Environmental Engineering Reference
In-Depth Information
to avoid the time delay due to online diagnosis of the faults and reconfiguration of
the controller. In fact, this is very important in practice where the time windows
during which the system remains stabilisable in the presence of faults are very
short, e.g. the unstable double inverted pendulum example [
27
,
50
]. Most of the
PFTC methods have been proposed mainly based on robust control theory.
However, the fundamental difference between the traditional robust control and
the PFTC lies in the fact that robust control deals with small parameter variations
or model uncertainties, whilst PFTC deals with more drastic changes in system
configuration caused by faults [
4
,
17
,
34
,
41
,
47
,
48
,
53
]. It should be noted that
in the early literature the PFTC approach was referred to as 'reliable control'
[
47
,
48
].
In order to improve the post-fault control performance and cope with severe
faults that break the control loop, it is commonly advantageous to switch to a new
controller that is on-line or designed off-line to control the faulty plant.
In the AFTC approach, two conceptual steps are required: FDD and controller
adjustment so that the control law is reconfigured to achieve performance
requirements, subsequent to faults [
6
,
32
,
36
,
56
]. An AFTC system compensates
for faults either by selecting a pre-computed control law (projection-based
approaches) [
7
,
25
,
40
,
55
] or by synthesizing a new control strategy online (online
automatic controller redesign approaches) [
1
,
2
,
22
,
23
,
37
,
52
,
57
]. Another widely
studied method is the fault compensation approach, where a fault compensation
input is superimposed on the nominal control input [
8
,
15
,
19
,
28
,
31
,
54
].
It should be noted that owing to the ability of the traditional adaptive control
methods to automatically adapt controller parameters to changes in system
parameters, adaptive control has been considered as a special case of AFTC that
obviates the need for diagnosis and controller re-design steps [
45
,
51
]. Figure
7.3
shows a general overview of the main approaches used to achieve FTC for each
class.
7.3 Wind Turbine Modelling
The principle aim of control in the wind turbine systems operation is to optimise
wind energy conversion to mechanical energy which in turn is used to produce
electricity. These systems are characterised by non-linear aerodynamic behaviour
and depend on a stochastic uncontrollable wind force as a driving signal. To
conceptualise the system from analysis and control designs to real application, an
accurate overall mathematical model of the turbine dynamics is required. Nor-
mally, the model is obtained by combining the constituent subsystem models that
together make up the overall wind turbine dynamics. This Section describes the
combination of a flexible low speed shaft model together with a two-mass con-
ceptual model of a wind turbine.
The aerodynamic torque (T
a
) acting within the rotor represents the principal
source of non-linearity of the wind turbine. T
a
depends on the rotor speed x
r
, the
