing the efficiency of the hydropower plant as a whole.
Description
Keywords
Nonlinear adaptive control, Reynolds number, Prandtl number, Nusselt number, ANN