U.S. patent application number 15/103822 was filed with the patent office on 2016-10-27 for improvements in or relating to converters for use in high voltage direct current power transmission.
This patent application is currently assigned to General Electric Technology GmbH. The applicant listed for this patent is GENERAL ELECTRIC TECHNOLOGY GMBH. Invention is credited to Michael Marc Claude Merlin, Kevin Dyke, Timothy Charles Green, Omar Fadhel Jasim, Francisco Jose Moreno Munoz.
Application Number | 20160315548 15/103822 |
Document ID | / |
Family ID | 49880638 |
Filed Date | 2016-10-27 |
United States Patent
Application |
20160315548 |
Kind Code |
A1 |
Jasim; Omar Fadhel ; et
al. |
October 27, 2016 |
IMPROVEMENTS IN OR RELATING TO CONVERTERS FOR USE IN HIGH VOLTAGE
DIRECT CURRENT POWER TRANSMISSION
Abstract
In the field of high voltage direct current (HVDC) power
transmission, a converter comprises three converter limbs, each
corresponding to a respective phase of the converter, each
extending between first and second DC terminals and each including
first and second limb portions separated by an AC terminal. Each
limb portion includes a chain-link converter that is operable to
provide a stepped variable voltage source, and a primary switching
element to selectively switch the respective limb portion into and
out of circuit. The converter also includes a first controller that
is programmed to selectively operate for one converter limb at a
time the primary switching element in each limb portion thereof to
simultaneously switch both the first and second limb portions into
circuit and thereby define a fully-conducting converter limb to
sequentially route via each said fully-conducting converter limb a
DC current demand (I.sub.DC) between the first and second DC
terminals.
Inventors: |
Jasim; Omar Fadhel;
(Nottingham, GB) ; Moreno Munoz; Francisco Jose;
(Bera, Navarra, ES) ; Claude Merlin; Michael Marc;
(Dourdan, FR) ; Green; Timothy Charles; (Haywards
Heath, GB) ; Dyke; Kevin; (Stafford, GB) |
|
Applicant: |
Name |
City |
State |
Country |
Type |
GENERAL ELECTRIC TECHNOLOGY GMBH |
Baden |
|
CH |
|
|
Assignee: |
General Electric Technology
GmbH
Baden
CH
|
Family ID: |
49880638 |
Appl. No.: |
15/103822 |
Filed: |
December 12, 2014 |
PCT Filed: |
December 12, 2014 |
PCT NO: |
PCT/EP2014/077636 |
371 Date: |
June 10, 2016 |
Current U.S.
Class: |
1/1 |
Current CPC
Class: |
H02M 1/00 20130101; H02M
2001/0003 20130101; H02J 3/36 20130101; H02M 7/72 20130101; H02M
5/458 20130101; H02M 7/483 20130101; H02M 7/797 20130101; H02M
2007/4835 20130101; H02M 2001/0025 20130101 |
International
Class: |
H02M 5/458 20060101
H02M005/458 |
Foreign Application Data
Date |
Code |
Application Number |
Dec 12, 2013 |
EP |
13275311.2 |
Claims
1. A converter, for use in high voltage direct current power
transmission, comprising: three converter limbs, each corresponding
to a respective phase of the converter, each extending between
first and second DC terminals and each including first and second
limb portions separated by an AC terminal, each limb portion
including a chain-link converter operable to provide a stepped
variable voltage source and a primary switching element to
selectively switch the respective limb portion into and out of
circuit; and a first controller programmed to selectively operate
for one converter limb at a time the primary switching element in
each limb portion thereof to simultaneously switch both the first
and second limb portions into circuit and thereby define a
fully-conducting converter limb to sequentially route via each said
fully-conducting converter limb a DC current demand between the
first and second DC terminals.
2. A converter according to claim 1 wherein the first controller is
programmed to sequentially define fully-conducting converter limbs
at regular intervals of around 60 electrical degrees.
3. A converter according to claim 1 wherein the first controller is
additionally programmed, while selectively operating for a given
converter limb the primary switching element in each limb portion
thereof to simultaneously switch both the first and second limb
portions into circuit and thereby define a fully-conducting
converter limb, to concurrently operate for each other converter
limb the primary switching element in one or both limb portions
thereof to switch a single limb portion into circuit and thereby
define a partially-conducting converter limb to direct respective
AC current demand phase waveforms towards a given AC terminal
whereby the respective AC current demand phase waveforms sum to
zero.
4. A converter according to claim 1 further including a second
controller programmed to: (a) obtain a respective AC current demand
phase waveform for each converter limb which the corresponding
converter limb is required to track, and a DC current demand which
each converter limb is also required to track; and (b) carry out
mathematical optimization to determine an optimal limb portion
current for each limb portion that the limb portion must contribute
to track the corresponding required AC current demand phase
waveform and the required DC current demand.
5. A converter according to claim 4 wherein the second controller
is programmed to carry out mathematical optimization by creating an
equivalent converter configuration which represents the flow of
current through the converter.
6. A converter according to claim 5 wherein the second controller
is programmed to create an equivalent converter configuration which
represents the flow of current through the converter by mapping
possible current flow paths through the converter.
7. A converter according to claim 1 wherein the second controller
is programmed to carry out mathematical optimization by applying a
current weighting to the relative current contribution provided by
a plurality of limb portions.
8. A converter according to claim 7 wherein the second controller
is programmed to determine the or each weighting according to
measured operating parameters of the converter.
9. A converter according to claim 7 wherein when controlling the
converter under a particular operating condition the second
controller is programmed to apply a weighting by applying a
different weighting to at least one limb portion such that the or
each said limb portion provides a different contribution to the
other limb portions.
10. A converter according to claim 4 wherein the second controller
is programmed to carry out mathematical optimization to determine
one or more minimum individual limb portion currents that the
corresponding limb portion must contribute to track the
corresponding required AC current demand phase waveform and the
required DC current demand.
11. A converter according to claim 4 wherein the second controller
is further programmed to carry out mathematical optimization to
provide optimal limb portion voltage sources.
Description
[0001] This invention relates to a converter for use in high
voltage direct current power transmission.
[0002] In high voltage direct current (HVDC) power transmission
networks alternating current (AC) power is typically converted to
direct current (DC) power for transmission via overhead lines
and/or under-sea cables. This conversion removes the need to
compensate for the AC capacitive load effects imposed by the power
transmission medium, i.e. the transmission line or cable, and
reduces the cost per kilometre of the lines and/or cables, and thus
becomes cost-effective when power needs to be transmitted over a
long distance.
[0003] The conversion between DC power and AC power is utilized in
power transmission networks where it is necessary to interconnect
the DC and AC electrical networks. In any such power transmission
network, converters are required at each interface between AC and
DC power to effect the required conversion; AC to DC or DC to
AC.
[0004] According to a first aspect of the invention there is
provided a converter, for use in high voltage direct current power
transmission, comprising: [0005] three converter limbs, each
corresponding to a respective phase of the converter, each
extending between first and second DC terminals and each including
first and second limb portions separated by an AC terminal, each
limb portion including a chain-link converter operable to provide a
stepped variable voltage source and a primary switching element to
selectively switch the respective limb portion into and out of
circuit; and [0006] a first controller programmed to selectively
operate for one converter limb at a time the primary switching
element in each limb portion thereof to simultaneously switch both
the first and second limb portions into circuit and thereby define
a fully-conducting converter limb to sequentially route via each
said fully-conducting converter limb a DC current demand between
the first and second DC terminals.
[0007] Sequentially routing, via each fully-conducting converter
limb in turn, a DC current demand between the first and second DC
terminals allows the DC current demand to flow continuously between
the first and second DC terminals, and thereby permits the
converter of the invention to continuously exchange power with a DC
network, connected in use with the first and second DC terminals,
throughout an operating cycle of the converter.
[0008] Meanwhile, having only one fully-conducting converter limb
at a time avoids the creation of current paths between respective
converter limbs, and so no transient circulating currents between
converter limbs arise. The absence of such circulating currents
permits the removal of a passive inductor from each limb portion
that would otherwise be needed to limit the level of aforementioned
circulating current between the converter limbs.
[0009] Such passive inductors are physically very large, and so
omitting them allows for a significant reduction, e.g. of around
20%, in the overall footprint of a converter station including the
converter of the invention. This in turn helps to reduce
considerably the cost of the converter station.
[0010] In addition, omitting the passive inductor from each limb
portion also means that it is possible to connect a transformer
directly with the converter without the need for interconnecting
bushings which are typically large and expensive, and so offers
further space- and cost-saving opportunities.
[0011] In addition, omitting an inductor from each limb portion
allows for an increase in the level of AC voltage that can be
generated by the converter, while permitting a reduction in the
level of AC current, i.e. the level of individual AC current demand
phase waveforms, that must be provided, and so gives rise to an
increase in the efficiency of the converter.
[0012] Preferably the first controller is programmed to
sequentially define fully-conducting converter limbs at regular
intervals of around 60 electrical degrees.
[0013] The inclusion of a first controller so programmed results in
the DC current demand routed between the first and second DC
terminal being smooth and continuous.
[0014] Optionally the first controller is additionally programmed,
while selectively operating for a given converter limb the primary
switching element in each limb portion thereof to simultaneously
switch both the first and second limb portions into circuit and
thereby define a fully-conducting converter limb, to concurrently
operate for each other converter limb the primary switching element
in one or both limb portions thereof to switch a single limb
portion into circuit and thereby define a partially-conducting
converter limb to direct respective AC current demand phase
waveforms towards a given AC terminal whereby the respective AC
current demand phase waveforms sum to zero.
[0015] Having the AC current demand phase waveforms sum to zero at
a given AC terminal, i.e. within the converter of the invention,
eliminates the inclusion of any AC components in the DC current
demand routed between the first and second DC terminals, and so
avoids the need to filter this current before, e.g. passing it to a
DC network connected in use to the first and second DC
terminals.
[0016] Any kind of filter in a HVDC installation has major
implications with regards to the footprint of a resulting converter
station, and so avoiding such filters is very beneficial.
[0017] A converter according to preferred embodiment of the
invention further includes a second controller programmed to:
[0018] (a) obtain a respective AC current demand phase waveform for
each converter limb which the corresponding converter limb is
required to track, and a DC current demand which each converter
limb is also required to track; and [0019] (b) carry out
mathematical optimization to determine an optimal limb portion
current for each limb portion that the limb portion must contribute
to track the corresponding required AC current demand phase
waveform and the required DC current demand.
[0020] Carrying out the aforementioned mathematical optimization,
i.e. selecting the best individual limb portion current (with
regard to chosen criteria) from a set of available alternatives,
allows the AC and DC current demands to be controlled independently
of one another, e.g. by a higher level controller.
[0021] It also permits individual limb portion currents to vary
independently of one another to accommodate different current flow
paths through the converter, e.g. as occasioned by the sequential
definition of a fully-conducting converter limb and respective
partially-conducting converter limbs throughout each operating
cycle of the converter.
[0022] Moreover, the second controller is able to carry out steps
(a) and (b) in real time so as to permit robust control of the
converter of the invention.
[0023] The second controller may be programmed to carry out
mathematical optimization by creating an equivalent converter
configuration which represents the flow of current through the
converter.
[0024] Creating an equivalent converter configuration in the
aforementioned manner imposes constraints on the way in which the
converter can be controlled and so assists in carrying out
mathematical optimization to determine each optimal limb portion
current.
[0025] Optionally the second controller is programmed to create an
equivalent converter configuration which represents the flow of
current through the converter by mapping possible current flow
paths through the converter.
[0026] Mapping the possible current flow paths through the
converter helps the second controller to tailor the mathematical
optimization it provides to the topology, i.e. structure, of the
converter of the invention.
[0027] Preferably the second controller is programmed to carry out
mathematical optimization by applying a current weighting to the
relative current contribution provided by a plurality of limb
portions.
[0028] Applying such weightings allows variations in the
performance of each limb portion to be further accommodated while
continuing to optimise the operation of the converter as a
whole.
[0029] The second controller may be programmed to determine the or
each weighting according to measured operating parameters of the
converter.
[0030] Determining the weightings in the aforementioned manner
allows the second controller to take into account environmental
factors which might affect the healthy operation of the converter,
and to alter the optimal limb portion currents that are determined
in an effort to overcome the environmental factors and alleviate
the associated impact on the operation of the converter. Examples
of such environmental factors include the components in one limb
portion running hot, or a limb portion suffering component damage
or failure such that its performance is degraded.
[0031] In another preferred embodiment of the invention the second
controller, when controlling the converter under a particular
operating condition, is programmed to apply a weighting by applying
a different weighting to at least one limb portion such that the or
each said limb portion provides a different contribution to the
other limb portions.
[0032] Such a feature allows the second controller to distinguish
between one limb portion and another, e.g. according to how well a
given limb portion is performing.
[0033] This is useful in circumstances where it becomes desirable
to reduce the level of current contributed by a given limb portion,
e.g. because the cooling associated with the limb portion is
operating at a reduced capacity, and temporarily increase the level
of current provided by one or more other limb portions so as to
allow the converter to continue to operate and provide a high level
of power conversion.
[0034] It can also be used to reduce the limb portion voltage that
a given limb portion must provide, e.g. in circumstances where a
fault or other damage has degraded the performance of the given
limb portion, such that the converter remains able to continue
operating and provide a high level of power conversion.
[0035] Preferably the second controller is programmed to carry out
mathematical optimization to determine one or more minimum
individual limb portion currents that the corresponding limb
portion must contribute to track the corresponding required AC
current demand phase waveform and the required DC current
demand.
[0036] Determining one or more minimum individual limb portion
currents reduces the conduction and switching losses in each limb
portion because ordinarily such losses are proportional to current
squared, i.e. I.sup.2.
[0037] In a still further preferred embodiment of the invention the
second controller is further programmed to carry out mathematical
optimization to provide optimal limb portion voltage sources.
[0038] The inclusion of a second controller so programmed assists
in the provision, in the most efficient manner possible, of
individual limb portion currents that vary independently of one
another.
[0039] There now follows a brief description of preferred
embodiments of the invention, by way of non-limiting example, with
reference being made to the following figures in which:
[0040] FIG. 1 shows a schematic view of a converter according to a
first embodiment of the invention;
[0041] FIG. 2 shows a preferred switching sequence of primary
switching elements within the converter shown in FIG. 1 during an
operating cycle of the converter;
[0042] FIG. 3 illustrates schematically the selective definition of
a fully-conducting converter limb and respective
partially-conducting converter limbs in the converter shown in FIG.
1;
[0043] FIG. 4 shows a flow diagram that illustrates the principal
steps a second controller in the converter shown in FIG. 1 is
programmed to carry out;
[0044] FIG. 5 shows a schematic representation of an equivalent
converter configuration corresponding to the converter shown in
FIG. 1;
[0045] FIG. 6(a) shows a flow diagram that illustrates the
principal steps a further second controller may be programmed to
carry out; and
[0046] FIG. 6(b) shows a schematic view of a feedback loop which
forms a part of the principal steps the further controller may be
programmed to carry out.
[0047] A converter according to a first embodiment of the invention
is designated generally by reference numeral 10, as shown in FIG.
1.
[0048] The converter 10 includes three converter limbs 12A, 12B,
12C, each of which corresponds to a respective phases A, B, C of
the converter 10.
[0049] Each converter limb 12A, 12B, 12C extends between first and
second DC terminals 14, 16, and each converter limb 12A, 12B, 12C
includes a first limb portion 12A+, 12B+, 12C+ and a second limb
portion 12A-, 12B-, 12C- which are separated by an AC terminal 18A,
18B, 18C.
[0050] In use, the first and second DC terminals 14, 16 are
connected to a DC network 20, with the first DC terminal 14
carrying a voltage of V.sub.DC+ and the second DC terminal 16
carrying a voltage of V.sub.DC-, while the AC terminal 18A, 18B,
18C is connected to a corresponding phase A, B, C of a three-phase
AC network 22 and carries a corresponding AC voltage phase waveform
V.sub.A, V.sub.B, V.sub.C.
[0051] Each limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C-
includes a chain-link converter 24A+, 24A-, 24B+, 24B-, 24C+, 24C-
that includes a chain of modules 26 connected in series. The number
of modules 26 in each chain-link converter 24A+, 24A-, 24B+, 24B-,
24C+, 24C- depends on the required voltage rating of the respective
limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C-.
[0052] Each module 26 of each chain-link converter 24A+, 24A-,
24B+, 24B-, 24C+, 24C- includes two pairs of secondary switching
elements (not shown) connected in parallel with an energy storage
device, in the form of a capacitor (not shown), to define a
4-quadrant bipolar module 26 that can provide negative, zero or
positive voltage and can conduct current in two directions.
[0053] In use, the secondary switching elements of the modules 26
of each chain-link converter 24A+, 24A-, 24B+, 24B-, 24C+, 24C- are
operated to enable each chain-link converter 24A+, 24A-, 24B+,
24B-, 24C+, 24C- to provide a stepped variable voltage source. The
secondary switching elements are also desirably switched at near to
the fundamental frequency of the AC network 22.
[0054] The capacitor of each module 26 may be bypassed or inserted
into the respective chain-link converter 24A+, 24A-, 24B+, 24B-,
24C+, 24C- by changing the state of the secondary switching
elements.
[0055] The capacitor of each module 26 is bypassed when the pairs
of secondary switching elements are configured to form a short
circuit in the module 26. This causes current in the corresponding
limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C- of the converter 10
to pass through the short circuit and bypass the capacitor, and so
the module 26 is able to provide a zero voltage.
[0056] The capacitor of each module 26 is inserted into the
respective chain-link converter 24A+, 24A-, 24B+, 24B-, 24C+, 24C-
when the pair of secondary switching elements is configured to
allow the aforementioned current to flow into and out of the
capacitor. The capacitor is then able to charge or to discharge its
stored energy so as to provide a voltage. The bidirectional nature
of the 4-quadrant bipolar module 26 means that the capacitor may be
inserted into the module 26 in either forward or reverse directions
so as to provide a positive or negative voltage.
[0057] It is therefore possible to build up a combined voltage
across each chain-link converter 24A+, 24A-, 24B+, 24B-, 24C+, 24C-
which is higher than the voltage available from each individual
module 26 via the insertion of the capacitors of multiple modules
26, each providing its own voltage, into the chain-link converter
24A+, 24A-, 24B+, 24B-, 24C+, 24C-.
[0058] The ability of a 4-quadrant bipolar module 26 to provide
positive or negative voltages means that the voltage across each
chain-link converter 24A+, 24A-, 24B+, 24B-, 24C+, 24C- may be
built up from a combination of modules 26 providing positive or
negative voltages. The energy levels in individual capacitors may
be maintained therefore at optimal levels by controlling the
modules 26 to alternate between providing positive or negative
voltage.
[0059] It is possible to vary the timing of switching operations
for each module 26 such that the insertion and/or bypass of the
capacitors of individual modules 26 in the chain-link converter
24A+, 24A-, 24B+, 24B-, 24C+, 24C- results in the generation of a
voltage waveform at a corresponding AC terminal 18A, 18B, 18C. For
example, insertion of the capacitors of the individual modules 26
may be staggered to generate a sinusoidal waveform. Other waveform
shapes may be generated by adjusting the timing of switching
operations for each module 26 in the chain-link converter 24A+,
24A-, 24B+, 24B-, 24C+, 24C-.
[0060] In this manner the chain-link converters 24A+, 24A-, 24B+,
24B-, 24C+, 24C- are able to facilitate power transfer between the
AC and DC networks 22, 20.
[0061] In addition to the foregoing, the inclusion of 4-quadrant
bipolar modules 26 means that it is possible for each chain-link
converter 24A+, 24A-, 24B+, 24B-, 24C+, 24C- to block a DC fault
current in the event of e.g. a pole to pole fault in the DC network
20, with such blocking being achieved by opening appropriate
secondary switching elements in each module 26 to prevent the flow
of fault current through each said module 26.
[0062] Each limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C- also
includes a primary switching element Sw.sub.1, Sw.sub.4, Sw.sub.3,
Sw.sub.6, Sw.sub.5, Sw.sub.2 that is connected in series with the
corresponding chain-link converter 24A+, 24A-, 24B+, 24B-, 24C+,
24C-. In use, each primary switching element Sw.sub.1, Sw.sub.2,
Sw.sub.3, Sw.sub.4, Sw.sub.5, Sw.sub.6 selectively switches the
respective limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C- into and
out of circuit within the converter 10.
[0063] In other embodiments of the invention (not shown) each
primary switching element may include a plurality of, e.g.
series-connected, switching elements depending on the required
voltage rating of each limb portion 12A+, 12A-, 12B+, 12B-, 12C+,
12C-.
[0064] In addition, the series connection between each primary
switching element Sw.sub.1, Sw.sub.2, Sw.sub.3, Sw.sub.4, Sw.sub.5,
Sw.sub.6 and the corresponding chain-link converter 24A+, 24A-,
24B+, 24B-, 24C+, 24C- in each limb portion 12A+, 12A-, 12B+, 12B-,
12C+, 12C- allows, in other embodiments of the invention, the
respective primary switching element Sw.sub.1, Sw.sub.2, Sw.sub.3,
Sw.sub.4, Sw.sub.5, Sw.sub.6 and corresponding chain-link converter
24A+, 24A-, 24B+, 24B-, 24C+, 24C- to be connected in a reverse
order between the corresponding AC terminal 18A, 18B, 18C and the
respective first or second DC terminal 14, 16.
[0065] In the embodiment shown, each primary switching element
Sw.sub.1, Sw.sub.2, Sw.sub.3, Sw.sub.4, Sw.sub.5, Sw.sub.6 and each
of the secondary switching elements in the respective chain-link
converters 24A+, 24A-, 24B+, 24B-, 24C+, 24C- is an insulated gate
bipolar transistor (IGBT) 28 connected in parallel with an
anti-parallel diode 30.
[0066] In other embodiments of the invention (not shown) one or
more of the primary and secondary switching elements Sw.sub.1,
Sw.sub.2, Sw.sub.3, Sw.sub.4, Sw.sub.5, Sw.sub.6 may include a
different semiconductor device, such as a field effect transistor,
a gate-turn-off thyristor, an injection gate enhanced thyristor, an
integrated gate commutated transistor or another
externally-commutated semiconductor switch, i.e. a semiconductor
switch which is turned off by one or more external components
causing the current flowing through the semiconductor switch to
fall to zero. Such other externally-commutated semiconductor
switches can include so-called `forced commutated` and `self
commutated` semiconductor switches. In each instance the
semiconductor device is preferably connected in parallel with an
anti-parallel diode.
[0067] For the reasons set out hereinabove, each limb portion 12A+,
12A-, 12B+, 12B-, 12C+, 12C- omits any form of physical, passive
inductor component, which in turns provides considerable benefits
in terms of reducing the overall footprint of a resulting converter
station in which the converter of the invention is
incorporated.
[0068] In addition to the foregoing, the converter 10 includes a
first controller 32 that is arranged in operative communication
with each primary switching element Sw.sub.1, Sw.sub.2, Sw.sub.3,
Sw.sub.4, Sw.sub.5, Sw.sub.6.
[0069] The first controller 32 is a programmable device, such as a
microcontroller, and more particularly is programmed to operate,
for one converter limb 12A, 12B, 12C at a time, the primary
switching element Sw.sub.1, Sw.sub.2, Sw.sub.3, Sw.sub.4, Sw.sub.5,
Sw.sub.6 in each limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C-
thereof to simultaneously switch both the first limb portions 12A+,
12B+, 12C+ and the second limb portions 12A-, 12B-, 12C- into
circuit and thereby define a fully conducting converter limb 12A,
12B, 12C.
[0070] In this way the first controller 32 sequentially routes, via
each said fully-conducting converter limb 12A, 12B, 12C, a DC
current demand I.sub.DC (i.e. the DC current that the converter
limbs 12A, 12B, 12C are required to track) between the first and
second DC terminals 14, 16.
[0071] More particularly still, the first controller 32 is
programmed to sequentially define fully-conducting converter limbs
12A, 12B, 12C at regular intervals 34.sub.1, 34.sub.2, 34.sub.3,
34.sub.4, 34.sub.5, 34.sub.6 of around 60 electrical degrees. In
this regard, in an ideal case each interval is 60 electrical
degrees, although for practical implementation purposes each
interval can lie in the ranges 60.+-.1 electrical degrees, or
60.+-.2 electrical degrees.
[0072] The first controller 32 is also additionally programmed,
while selectively operating for a given converter limb 12A, 12B,
12C the primary switching element Sw.sub.1, Sw.sub.2, Sw.sub.3,
Sw.sub.4, Sw.sub.5, Sw.sub.6 in each limb portion thereof 12A+,
12A-, 12B+, 12B-, 12C+, 12C- to simultaneously switch both the
first and second limb portions 12A+, 12B+, 12C+, 12A-, 12B-, 12C-
into circuit and thereby define a fully-conducting converter limb
12A, 12B, 12C, to concurrently operate for each other converter
limb 12A, 12B, 12C the primary switching element Sw.sub.1,
Sw.sub.2, Sw.sub.3, Sw.sub.4, Sw.sub.5, Sw.sub.6 in one or both
limb portions 12A+, 12A-, 12B+, 12B-, 12C+, 12C- thereof to switch
a single limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C- into
circuit and thereby define a partially-conducting converter limb
12A, 12B, 12C.
[0073] In this manner the first controller 32 is programmed to
direct respective AC current demand phase waveforms I.sub.A,
I.sub.B, I.sub.C, i.e. respective AC phase currents that the
converter 10 is required to track, towards a given AC terminal 18A,
18B, 18C whereby the AC current demand phase waveforms I.sub.A,
I.sub.B, I.sub.C sum to zero.
[0074] FIG. 2 shows one example switching sequence of the primary
switching elements Sw.sub.1, Sw.sub.2, Sw.sub.3, Sw.sub.4,
Sw.sub.5, Sw.sub.6 which is implemented by the first controller 32.
Other switching sequences may, however, also be implemented.
[0075] In the embodiment shown, the first controller 32 utilises a
phase locked loop (PLL) control scheme to coordinate the switching
sequence with respective AC voltage phase waveforms V.sub.A,
V.sub.B, V.sub.C of the AC network 22.
[0076] More particularly, during a first interval 34.sub.1 the
first controller 32 operates the primary switching element
Sw.sub.1, Sw.sub.4 in each limb portion 12A+, 12A- of a first
converter limb 12A to simultaneously switch both the first and
second limb portions 12A+, 12A- into circuit and thereby define a
fully-conducting converter limb 12A.
[0077] At the same time, i.e. concurrently with the foregoing, the
first controller 32 operates the primary switching element Sw.sub.6
in the second limb portion 12B- of a second converter limb 12B to
switch the second limb portion 12B- into circuit and thereby define
a partially-conducting converter limb 12B. The primary switching
element Sw.sub.3 in the first limb portion 12B+ of the second
converter limb 12B remains switched out of circuit.
[0078] The first controller 32 also, at the same time, operates the
primary switching element Sw.sub.5 in the first limb portion 12C+
of a third converter limb 12C to switch the first limb portion 12C+
into circuit and thereby define another partially-conducting
converter limb 12C. The primary switching element Sw.sub.2 in the
second limb portion 12C- of the third converter limb 12C remains
switched out of circuit.
[0079] During a second interval 34.sub.2 the first controller 32
leaves the first limb portion 12C+ of the third converter limb 12C
switched into circuit while operating the primary switching element
Sw.sub.2 in the second limb portion 12C- of the third converter
limb 12C to switch the second limb portion 12C- into circuit and
thereby define a fully-conducting converter limb 12C. At the same
time the first controller leaves the first limb portion 12A+ of the
first converter limb 12A switched into circuit while operating the
switching element Sw.sub.4 in the second limb portion 12A- to
switch the second limb portion 12A- out of circuit and thereby
define a partially-conducting converter limb 12A. The first
controller 32 also, at the same time, leaves the second limb
portion 12B- of the second converter limb 12B switched into circuit
to continue to define a partially-conducting converter limb 12B.
The primary switching element Sw.sub.3 in the first limb portion
12B+ of the second converter limb 12B remains switched out of
circuit.
[0080] During a third interval 34.sub.3 the first controller 32
operates the primary switching element Sw.sub.3 in the first limb
portion 12B+ of the second converter limb 12B to switch the first
limb portion 12B+ into circuit and thereby define, along with the
already switched into circuit second limb portion 12B-, a
fully-conducting converter limb 12B. At the same time the first
controller 32 continues to leave the first limb portion 12A+ of the
first converter limb 12A switched into circuit and the second limb
portion 12A- of the second limb portion 12A- switched out of
circuit to continue to define a partially-conducting converter limb
12A. The first controller 32 also, at the same time, leaves the
second limb portion 12C- of the third converter limb 12C switched
into circuit while operating the switching element Sw.sub.5 in the
first limb portion 12C+ of the third converter limb 12C to switch
the first limb portion 12C+ out of circuit to thereby define a
partially-conducting converter limb 12C.
[0081] During a fourth interval 34.sub.4 the first controller 32
operates the primary switching element Sw.sub.4 in the second limb
portion 12A- of the first converter limb 12A to switch the second
limb portion 12A- into circuit and thereby define, along with the
already switched into circuit first limb portion 12A+, a
fully-conducting converter limb 12A, as shown by way of example in
FIG. 3. As also shown in FIG. 3, the aforesaid fully-conducting
converter limb 12A routes a DC current demand I.sub.DC between the
first and second DC terminals 14, 16.
[0082] At the same time the first controller 32 continues to leave
the first limb portion 12B+ of the second converter limb 12B
switched into circuit while operating the switching element
SW.sub.6 of the second limb portion 12B- of the second converter
limb 12B to switch the second limb portion 12B- out of circuit to
thereby define a partially-conducting converter limb 12B, as also
shown in FIG. 3. As additionally shown in FIG. 3, the
partially-conducting converter limb 12B, i.e. the first limb
portion 12B+ thereof, directs an AC current demand phase waveform
I.sub.B, i.e. an AC phase current I.sub.B, towards a first AC
terminal 18A.
[0083] The first controller 32 also, at the same time, leaves the
first limb portion 12C+ of the third converter limb 12C switched
out of circuit and the second limb portion 12C- of the third
converter limb 12C switched into circuit to continue to define a
partially-conducting converter limb 12C, as again shown in FIG. 3.
As shown in FIG. 3, the partially conducting converter limb 12C,
i.e. the second limb portion 12C- thereof, also directs an AC
current demand phase waveform I.sub.C, i.e. an AC phase current
I.sub.C, towards the first AC terminal 18A.
[0084] Each of the aforementioned AC current demand phase waveforms
I.sub.B, I.sub.C, together with a further AC current demand phase
waveform I.sub.A sum to zero at the first AC terminal 18A, and
thereby cancel one another out such that they do not impact
adversely on the quality, i.e. smoothness, of the DC current demand
I.sub.DC routed between the first and second DC terminals 14,
16.
[0085] During a fifth interval 34.sub.5 the first controller 32
operates the primary switching element Sw.sub.5 in the first limb
portion 12C+ of the third converter limb 12C to switch the first
limb portion 12C+ into circuit and thereby define, along with the
already switched into circuit second limb portion 12C-, a
fully-conducting converter limb 12C. At the same time the first
controller 32 continues to leave the second limb portion 12A- of
the first converter limb 12A switched into circuit while operating
the switching element Sw.sub.1 of the first limb portion 12A+ of
the first converter limb 12A to switch the first limb portion 12A+
out of circuit to thereby define a partially-conducting converter
limb 12A. The first controller 32 also, at the same time, leaves
the first limb portion 12B+ of the second converter limb 12B
switched into circuit and the second limb portion 12B- of the
second converter limb 12B switched out of circuit to continue to
define a partially-conducting converter limb 12B.
[0086] During a sixth and final interval 34.sub.6 the first
controller 32 operates the primary switching element Sw.sub.6 in
the second limb portion 12B- of the second converter limb 12B to
switch the second limb portion 12B- into circuit and thereby
define, along with the already switched into circuit first limb
portion 12B+, a fully-conducting converter limb 12B. At the same
time the first controller 32 continues to leave the second limb
portion 12A- of the first converter limb 12A switched into circuit
and the first limb portion 12A+ switched out of circuit to continue
to define a partially-conducting converter limb 12A. The first
controller 32 also, at the same time, leaves the first limb portion
12C+ of the third converter limb 12C switched into circuit while
operating the switching element Sw.sub.2 of the second limb portion
12C- of the third limb portion 12C to switch the second limb
portion 12C- out of circuit and thereby define a
partially-conducting converter limb 12C.
[0087] It follows that the first controller 32 sequentially defines
single, individual first, second, third, fourth, fifth and sixth
fully-conducting converter limbs 12A, 12C, 12B, 12A, 12C, 12B
during corresponding first, second, third, fourth, fifth and sixth
intervals 34.sub.1, 34.sub.2, 34.sub.3, 34.sub.4, 34.sub.5,
34.sub.6 of a complete operating cycle 36 of the converter 10.
[0088] The converter 10 also includes a second controller 38 that
is arranged in communication with the first controller 32 and each
of the chain-link converters 24A+, 24A-, 24B+, 24B-, 24C+, 24C-.
The second controller 38 is similarly a programmable device, such
as a microcontroller. Although in the embodiment described, the
first and second controllers 32, 38 are shown as separate items,
they may in other embodiments of the invention form individual
parts or a single part of a larger controller or controller
arrangement.
[0089] Returning to the embodiment shown, the second controller 38
is programmed to: [0090] (a) obtain a respective AC current demand
phase waveform I.sub.A, I.sub.B, I.sub.C for each converter limb
12A, 12B, 12C which the corresponding converter limb 12A, 12B, 12C
is required to track, and a DC current demand I.sub.DC which each
converter limb 12A, 12B, 12C is also required to track; and [0091]
(b) carry out mathematical optimization to determine an optimal
limb portion current I.sub.A+, I.sub.A-, I.sub.B+, I.sub.B-,
I.sub.C+, I.sub.C- for each limb portion 12A+, 12A-, 12B+, 12B-,
12C+, 12C- that the limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C-
must contribute to track the corresponding required AC current
demand phase waveform I.sub.A, I.sub.B, I.sub.C and the required DC
current demand I.sub.DC.
[0092] The second controller 38 is also further programmed to (c)
carry out mathematical optimization to provide optimal limb portion
voltage sources V.sub.A+, V.sub.A-, V.sub.B+, V.sub.B-, V.sub.C+,
V.sub.C-, with these principal steps (a), (b) and (c) being
illustrated in a first flow diagram 40 shown in FIG. 4.
[0093] As set out above, the second controller 38 is programmed to
first obtain a respective AC current demand phase waveform I.sub.A,
I.sub.B, I.sub.C for each converter limb 12A, 12B, 12C which each
converter limb 12A, 12B, 12C is required to track, and then obtain
a DC current demand I.sub.DC which the converter limbs 12A, 12B,
12C are also required to track.
[0094] The various AC current demand phase waveforms I.sub.A,
I.sub.B, I.sub.C and the DC current demand I.sub.DC may be obtained
directly from a higher-level controller (not shown) within a
converter or from some other external entity. Alternatively the
converter 10 may obtain it directly by carrying out its own
calculations.
[0095] The second controller 38 is also programmed to, as a second
step (and as indicated by a first process box 42 in the first flow
diagram 40), carry out mathematical optimization to determine an
optimal limb portion current I.sub.A+, I.sub.A-, I.sub.B+,
I.sub.B-, I.sub.C+, I.sub.C- for each limb portion 12A+, 12A-,
12B+, 12B-, 12C+, 12C- that the limb portion 12A+, 12A-, 12B+,
12B-, 12C+, 12C- must contribute to track the corresponding
required AC current demand phase waveform I.sub.A, I.sub.B, I.sub.C
and the required DC current demand I.sub.DC.
[0096] The second controller 38 is programmed to carry out such
mathematical optimization by creating an equivalent converter
configuration 100, as shown in FIG. 5, which represents the flow of
current through the corresponding converter 10 of the
invention.
[0097] The equivalent converter configuration 100 includes similar
features to the converter 10 of the invention and these like
features share the same reference numerals. To that end the
equivalent converter configuration 100 includes three converter
limbs 12A, 12B, 12C, each of which corresponds to a respective
phase A, B, C of the converter 10 of the invention.
[0098] In the equivalent converter configuration 100 each converter
limb 12A, 12B, 12C similarly extends between first and second DC
terminals 14, 16, and each converter limb 12A, 12B, 12C includes a
first limb portion 12A+, 12B+, 12C+ and a second limb portion 12A-,
12B-, 12C-. Each pair of first and second limb portions 12A+, 12A-,
12B+, 12B-, 12C+, 12C- in each converter limb 12A, 12B, 12C is
separated by a corresponding AC terminal 18A, 18B, 18C.
[0099] The equivalent converter configuration 100 also represents
the respective AC current demand phase waveforms I.sub.A, I.sub.B,
I.sub.C that each converter limb 12A, 12B, 12C is required to
track, e.g. match as closely as possible, and the DC current demand
I.sub.DC that the converter limbs 12A, 12B, 12C are also required
to track.
[0100] In practice each converter limb 12A, 12B, 12C also operates
within the constraints of the corresponding AC voltage phase
waveforms V.sub.A, V.sub.B, V.sub.C of the AC network 22, as well
as a DC voltage V.sub.DC of the DC electrical network 20, to which
the converter 10 is, in use, connected, and so the equivalent
converter configuration 100 may also represent these elements.
[0101] The second controller 38 is programmed to create an
equivalent converter configuration 100 which represents the flow of
current through the converter 10 by mapping possible current flow
paths through the converter 10.
[0102] One way in which the possible current flow paths through the
converter 10 may be mapped is by conducting a Kirchhoff analysis of
the equivalent converter configuration 100 obtain the following
equations:
I.sub.A=.alpha..sub.A.sup.+I.sub.A+-.alpha..sub.A.sup.-I.sub.A-
I.sub.B=.alpha..sub.B.sup.+I.sub.B+-.alpha..sub.B.sup.-I.sub.B-
I.sub.C=.alpha..sub.C.sup.+I.sub.C+-.alpha..sub.C.sup.-I.sub.C-
I.sub.DC+=.alpha..sub.A.sup.+I.sub.A++.alpha..sub.B.sup.+I.sub.B++.alpha-
..sub.C.sup.+I.sub.C+
I.sub.DC-=.alpha..sub.A.sup.-I.sub.A-+.alpha..sub.B.sup.-I.sub.B-+.alpha-
..sub.C.sup.-I.sub.C-
where [0103] the binary variables .alpha..sub.k.sup..+-. indicate
whether a given limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C- is
switched into circuit (.alpha..sub.k.sup..+-.=1) or out of circuit
(.alpha..sub.k.sup..+-.=0), i.e. .alpha..sub.A.sup.+,
.alpha..sub.A.sup.-, .alpha..sub.B.sup.+, .alpha..sub.B.sup.-,
.alpha..sub.C.sup.+, .alpha..sub.C.sup.- represent the state of the
corresponding primary switching element Sw.sub.1, Sw.sub.2,
Sw.sub.3, Sw.sub.4, Sw.sub.5, Sw.sub.6 in each limb portion 12A+,
12A-, 12B+, 12B-, 12C+, 12C- of the converter 10 (details of which
are provided to the second controller 38 by the first controller
32); [0104] I.sub.DC+ is the sum of currents in the first limb
portions 12A+, 12B+, 12C+, i.e. as shown in FIG. 3; [0105]
I.sub.DC- is the sum of currents in the second limb portions 12A-,
12B-, 12C-, i.e. as also shown in FIG. 3; and
[0105] I.sub.DC+=I.sub.DC-=I.sub.DC
[0106] The preceding equations are then combined and simplified
into
I.sub.A=.alpha..sub.A.sup.+I.sub.A+-.alpha..sub.A.sup.-I.sub.A-
I.sub.B=.alpha..sub.B.sup.+I.sub.B+-.alpha..sub.B.sup.-I.sub.B-
I.sub.C=.alpha..sub.C.sup.+I.sub.C+-.alpha..sub.C.sup.-I.sub.C-
I.sub.DC=.alpha..sub.A.sup.+I.sub.A++.alpha..sub.B.sup.+I.sub.B++.alpha.-
.sub.C.sup.+I.sub.C+
[0107] The possible current flow paths through the converter 10 are
then mapped by expressing the latter equations in a matrix form,
i.e.:
( .alpha. A + - .alpha. A - 0 0 0 0 0 0 .alpha. B + - .alpha. B - 0
0 0 0 0 0 .alpha. C + - .alpha. C - .alpha. A + 0 .alpha. B + 0
.alpha. C + 0 ) A ( I A + I A - I B + I B - I C + I C - ) x = ( I A
I B I C I DC ) b ##EQU00001##
such that A is a matrix which maps the possible current flow paths
provided by the limb portions 12A+, 12A-, 12B+, 12B-, 12C+,
12C-.
[0108] Other equivalent converter configurations and corresponding
analysis techniques are, however, also possible.
[0109] The second controller 38 is also programmed to carry out
mathematical optimization by applying a current weighting to the
relative current contribution provided by each limb portion 12A+,
12A-, 12B+, 12B-, 12C+, 12C-. The respective current weighting for
each limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C- is determined
according to measured operating parameters of the converter 10
during its operation. The various current weightings can be
determined throughout operation of the said converter 10 so as to
permit an updating of the current weightings, e.g. in response to
changing environmental conditions. As a result the various current
weightings can vary as the converter 10 operates.
[0110] During normal operation of the converter 10 an identical
current weighting is applied to each limb portion current I.sub.A+,
I.sub.A-, I.sub.B+, I.sub.B-, I.sub.C+, I.sub.C-.
[0111] However, when the converter 10 is operating under certain
conditions, e.g. an abnormal operating condition, a different
current weighting may be applied to the current contribution, i.e.
the limb portion current I.sub.A+, I.sub.A-, I.sub.B+, I.sub.B-,
I.sub.C+, I.sub.C-, provided by at least one limb portion 12A+,
12A-, 12B+, 12B-, 12C+, 12C-. For example, a larger current
weighting may be applied to the optimal limb portion current
I.sub.A+, I.sub.A-, I.sub.B+, I.sub.B-, I.sub.C+, I.sub.C- that a
particular limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C must
contribute, so as to reduce an actual limb portion current that the
said limb portion contributes relative to an actual current
contribution of each of the other limb portions, which are
otherwise all the same as one another.
[0112] In addition to the foregoing the second controller 38 is
programmed to carry out mathematical optimization to determine a
minimum individual limb portion current I.sub.A+, I.sub.A-,
I.sub.B+, I.sub.B-, I.sub.C+, I.sub.C- that each of the limb
portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C- must contribute so as to
track the corresponding required AC current demand phase waveform
I.sub.A, I.sub.B, I.sub.C and the required DC current demand
I.sub.DC-One way in which minimum individual limb portion currents
I.sub.A+, I.sub.A-, I.sub.B+, I.sub.B-, I.sub.C+, I.sub.C-, i.e. x
in the Ax=b equation set out above, may be determined, and the
aforementioned individual current weightings applied to the minimal
individual limb portion currents I.sub.A+, I.sub.A-, I.sub.B+,
I.sub.C+, I.sub.C-, is by solving a nonlinear optimization of the
general form:
min x J Current = .PSI. ( x ( t 1 ) ) + .intg. t 0 t 1 f ( x ( t )
, t ) t ##EQU00002##
subject to the equality constrained equation of the form:
Ax=b
where [0113] J.sub.Current is the current objective function to be
minimized; [0114] .PSI. is the current weighting at time t.sub.1;
[0115] f is the current cost function which in the embodiment
described includes a current weighting matrix Q.sub.1; [0116] x is
the transpose of [I.sub.A+, I.sub.A-, I.sub.B+, I.sub.B-, I.sub.C+,
I.sub.C-], i.e. [I.sub.A+, I.sub.A-, I.sub.B+, I.sub.B-, I.sub.C-]
reflected in a column vector; [0117] t.sub.0 is the time at which a
particular period of control of the converter 10 starts; and [0118]
t.sub.1 is the time at which a particular period of control of the
converter 10 ends.
[0119] The current weighting matrix Q.sub.1 is determined according
to measured operating parameters of the converter 10, and may be so
determined throughout the operation of the converter 10, such that
it can vary as the said converter 10 is controlled in response to
changes in the operation of the converter 10.
[0120] When subject only to an equality constrained equation, as
mentioned above, the Lagrangian (or the method of Lagrange
multipliers) is a technique for solving the above-identified
nonlinear optimization in order to find local minima of the current
objective function J.sub.Current. It may also be solved using other
optimization algorithms, including iterative and programming
algorithms.
[0121] As a general optimal control problem, the aforementioned
nonlinear optimization could additionally include one or more
inequality constraints in which case it could be solved by using
the further method of Hamiltonian (Pontryagin's minimum
principle).
[0122] One example of such an inequality constraint is:
C ( I A + I A - I B + I B - I C + I C - ) x .ltoreq. ( I A + max I
A - max I B + max I B - max I C + max I C - max ) d
##EQU00003##
where [0123] C is a matrix which maps possible maximum current flow
paths provided by the limb portions 12A+, 12A-, 12B+, 12B-, 12C+,
12C-; and [0124] d is a vector representing of the maximum desired
current in each limb portion 12A+, 12A-, 12B+, 12B-, 12C+,
12C-.
[0125] In either case the minimum individual limb portion currents
I.sub.A+, I.sub.A-, I.sub.B+, I.sub.B-, I.sub.C+, I.sub.C- may also
be determined by solving a nonlinear optimization of the form
max x { - J Current } . ##EQU00004##
[0126] Meanwhile, as mentioned above, the second controller 38 is
also programmed to carry out mathematical optimization to provide
an optimal limb portion voltage source V.sub.A+, V.sub.A-,
V.sub.B+, V.sub.B-, V.sub.C+, V.sub.C- for each limb portion 12A+,
12A-, 12B+, 12B-, 12C+, 12C- to achieve the corresponding
mathematically optimized minimum limb portion current I.sub.A+,
I.sub.A-, I.sub.B+, I.sub.B-, I.sub.C+, I.sub.C-. In other
embodiments of the method of the invention, however, such
mathematical optimization of the limb portion voltage sources need
not take place.
[0127] The second controller 38 is programmed to carry out
mathematical optimization to provide optimal limb portion voltage
sources V.sub.A+, V.sub.A-, V.sub.B+, V.sub.B-, V.sub.C+, V.sub.C-
by creating an equivalent converter configuration 100 which
represents voltage conditions in the converter 10.
[0128] Representing the voltage conditions in the converter 10
portrayed in the equivalent converter configuration 100
additionally includes mapping a limb portion voltage source
V.sub.A+, V.sub.A-, V.sub.B+, V.sub.B-, V.sub.C+, V.sub.C- and an
inductive component for each limb portion 12A+, 12A-, 12B+, 12B-,
12C+, 12C-.
[0129] In the embodiment described, each limb portion voltage
source V.sub.A+, V.sub.A-, V.sub.B+, V.sub.B-, V.sub.C+, V.sub.C-
corresponds to a respective chain-link converter 24A+, 24A-, 24B+,
24B-, 24C+, 24C- which is switchable into and out of the
corresponding limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C- via
operation of the corresponding primary switching element Sw.sub.1,
Sw.sub.2, Sw.sub.3, Sw.sub.4, Sw.sub.5, Sw.sub.6. Each limb portion
voltage source V.sub.A+, V.sub.A-, V.sub.B+, V.sub.B-, V.sub.C+,
V.sub.C- is therefore variable in magnitude between zero (i.e.
equivalent to being switched out of the corresponding limb portion
12A+, 12A-, 12B+, 12B-, 12C+, 12C-) and an upper voltage limit.
[0130] Meanwhile an inductive component for each limb portion 12A+,
12A-, 12B+, 12B-, 12C+, 12C- within the equivalent converter
configuration 100 represents the inductance associated with the
corresponding limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C- of
the actual converter 10. Such inductances do not include a passive
limb inductor within each limb portion 12A+, 12A-, 12B+, 12B-,
12C+, 12C- since they are no longer needed to control the level of
circulating current between the converter limbs 12A, 12B, 12C.
Instead, the respective inductances take the form of a phase
inductance 44A, 44B, 44C and a DC line inductance 46 (each of which
may be made up of a physical passive inductor component and any
stray inductance within the associated electrical structure of the
converter), and a very small remaining stray inductance within the
corresponding limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C-.
[0131] The aforementioned inductive component of each limb portion
12A+, 12A-, 12B+, 12B-, 12C+, 12C- is represented in the equivalent
converter configuration 100 as an inductive voltage portion
U.sub.A+, U.sub.A-, U.sub.B+, U.sub.B-, U.sub.C+, U.sub.C- that is
made up of the voltage arising from the flow of current through the
aforementioned inductance, i.e. phase inductance 44A, 44B, 44C and
DC line inductance 46 only, associated with a corresponding limb
portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C-.
[0132] In other embodiments of the invention representing of the
voltage conditions in the converter 10 may additionally include
mapping a resistive component for each limb portion 12A+, 12A-,
12B+, 12B-, 12C+, 12C-.
[0133] Such a resistive component represents the resistance
associated with a given limb portion 12A+, 12A-, 12B+, 12B-, 12C+,
12C-, and similarly may take the form of a resistor within a given
limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C-, i.e. a limb
portion resistance, or the form of a resistance electrically
associated with a given limb portion 12A+, 12A-, 12B+, 12B-, 12C+,
12C-, e.g. a phase resistance and/or a DC line resistance.
[0134] Mapping the limb portion voltage sources V.sub.A+, V.sub.A-,
V.sub.B+, V.sub.B-, V.sub.C+, V.sub.C- and inductive voltage
portions U.sub.A+, U.sub.A-, U.sub.B+, U.sub.B-, U.sub.C+, U.sub.C-
again similarly includes conducting a Kirchhoff analysis of the
equivalent converter configuration 100, although other equivalent
converter configurations and corresponding analysis techniques are
also possible. In applying the Kirchhoff analysis the following
equation, in matrix form, is obtained:
M V ( V A + V A - V B + V B - V C + V C - ) - M U ( U A + U A - U B
+ U B - U C + U C - V DC V AB V CB ) = ( 0 0 0 0 0 )
##EQU00005##
where:
M V = ( 1 1 - 1 - 1 0 0 0 0 1 1 - 1 - 1 0 0 0 0 1 1 1 0 - 1 0 0 0 0
0 1 0 - 1 0 ) ##EQU00006##
i.e. M.sub.V is a matrix which maps the position of the limb
portion voltage sources V.sub.A+, V.sub.A-, V.sub.B+, V.sub.B-,
V.sub.C+, V.sub.C- within the particular converter structure;
M U = ( 1 - 1 - 1 1 0 0 0 0 0 0 0 1 - 1 - 1 1 0 0 0 0 0 0 0 1 - 1 -
1 0 0 1 0 - 1 0 0 0 0 1 0 0 0 1 0 - 1 0 0 0 - 1 ) ##EQU00007##
i.e. M.sub.u is a matrix which maps the position of the inductive
voltage portions U.sub.A+, U.sub.A-, U.sub.B+, U.sub.B-, U.sub.C+,
U.sub.C- within the particular converter structure; [0135] V.sub.DC
is the DC voltage, i.e. the voltage difference between the first
and second DC terminals 14, 16; [0136] V.sub.AB is the voltage
difference between the first and second converter limbs 12A, 12B;
and [0137] V.sub.CB is the voltage difference between the third and
second converter limbs 12C, 12B.
[0138] The second controller 38 is further programmed to carry out
mathematical optimization to provide an optimal limb portion
voltage source V.sub.A+, V.sub.A-, V.sub.B+, V.sub.B-, V.sub.C+,
V.sub.C- for each limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C-
by reducing any deviation in an actual measured limb portion
current I'.sub.A+, I'.sub.A-, I'.sub.B+, I'.sub.B-, I'.sub.C+,
I'.sub.C- of a given limb portion 12A+, 12A-, 12B+, 12B-, 12C+,
12C- from the corresponding determined optimal limb portion current
I.sub.A+, I.sub.A-, I.sub.B+, I.sub.B-, I.sub.C+, I.sub.C- for the
said given limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C-.
[0139] The second controller 38 is still further programmed to
calculate the inductive voltage portion U.sub.A+, U.sub.A-,
U.sub.B+, U.sub.B-, U.sub.C+, U.sub.C- for each limb portion 12A+,
12A-, 12B+, 12B-, 12C+, 12C-. This calculation is based on the
corresponding determined optimal limb portion currents I.sub.A+,
I.sub.A-, I.sub.B+, I.sub.B-, I.sub.C+, I.sub.C-, together with the
inductance associated with the corresponding limb portion 12A+,
12A-, 12B+, 12B-, 12C+, 12C-.
[0140] Thereafter the calculated inductive voltage portion
U.sub.A+, U.sub.A-, U.sub.B+, U.sub.B-, U.sub.C+, U.sub.C- is
modified to drive the actual measured limb portion current
I'.sub.A+, I'.sub.A-, I'.sub.B+, I'.sub.B-, I'.sub.C+, I'.sub.C- to
follow the corresponding determined optimal limb portion current
I.sub.A+, I.sub.A-, I.sub.B+, I.sub.B-, I.sub.C+, I.sub.C-.
[0141] Such measuring and modification takes the form of a feedback
loop which provides closed-loop control, as illustrated
schematically by a second process box 48 in the first flow diagram
40 shown in FIG. 4. The feedback loop may additionally include a
feed-forward element which seeks to predict desirable future values
for one or more of the inductive voltage portions U.sub.A+,
U.sub.A-, U.sub.B+, U.sub.B-, U.sub.C+, U.sub.C- in order to
improve the performance of the closed-loop control.
[0142] The calculated inductive voltage portion U.sub.A+, U.sub.A-,
U.sub.B+, U.sub.B-, U.sub.C+, U.sub.C- of each limb portion 12A+,
12A-, 12B+, 12B-, 12C+, 12C- is utilised, as indicated by a third
process box 50 in the first flow diagram 40, when carrying out the
aforementioned mathematical optimization to provide the optimal
limb portion voltage sources V.sub.A+, V.sub.A-, V.sub.B+,
V.sub.B-, V.sub.C+, V.sub.C-. Such mathematical optimization also
includes applying a voltage weighting to the relative voltage
contribution provided by each limb portion voltage source V.sub.A+,
V.sub.A-, V.sub.B+, V.sub.B-, V.sub.C+, V.sub.C-. The voltage
weightings are determined according to measured operating
parameters of the converter 10, and may be so determined throughout
operation of the said converter 10. Such potentially repeated
determination of the voltage weightings permits the ongoing
optimization of the converter operation during, e.g. changing
environmental conditions.
[0143] For example, during normal operation of the said particular
converter structure an identical voltage weighting is applied to
the limb portion voltage source V.sub.A+, V.sub.A-, V.sub.B+,
V.sub.B-, V.sub.C+, V.sub.C- of each limb portion 12A+, 12A-, 12B+,
12B-, 12C+, 12C-.
[0144] However during, e.g. abnormal operating conditions, a
different voltage weighting can be applied to the limb portion
voltage source V.sub.A+, V.sub.A-, V.sub.B+, V.sub.B-, V.sub.C+,
V.sub.C- of one or more limb portions 12A+, 12A-, 12B+, 12B-, 12C+,
12C- to further alleviate the impact of, e.g. the abnormal
operating conditions.
[0145] More particularly, the second controller 38 is programmed to
carry out mathematical optimization to provide an optimal limb
portion voltage source V.sub.A+, V.sub.A-, V.sub.B+, V.sub.B-,
V.sub.C+, V.sub.C- for each limb portion 12A+, 12A-, 12B+, 12B-,
12C+, 12C- by determining a minimum individual limb portion voltage
source V.sub.A+, V.sub.A-, V.sub.B+, V.sub.B-, V.sub.C+, V.sub.C-
for each limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C- that is
required to achieve the corresponding minimum limb portion current
I.sub.A+, I.sub.A-, I.sub.B+, I.sub.B-, I.sub.C+, I.sub.C-
previously determined.
[0146] One way in which minimum individual limb portion voltage
sources V.sub.A+, V.sub.A-, V.sub.B+, V.sub.B-, V.sub.C+, V.sub.C-
(i.e. the minimum level of voltage a variable voltage source within
a given limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C- must
provide) may be determined, and the aforementioned individual
voltage weightings applied thereto, is by solving for x (where x is
the transpose of [V.sub.A+, V.sub.A-, V.sub.B+, V.sub.B-, V.sub.C+,
V.sub.C-]) a nonlinear optimization of the general form:
min x J Voltage = .PSI. ( x ( t 1 ) ) + .intg. t 0 t 1 f ( x ( t )
, t ) t ##EQU00008##
subject to an equality constrained equation M.sub.Vx=b, where b is
known, of the form:
M V ( V A + V A - V B + V B - V C + V C - ) x = M U ( U A + U A - U
B + U B - U C + U C - V DC V AB V CB ) b ##EQU00009##
and where [0147] J.sub.Voltage is the voltage objective function to
be minimized; [0148] .PSI. is the voltage weighting at time t.sub.1
[0149] f is the voltage cost function which in the embodiment
described includes a voltage weighting matrix Q.sub.V; [0150]
t.sub.0 is the time at which a particular period of control of the
converter 10 starts; and [0151] t.sub.1 is the time at which a
particular period of control of the converter 10 ends.
[0152] The voltage weighting matrix Q.sub.V is similarly determined
according to measured operating parameters of the converter 10, and
may be so determined throughout the operation of the converter 10.
As such it too can vary as the said converter 10 is controlled.
[0153] Solving the nonlinear optimization mentioned above may also
be made subject to an inequality equation of the form:
Cx.ltoreq.d
where [0154] C is a matrix which maps the position of possible
maximum limb portion voltage sources in the limb portions 12A+,
12A-, 12B+, 12B-, 12C+, 12C-; and [0155] d is a vector representing
of the maximum desired voltage in each limb portion 12A+, 12A-,
12B+, 12B-, 12C+, 12C-.
[0156] An alternative second controller (not shown), which may
instead be included in the converter 10, is similarly programmed
to: [0157] (a) obtain a respective AC current demand phase waveform
I.sub.A, I.sub.B, I.sub.C for each converter limb 12A, 12B, 12C
which the corresponding converter limb 12A, 12B, 12C is required to
track, and a DC current demand I.sub.DC which each converter limb
12A, 12B, 12C is also required to track; and [0158] (b) carry out
mathematical optimization to determine an optimal limb portion
current I.sub.A+, I.sub.A-, I.sub.B+, I.sub.B-, I.sub.C+, I.sub.C-
for each limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C- that the
limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C- must contribute to
track the corresponding required AC current demand phase waveform
I.sub.A, I.sub.B, I.sub.C and the required DC current demand
I.sub.DC.
[0159] The aforementioned steps are again, similarly indicated by a
first process box 42 in a second flow diagram 60 shown in FIG.
6(a).
[0160] Thereafter, however, the alternative second controller is
programmed to apply a control algorithm to directly establish
optimal limb portion voltage sources V.sub.A+, V.sub.A-, V.sub.B+,
V.sub.B-, V.sub.C+, V.sub.C- from each of the corresponding
determined minimum limb portion current I.sub.A+, I.sub.A-,
I.sub.B+, I.sub.B-, I.sub.C+, I.sub.C-, i.e. as shown by a single
fourth process box 62 in the second flow diagram 60.
[0161] Applying such a control algorithm includes reducing any
deviation in the actual measured limb portion current I'.sub.A+,
I'.sub.A-, I'.sub.B+, I'.sub.B-, I'.sub.C+, I'.sub.C- of a given
limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C- from the
corresponding determined minimum limb portion current I.sub.A+,
I.sub.A-, I.sub.B+, I.sub.B-, I.sub.C+, I.sub.C- for the said given
limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C-.
[0162] One way in which the deviation in the actual measured limb
portion current I'.sub.A+, I'.sub.A-, I'.sub.B+, I'.sub.B-,
I'.sub.C+, I'.sub.C- of a given limb portion 12A+, 12A-, 12B+,
12B-, 12C+, 12C- from the corresponding determined minimum limb
portion current I.sub.A+, I.sub.A-, I.sub.B+, I.sub.B-, I.sub.C+,
I.sub.C- may be reduced, and preferably eliminated, is by
establishing a feedback loop 70 as shown schematically in FIG.
6(b).
[0163] In the embodiment shown the feedback loop 70 compares a
respective actual measured limb portion current I'.sub.A+,
I'.sub.A-, I'.sub.B+, I'.sub.B-, I'.sub.C+, I'.sub.C- with the
corresponding determined minimum limb portion current I.sub.A+,
I.sub.A-, I.sub.B+, I.sub.B-, I.sub.C+, I.sub.C- and calculates a
corresponding limb portion error e.sub.A+, e.sub.A-, e.sub.B+,
e.sub.B-, e.sub.C+, e.sub.C-. The feedback loop 70 then applies a
correction factor K to each limb portion error e.sub.A+, e.sub.A-,
e.sub.B+, e.sub.B-, e.sub.C+, e.sub.C- to thereby establish
directly the corresponding limb portion voltage source V.sub.A+,
V.sub.A-, V.sub.B+, V.sub.B-, V.sub.C+, V.sub.C- which is required
to drive the error e.sub.A+, e.sub.A-, e.sub.B+, e.sub.B-,
e.sub.C+, e.sub.C- towards zero.
[0164] The correction factor K may take the form of a control
system matrix, such as a gain matrix (not shown), which sets out
individual correction factors that each limb portion error
e.sub.A+, e.sub.A-, e.sub.B+, e.sub.B-, e.sub.C+, e.sub.C- is, e.g.
multiplied by in the case of a gain matrix, to establish the
corresponding limb portion voltage source V.sub.A+, V.sub.A-,
V.sub.B+, V.sub.B-, V.sub.C+, V.sub.C-.
[0165] One way in which such individual correction factors may be
established is by creating an equivalent converter configuration
that represents the voltage conditions in the particular
three-phase converter structure under control and thereafter
considering the dynamics of such an equivalent converter
configuration.
[0166] More particularly, in relation to the embodiment described
hereinabove, the foregoing steps may be achieved by creating the
equivalent converter configuration 100 shown in FIG. 5 and mapping
a limb portion voltage source V.sub.A+, V.sub.A-, V.sub.B+,
V.sub.B-, V.sub.C+, V.sub.C- and an inductive component for each
limb portion 12A+, 12A-, 12B+, 12B-, 12C+, 12C- in the equivalent
converter configuration 100.
[0167] Thereafter such mapping may include conducting a Kirchhoff
analysis of the equivalent converter configuration 100 (although
other equivalent converter configurations and corresponding
analysis techniques are also possible) that applies Kirchhoff's
current and voltage laws to describe the dynamics of the equivalent
converter configuration 100 as:
v = M I t + N .xi. ##EQU00010##
where: [0168] v is the transpose of [V.sub.A+, V.sub.A-, V.sub.B+,
V.sub.B-, V.sub.C+, V.sub.C-], [0169] M is a coupled-inductance
matrix which maps the inductive component of each limb portion, and
more particularly maps each of the phase and DC line inductances
associated with each limb portion; e.g.
[0169] M = 10 - 3 ( 3 - 2 1 0 1 - 1 2 1 0 1 1 0 3 - 2 1 1 0 - 1 2 1
- 1 2 - 1 2 1 3 - 2 3 - 2 1 ) ; ##EQU00011## [0170] I is the
transpose of [I'.sub.A+, I'.sub.A-, I'.sub.B+, I'.sub.B-,
I'.sub.C+, I'.sub.C- ], i.e. the transpose of a currents vector
representing the actual measured limb portion currents I'.sub.A+,
I'.sub.A-, I'.sub.B+, I'.sub.B-, I'.sub.C+, I'.sub.C-; [0171] N is
an input voltage matrix which maps the position of various input
voltages within the particular converter structure, e.g.
[0171] N = 1 6 [ 3 - 4 2 3 4 - 2 3 2 2 3 - 2 - 2 3 2 - 4 3 - 2 4 ]
; ##EQU00012##
and [0172] .xi. is an input voltages vector representing external
disturbances, e.g.
[0172] .xi. = [ V DC V AB V CB ] ##EQU00013##
where [0173] V.sub.DC is the DC voltage, i.e. the voltage
difference between the first and second DC terminals 14, 16; [0174]
V.sub.AB is the voltage difference between the first and second
converter limbs 12A, 12B; and [0175] V.sub.CB is the voltage
difference between the third and second converter limbs 12C,
12B.
[0176] In this way, conducting the aforesaid Kirchhoff analysis
makes it possible to take into account all of the factors mentioned
above relating to the converter 10, i.e. M, I, N, .xi., when
considering what impact a change in one or more individual limb
portion voltage sources V.sub.A+, V.sub.A-, V.sub.B+, V.sub.B-,
V.sub.C+, V.sub.C- will have, e.g. on the actual measured limb
portion current I'.sub.A+, I'.sub.A-, I'.sub.B+, I'.sub.B-,
I'.sub.C+, I'.sub.C-. This ability renders the alternative second
controller robust against controller uncertainties and modelling
errors.
[0177] Moreover, as a result it is possible thereafter to establish
each of the individual correction factors by considering what
change needs to be made to a given individual limb portion voltage
source V.sub.A+, V.sub.A-, V.sub.B+, V.sub.B-, V.sub.C+, V.sub.C-
to desirably alter the corresponding limb portion current provided
by the converter 10 under control, i.e. the corresponding actual
measured limb portion current I'.sub.A+, I'.sub.A-, I'.sub.B+,
I'.sub.B-, I'.sub.C+, I'.sub.C-, in order to drive the actual
measured limb portion current I'.sub.A+, I'.sub.A-, I'.sub.B+,
I'.sub.B-, I'.sub.C+, I'.sub.C- towards the determined minimum limb
portion current I.sub.A+, I.sub.A-, I.sub.B+, I.sub.B-, I.sub.C+,
I.sub.C-, i.e. in order to reduce the corresponding limb portion
error e.sub.A+, e.sub.A-, e.sub.B+, e.sub.B-, e.sub.C+, e.sub.C-
towards zero.
[0178] Once such individual correction factors have been
established for the converter 10, e.g. at initial design and
commissioning stages, there is not normally a need to determine
them again. As a result the feedback loop 70 involves minimal
computational effort since in each cycle it is simply required to
multiply a given limb portion error e.sub.A+, e.sub.A-, e.sub.B+,
e.sub.B-, e.sub.C+, e.sub.C- by the corresponding individual
correction factor which has already been determined.
* * * * *