Information Technology Reference
In-Depth Information
wind
WIND
Wind input
For
Fuzzy if-then rules
detection
Tm
Wind input
For
MLP training
WIND TURBINE
Control signal
pitch angle
Fuzzy if-then
rule
detector
Pcap
Fuzzy if-then
rule box
MLP
Algorithm
box
Fuzzy if-then
rule
Fig. 9 Block diagram of FRENGA controllers
Figure
10
d shows the evolution of the output power of wind turbine in Pu that is
controlled by FRENGA.
By analyzing the simulation results, it can be concluded that the present models
allow an accurate approximation of the dynamic response of the wind turbine
operating with different winds, although the wind turbines generate the maximum
reactive power.
9 Hybrid Optimal Control Strategy
The annoyance with understanding the method is the basic obstacle in some
application of neural networks (NNs). Thus extracting knowledge from NN in the
comprehensible way has been developed (Santos et al.
2000
; Mitra and Hayashi
2000
). Generally, it has the form of propositional rules. Many rule extraction
approaches were advanced in the last few years. Framework of these methods
determines the expressive power of extracted rules.
In this section,
first MLP neural network has been trained with data extracted
from authoritative articles next we extracted Fuzzy If-Then rules from Multi-Layer
Perceptron NN using FGS optimization, which called FRENGA. In the case of rule
extraction, this work is interested in receiving the set of Fuzzy If-Then rules and
satisfying different criteria. Two approaches proposed a signal control to manage
output power and torque.
At the end online form both methods were tested on a turbine and one of them
that product better respond is selected. Figure
11
shows block diagram of the
proposed Hybrid controller (Kasiri
2011b
). Figure
12
a reveals controlled torque of
wind turbine blades that has been structured by Hybrid controller very well.
Figure
12
b presented output power of wind turbine in Pu that is controlled by
Hybrid controller.
Search WWH ::
Custom Search