Information Technology Reference
In-Depth Information
introduced. Two sets of subsystem controllers are designed and switched according
to the following law:
K i 0
for
τ<
T D
K i =
(6.20)
K i D
for
τ
T D
where:
the subsystem controllers K i 0 are designed to be robust against model uncertainties;
the subsystem controllers K i D are designed to be robust against model uncertainties
and faults.
This approach has the advantage of being less conservative in the nominal case.
6.3.5 Active FTC with Controller Reconfiguration
and Fault Detection and Isolation
In this case, the faults f
are considered to be uncertain parameters, and a fault
detection and isolation (FDI) algorithm can detect the fault occurrence at time instant
T D and isolate the fault in the set
∈ F
2 sets of subsystem
controllers are designed and switched according to the following law:
F
at time instant T I . Then, N
+
K i 0
K i D
K i I
for
T D
for T D τ<
τ<
K i =
T I
(6.21)
for
τ
T I
where:
the subsystem controllers K i 0 are designed to be robust against model uncertainties;
the subsystem controllers K i D are designed to be robust against model uncertainties
and all the possible faults in
F
;
are such that the ith controller K i I
the N subsystem controllers K i I ,...,
K i I
is
designed to be robust against model uncertainties and the jth fault F j .
6.3.6 Active FTC with Controller Reconfiguration
and Fault Detection, Isolation and Estimation
In this case, an estimation of the faults f
is provided by a fault estimation
algorithm, and such an estimation can be used as a premise variable for the controller.
Moreover, it is assumed that the FDI algorithm can detect the fault occurrence at time
instant T D and isolate the fault in the set
∈ F
2setsof
subsystem controllers are designed and switched according to the following law:
F
at time instant T I . Then, N
+
Search WWH ::




Custom Search