GB2429306A

GB2429306A - A controller for a parallel connected inverter

Info

Publication number: GB2429306A
Application number: GB0516827A
Authority: GB
Inventors: Jamie Kelly
Original assignee: Turbo Genset Co Ltd
Current assignee: Turbo Genset Co Ltd
Priority date: 2005-08-17
Filing date: 2005-08-17
Publication date: 2007-02-21
Anticipated expiration: 2025-08-17
Also published as: GB2429306B; GB0516827D0

Abstract

A controller 2 for regulating an inverter 13 that produces one or more phases 15 of AC power derives a phase frequency reference signal 23 and resolves relative to it both in-phase and quadrature phase components of current and voltage. The phase and quadrature components are compared to target values and inverter 13 is controlled in response. The target values of direct and quadrature components are determined from a desired droop characteristic for the generator. Controller 2 enables inverter 13 to be parallel connected to another inverter, a generator, a converter or a distribution network. Phase reference signal 23 may be derived from the mean phase of output voltages 15 by a phase locked loop 3.

Description

A controller for a parallel connected inverter The present invention

relates to a controller for sharing power between parallel connected AC voltage output inverters.

Electrical power is generally provided throughout a region by a regionally distributed grid network. The network is fed by a number of electricity generation stations which are spread out within the region. A single generation station will typically have a capacity of between 100MW and 2GW. If a load is far from an electricity generation source and the capacity of the grid network at this load is approaching its limit, then the network at the load is described as a weak grid', which can make the grid network at this load susceptible to a temporary drop in voltage, known as a brown out', or a complete loss of power, known as a blackout', which may result from a surge in power demand. For this reason it is often necessary to prop-up' the distributed grid network in such areas by providing smaller electricity generation facilities, to increase the capacity and hence the reliability of the network.

In some situations a stand-alone power source may be required, for example in a location where the distributed grid network is not present. Such a power source may be comprised of one or more generators connected in parallel. In cases where a load exceeds the power rating of the existing stand-alone power source, rather than replacing it with a larger rated device, it can be more convenient to add another power source in parallel with the original power source. In cases where reliability is paramount, it is preferable to supply a load from a number of parallel connected power sources to form a local grid network, so that failure of one or more of the power sources leaves sufficient capacity to maintain the supply. Paralleling can lead to power sources being treated as more of an off-the-shelf item, where, within reason, a load is supplied by a number of stand-alone power sources. Deploying a number of power sources can be advantageous in terms of reliability and resilience to fluctuations in capacity, allowing the generators to share the load and introducing redundancy to make the local grid network more robust to failure of an individual generating unit.

A stand-alone generator powered by a conventional energy source may take the form of a high-speed axial field generator directly driven by a gas turbine engine, with a power rating in the region of 500kW. The gas turbine is a compact power source, capable of utilising different types of fuels. Best performance is achieved at stable high speed of the order of 30,000rpm, so the ideal generator would also operate at high speed to avoid having to include a reduction gearbox. The electrical output of generators of this type is high frequency AC, which is unsuitable for direct connection to the regional or local grid network. Therefore the AC output is rectified to DC before being converted by inverters to grid frequency AC. An advantage of AC to DC conversion followed by DC to AC conversions is that the frequency of the inverter output is isolated from fluctuations in the frequency of the generator windings output, which would occur as a result of variations in the turbine speed.

When a load is applied to a non-ideal power source such as the standalone generator described above, the output voltage of the voltage source will drop as current which flows through the load causes a voltage drop across the internal impedance of the power source. For an AC power source the phase of the output voltage may also shift where the load is inductive. This phenomenon is known as droop, and the droop characteristics of a real power source can be tested and recorded. For a real system the amount of drop in the magnitude of the supply voltage, known as the regulation', is typically set as a percentage of the maximum open circuit voltage.

A basic requirement when connecting power source inverters in parallel is that each inverter supplies its correct share of the load. The voltage of the AC output of an inverter should match the voltage of and be in phase with the output voltage of the other inverters and/or the grid network supply. The current of the AC output of the inverter should be in phase with the output current of the other inverters and/or the grid network supply, but will not necessarily have the same magnitude as the output current of the other inverters. Assuming that the inverters are meant to proportionally share the load, then if two 100kW rated inverters were parallel connected, and the load was 150kW, then each inverter should supply half of the power i.e. 75kW, by supplying equal current. If a 100kW rated inverter is parallel connected to a 200kW rated inverter to supply a 180kW load, then the 100kW rated inverter should supply 60kW and the 200kW rated inverter should supply 120kW, by supplying twice as much current as the 100kW rated inverter. In this way both are supplying 60% of their rated capacity. This proportional sharing can be managed by the inverters' controllers.

A situation may arise when the load is not being shared proportionally between the inverters. If the higher rated inverter is trying to produce an output voltage greater than the common voltage, it dumps' a larger current into the load and provides more than its proportional share of the power. This situation is inefficient and does not utilise the available resources appropriately and is to be avoided by tailoring the output currents and voltages.

A multi-phase system is said to be balanced when the real power and vars of each phase are identical. Conversely, in an unbalanced system the real power and/or vars of one phase differ from the real power and/or vars of the other phases. If each phase produces the same vars then there will be equal phase shifts between the output voltages, e.g. for a three-phase supply the phases will be separated by 120 .

When inverters are parallel connected, the output power connections provide a channel of communication. Various modes of communication may be employed, as long as they do not interfere with the basic requirement of the inverter to produce output voltages within defined limits (amplitude, harmonics, frequency etc.). When parallel connected, the voltage and frequency are common to all the inverters. In a three-phase system, the four variables to be controlled are the three voltages and frequency, since the frequency is common to all phases. In a balanced three phase system, only the real power and the vars need to be controlled, so it is seen that using these four variables gives enough control capability.

In the case of an unbalanced three phase system, three real power quantities and three var quantities need to be controlled. The four measured variables mentioned above do not provide enough information to control these six quantities. One approach is to add a further communication channel, independent of the power connections. This can bring many advantages, though it is an addition which adds cost, leads to difficulties in standardisation and may adversely affect reliability. There are master/slave approaches, where one inverter may act as a voltage source while other inverters supply current, but this would also require an extra communication channel.

Therefore, there is a need to provide an inverter which can link to a distributed grid network, supplying controlled real power and vars to a grid and also operate as a stand- alone three phase supply.

Therefore, according to the present invention there is provided a controller for an AC power source having one or more phase outputs, the controller comprising: voltage measuring means for measuring the voltage of the or each phase output; current measuring means for measuring the current of the or each phase output; frequency reference means for generating a phase frequency reference for the or each phase output; processing means for resolving the measured voltage of the or each phase output into direct and quadrature voltage components and for resolving the measured current of the or each phase output into direct and quadrature current components, wherein the direct components of the or each phase output are the components which are in phase with the phase frequency reference of the respective phase output and the quadrature components of the or each phase output are the components which are 90 degrees out of phase with the phase frequency reference of the respective phase output; error calculation means for comparing the direct and quadrature voltage components and the direct and quadrature current components of the or each phase output with target direct and quadrature components respectively to calculate direct and quadrature error values for the or each respective phase output; and control signal means for generating control signals for controlling the or each phase output of the power source, based upon said respective direct and quadrature error values.

The invention makes use of the droop characteristics of the grid network to correctly balance the real power and vars produced by the power supply for unbalanced three phase loads. No inverter in a network must act as a master, which helps to avoid single point failures.

If the magnitude of a phase voltage is measured, it will yield only one quantity, but by obtaining the Fourier components, A1 and B1, shown in the following equations, further information is obtained: A1 J-.JV(9). cos(9).d9 B1 =-.fV(G).sin(o).a'o The phase value, 0, of each voltage, is preferably based upon the frequency of the signal on the power output in a single phase system or the average frequency of each of the phases in a multi-phase system. Seven points of reference yield six variables, sufficient to fully control the balance of the real power and vars in an inverter.

Preferably the inverter would be capable of parallel connection with the minimum of communication when in the stand-alone mode and be capable of seamless changeover', i.e. it should be able to continue supplying power to a load in the event that the grid network supply fails. Advantageously, the inverter is able to reconnect to the grid network when the grid supply is re-established, is capable of operating with unbalanced supplies, whether grid connected or as a stand-alone supply and is capable of direct connection to the grid, without the need for a transformer. Preferably the inverter has built in redundancy, for better fault tolerance and detection of faults. Preferably the inverter is operated as either a three-leg or a four-leg inverter, with the ability to switch while supplying power.

Preferably the target direct and quadrature components are determined based upon desired droop characteristics of the power source.

Preferably the droop characteristics are comprised of a direct droop constant and a quadrature droop constant, wherein the direct droop constant is the desired change in the direct voltage component from an open circuit direct voltage component, per unit of direct current component, and the quadrature droop constant is the desired change in the quadrature voltage component from an open circuit quadrature voltage component, per unit of quadrature current component.

The present invention controls the droop rather than allowing it to depend upon the characteristics of the device. In this way, power sharing can be controlled effectively across a network.

Preferably the control signal means comprises means for summing the direct and quadrature error values for the or each phase output to produce cumulative direct and quadrature error values for the or each phase output.

Preferably the control signal means comprises conversion means for converting the cumulative direct and quadrature error values of the or each phase output into control signals for the or each phase output.

Preferably the conversion means comprises: a first and second multiplier; a combiner; and signal generator means for producing a sinusoidal wave of frequency equal to the frequency of the phase frequency reference and a shifted sinusoidal wave which is shifted by 90 degrees relative to the sinusoidal wave, wherein the first multiplier multiplies the sinusoidal wave with the cumulative direct error value and the second multiplier multiplies the shifted sinusoidal wave with the cumulative quadrature error value, and the combiner combines the outputs of the first and second multipliers.

Preferably the frequency reference means comprises a phase locked loop.

Where there are a plurality of phase outputs, preferably the frequency reference means comprises averaging means for averaging the phase of the voltages of the phase outputs for generating a primary phase frequency reference.

Preferably the frequency reference means comprises phase shifting means for generating the phase frequency reference for each of the phase outputs by applying a respective phase shift to the primary phase frequency reference, the phase shift depending on desired phase spacing between the voltages of the phase outputs.

A further aspect of the present invention provides an AC output voltage power source comprising a DC power source, an inverter and a controller as described above for controlling the inverter.

A further aspect of the present invention provides a controller for a multi-phase AC power source having a plurality of phase outputs, the controller comprising: voltage measuring means for measuring the voltage of each of the phase outputs; current measuring means for measuring the current of each of the phase outputs; frequency reference means for generating a phase frequency reference for each of the phase outputs; processing means for resolving the measured voltage of each of the phase outputs into direct and quadrature voltage components and for resolving the measured current of each of the phase outputs into direct and quadrature current components, wherein the direct components of each of the phase outputs are the components which are in phase with the phase frequency reference of the respective phase output and the quadrature components of each of the phase outputs are the components which are 90 degrees out of phase with the phase frequency reference of the respective phase output; error calculation means for comparing the direct and quadrature voltage components and the direct and quadrature current components of each of the phase outputs with target direct and quadrature components respectively to calculate direct and quadrature error values for each of the respective phase outputs; and control signal means for generating control signals for controlling each of the phase outputs of the power source, based upon said respective direct and quadrature error values.

A further aspect of the present invention provides a method of controlling an AC output power source having one or more phase outputs, the method comprising: measuring the voltage of the or each phase output; measuring the current of the or each phase output; generating a phase frequency reference for the or each phase output; resolving the measured voltage of the or each phase output into direct and quadrature voltage components and resolving the measured current of the or each phase output into direct and quadrature current components, wherein the direct components of the or each phase output are the components which are in phase with the phase frequency reference of the respective phase output and the quadrature components of the or each phase output are the components which are 90 degrees out of phase with the phase frequency reference of the respective phase output; comparing the direct and quadrature voltage components and the direct and quadrature current components of the or each phase output with target direct and quadrature components respectively and calculating direct and quadrature error values for the or each respective phase output; and generating control signals for controlling the or each phase output of the power source, based upon said respective direct and quadrature error values.

Preferably generating control signals further comprises summing the direct and quadrature error values for the or each phase output to produce cumulative direct and quadrature error values for the or each phase output.

Preferably generating control signals further comprises converting the cumulative direct and quadrature error values of the or each phase output into control signals.

Preferably converting the cumulative direct and quadrature error values of the or each phase output into control signals comprises: producing a sinusoidal wave of frequency equal to the frequency of the phase frequency reference; producing a shifted sinusoidal wave which is shifted by 90 degrees relative to the sinusoidal wave; multiplying the sinusoidal wave and the shifted sinusoidal wave with the cumulative direct and quadrature error values respectively; and combining the multiplied outputs.

Where there are a plurality of phase outputs, preferably generating a phase frequency reference comprises averaging the phase of the voltages of the phase outputs to generate a primary phase frequency reference.

Preferably generating a phase frequency reference for each of the phase outputs comprises applying a phase shift to the primary phase frequency reference, the respective phase shift depending on the desired phase spacing between the voltages of the phase outputs.

The present invention can be implemented either in hardware or on software in a general purpose computer. Further the present invention can be implemented in a combination of hardware and software. The present invention can also be implemented by a single processing apparatus or a distributed network of processing apparatuses.

Since the present invention can be implemented by software, the present invention encompasses computer code provided to a general purpose computer or processor on any suitable carrier medium. The carrier medium can comprise any storage medium such as a floppy disk, a CD ROM, a magnetic device or a programmable memory device, or any transient medium such as any signal e.g. an electrical, optical or microwave signal.

The invention will be further described by way of example, with reference to the accompanying drawings, in which: Figure 1 is a block diagram of an embodiment of an inverter and a controller for the inverter according to the present invention; Figure 2 is a block diagram of the phase locked loop block, PLL, of Figure 1; Figure 3 is a block diagram of the target current control blocks ITGTX, of Figure 1; Figure 4 is a block diagram of the real time to direct-quadrature (DQ) converter block, RT-DQ, of Figure 3; Figure 5 is a block diagram of the DQ to real time converter block, DQRT of Figure 3; and Figure 6 is a block diagram of the four-leg inverter and output filters of Figure 1.

Figure 1 shows an overview of an inverter system 1 incorporating a fourleg inverter 13 and a controller 2 for the inverter 13 according to the present invention. The main system blocks are a phase locked loop (PLL) 3, three target current control blocks, ITGTa 5, ITGTb 7 and ITGTC 9, a current loop and pulse width modulation (PWM) section 11 and a four-leg inverter and output filter block 13.

The inverter system 1 has four outputs 14, 15, namely a neutral terminal 14 and the three phase outputs A, B and C 15. The current out of each of these phase outputs A, B and C 15 is measured and signals a, lb and L representative of the output current values are fed back as inputs 16 to the system 1. The voltage of each of the three phase outputs A, B and C 15 is measured relative to the neutral terminal 14 and signals Van, Vbfl and V, representative of these values are also fed back as inputs 16 to the system 1.

A frequency reference generator 17 generates a frequency bias value fbI to provide a signal representative of a frequency to the PLL 3.

The system 1 operates in the following manner. The signals Van, Vbn and V representing the voltages on the three phase outputs A, B and C 15 of the inverter block 13 are monitored by the PLL 3 (Figure 2) to recover the average phase of the voltages of the phase outputs and hence of the grid to which the inverter is connected. The PLL 3 produces a primary phase frequency reference ePLL 23 which is used in the three target current control blocks ITGTa 5, 1TGT 7 and ITGTC 9. The primary phase frequency reference PLL 23 is fed directly into the phase A target current control block ITGTa 5.

A delay of 120 degrees is applied to 8PLL to provide a shifted phase frequency reference (OPLL-l20) which is fed into the phase B target current control block, ITGTb 7. A further 120 degrees delay is applied to (OPLL-l2O) to provide a second shifted phase reference (OPLL-240) which is fed into the phase C target current control block, ITGTC 9. The three phase references 8PLL, (OPLL-l2O) and (OPLL-240) are generically referred to a PLL.

Each of the three target current control blocks, JTGTa 5, ITGTb 7 and ITGTC 9 receives a signal Vxn (Van, Vb or V) representative of the voltage and a signal I, (Ii, lb or J) representative of the current of one of the three phase outputs A, B and C respectively.

Figure 5 shows a generic example of the target current control block, ITGTX. The output Jxtgt from each target current control block, ITGTX is fed into the current loop and PWM section 11 which produces duty cycle drive signals Da, Db, D and D, which are used to control the four-leg inverter 13 of Figure 6.

Figure 2 is a block diagram of the PLL 3 of Figure 1. If two or more inverters are connected in parallel, each inverter has its own PLL 3 which provides a primary phase frequency reference for the controller of the inverter by locking to the frequency of the paralleled supply and by averaging the phase of the voltages phase outputs A, B and C 15. Each PLL 3 will be biased to the same nominal reference frequency fb representing the desired AC frequency of the supply, normally 50Hz or 60Hz, within the tolerance of the individual frequency reference source. When all the PLLs 3 lock to one another, they will run collectively at the average centre frequency of the collection.

The phase error between one PLL 3 and another can be kept at an arbitrarily small level by selection of more accurate frequency reference sources, and each inverter can be considered to be in phase with one another. The frequency reference source in the variable frequency oscillator (VFO) 39 is typically a quartz crystal, which can be readily obtained at an accuracy of I Oppm.

In the PLL 3, signals Van, Vbn and V representative of the voltages of the three phase outputs A, B and C 15 are fed into a mains phase detector 31 which outputs the average detected phase em of the voltages. This detected phase is fed into a phase comparator 33 which compares the measured phase em with the primary phase frequency reference output from the PLL 3, ePLL. The phase difference between these two signals is calculated and fed into an amplifier 35 along with a frequency bias signal fbI 37. The amplifier 35 drives a VFO 39 which produces the primary phase frequency reference signal ePLL.

Figure 3 is a block diagram of a target current control block ITGT (ITGTa 5, ITGTb 7, ITGTC 9) of Figure 1. There is a target current control block ITGTX 5, 7, 9 for each phase output A, B and C 15 which are provided with phase frequency references ePLL, (ePLL- 120 ) and (ePLL-240 ) respectively. Reference values for the direct voltage Vdd and quadrature voltage Vqd are also provided. These reference values are the nominal direct and quadrature values of the voltage V, when no load is applied, and for a 240V rms system would typically be as shown in Table 1.

Phase Vdd Vqd A 240 0 B -120 -207.8 (-120.']3) C -120 207.8(12O./3) I Table 1 - typical values of Vdd and Vqd for 240 V rms system Each input V, , I, to the target current control block ITGTX 5, 7, 9 is fed into its own real time to direct-quadrature converter block, RT-DQ 51, 53.

The structure of the real time to direct-quadrature converter block, RTDQ 51, 53 is shown in Figure 4. A signal Y, is fed into the RT-DQ block 51, 53 and is split into two identical paths 43, 45. The signal Y, represents either a voltage V, or a current I,, both of which vary sinusoidally with frequency equal to the frequency of Om. In one path 43 the signal Y, is multiplied in multiplier 47 by a time varying cosine waveform of frequency equal to the frequency of ePLL and the signal in the other path 45 is multiplied in multiplier 49 by a time varying sine waveform of frequency equal to the frequency of e PLL. The frequency spectrum of the output of the multiplier will have a DC component corresponding to the frequency difference component (em - ePLL 0) and a frequency sum component with a frequency of(em + OPLL). The output of the multipliers 47, 49 is filtered by a low-pass filter to remove the frequency sum component, giving Yd and Yq which are DC values representing the magnitude of the direct Yd and quadrature Yq components of Y,.

Returning to Figure 3, the AC voltage signal V, is fed into a RT-DQ block 51 which resolves V, into its direct and quadrature components, Vd and Vxq respectively.

Similarly, the AC current signal I, is fed into a RT-DQ block 53 which resolves I, into its direct and quadrature components, xd and xq respectively.

The droop characteristics of a power source are represented in this invention as direct droop factor K and the quadrature droop factor K. The values of Kj and Kq are governed by issues such as the accuracy of measurements, the allowable inbalance of the current sharing between inverters and the required regulation on full load. For two identical inverters K(j and Kq should be the same. Typically K and are equal and may be set to give a full load droop of 4%, although the lower the droop value the better the regulation of the power supply. If an inverter is designed to be added in parallel to existing inverters using this sharing principle, Kj and K1 would have to give the same full load regulation as those existing inverters, to ensure that such power sources share the real power and vars proportionally by setting the inverter output voltages to the correct level depending on the load. If one voltage were to be set too high then this source would supply too much current and the sources will not share the load proportionally.

Hence, when a current I, flows, the current being comprised of a direct component xd and a quadrature component ixq, the target direct and quadrature voltage components are Vdd4QIxd) and (Vqd-KIxq) respectively. For example, for Vdd = 240V, \Tqd = 0, K1 = Kq = K = 1 and xd = ixq =1A, the target values of Vd and Vxq will be 239V and -lv respectively. The errors Vde and Vqe between the measured values Vd and Vqd and their target values respectively are calculated in summation blocks 55, 57 and the values are passed to the DQ to real time converter block, DQ-RT 59, which is shown in Figure 5.

In the DQ-RT block 59 of Figure 5, the values Vde and Vqe are passed to integrators 61 to convert the error voltage signals into a direct target current signal dtgt and a quadrature target current signal qtgt. The integrated values are then multiplied by cos(8 PLL) and sin(8 PLL) respectively, and the resulting alternating signals are then combined to produce an alternating target current signal, xtgt for each phase. Using the example in the previous paragraph, if Vd and Vxq are 240V and OV respectively, the error values Vde and Vqe will be -1V and -lv respectively, which will give non-zero integrator outputs, which are combined to give a non-zero alternating target current signal xtgt. The current loop 11 uses the target current signals atgt, btgt and ctgt to determine the required voltage and phase for each output voltage phase A, B and C, which is converted to PWM duty cycle signals Da, Db and D to produce the desired voltage output waveform on each of the inverter voltage outputs. The feedback loop of the inverter system 1 ensures that values of Vde and Vqe are driven towards zero.

Figure 6 is a circuit diagram of the four leg inverter 63. A gas turbinedrives a shaft on which are wound two sets of three phase windings. Each set of windings produces an isolated 2kHz three phase sine wave voltage supply Vi,, Vq and V1 65, which is rectified by a rectifier 67 to produce a DC supply 69. The DC supply 69 is between 800V and 1200V, and due to the high frequency of the AC voltage source 65, does not require very much filtering, namely a link inductor, LLINK, 71 of 4OpH and a link capacitor, CLINK, 73 of 2400 jtF.

The DC supply 69 feeds two inverters 63, which together form a 500kVA system. Each inverter can operate separately, giving the overall system some built in redundancy.

There is also some latitude for cross checking of measurements giving the ability to detect faults. This approach also forces the two generator sections to share current, provided the inverter halves 63 can be made to share the load current.

The duty cycle drive signals Da, Db, D and D drive the gate of the IGBTs 77 using PWM. A sine wave AC voltage is produced on each of the three legs A, B and C of the inverter 63, according to the desired voltage and phase characteristics. Each voltage is smoothed using the filter formed by an inductor L, and capacitor C,. The four output voltage lines 79 are connected to the regional or local grid network and are fed back and monitored as inputs of the controller 1. If the IGBTs of the neutral leg are switched off, the inverter becomes a three leg inverter. It may be advantageous to modulate the IGBTs 81 on the neutral line, for example by using space vector modulation at three times the transmission frequency to reduce the DC bus voltage 9 by 15%.

Although a specific example of a generator and inverter is given above, the controller of the present invention can be applied to different types of generator and inverter.

Various modifications will be apparent to those in the art and it is desired to include all such modifications as fall within the scope of the accompanying claims.

Claims

CLAIMS: 1. A controller for an AC power source having one or more phase

outputs, the controller comprising: voltage measuring means for measuring the voltage of the or each phase output; current measuring means for measuring the output current of the or each phase output; frequency reference means for generating a phase frequency reference for the or each phase output; processing means for resolving the measured voltage of the or each phase output into direct and quadrature voltage components and for resolving the measured current of the or each phase output into direct and quadrature current components, wherein the direct components of the or each phase output are the components which are in phase with the phase frequency reference of the respective phase output and the quadrature components of the or each phase output are the components which are 90 degrees out of phase with the phase frequency reference of the respective phase output; error calculation means for comparing the direct and quadrature voltage components and the direct and quadrature current components of the or each phase output with target direct and quadrature components respectively to calculate direct and quadrature error values for the or each respective phase output; and control signal means for generating control signals for controlling the or each phase output of the power source, based upon said respective direct and quadrature error values.
2. A controller as claimed in claim 1, wherein the target direct and quadrature components are determined based upon desired droop characteristics of the power source.
3. A controller as claimed in claim 2, wherein the droop characteristics are comprised of a direct droop constant and a quadrature droop constant, wherein: the direct droop constant is the desired change in the direct voltage component from an open circuit direct voltage component, per unit of direct current component; and the quadrature droop constant is the desired change in the quadrature voltage component from an open circuit quadrature voltage component, per unit of quadrature current component.
4. A controller as claimed in any one of the preceding claims, wherein the control signal means further comprises means for summing the direct and quadrature error values for the or each phase output to produce cumulative direct and quadrature error values for the or each phase output.
5. A controller as claimed in claim 4, wherein the control signal means further comprises conversion means for converting the cumulative direct and quadrature error values of the or each phase output into control signals for the or each phase output.
6. A controller as claimed in claim 5, wherein the conversion means comprises: a first and second multiplier; a combiner; and signal generator means for producing a sinusoidal wave of frequency equal to the frequency of the phase frequency reference and a shifted sinusoidal wave which is shifted by 90 degrees relative to the sinusoidal wave, wherein the first multiplier multiplies the sinusoidal wave with the cumulative direct error value and the second multiplier multiplies the shifted sinusoidal wave with the cumulative quadrature error value, and the combiner combines the outputs of the first and second multiplier.
7. A controller as claimed in any one of the preceding claims, wherein the frequency reference means comprises a phase locked loop.
8. A controller as claimed in any one of the preceding claims where there are a plurality of phase outputs, wherein the frequency reference means comprises averaging means for averaging the phase of the voltages of the phase outputs for generating a primary phase frequency reference.
9. A controller as claimed in claim 8, wherein the frequency reference means comprises phase shifting means for generating the phase frequency reference for each phase outputs by applying a respective phase shift to the primary phase frequency reference, the phase shift depending on desired phase spacing between the voltages of the phase outputs.
10. An AC output voltage power source comprising: a DC power source; an inverter; and a controller as claimed in any one of the preceding claims for controlling the inverter.
11. A controller adapted to control a multi-phase AC power source having a plurality of phase outputs, the controller comprising: voltage measuring means for measuring the voltage of each of the phase outputs; current measuring means for measuring the current of each of the phase outputs; frequency reference means for generating a phase frequency reference for each phase outputs; processing means for resolving the measured voltage of each of the phase outputs into direct and quadrature voltage components and for resolving the measured current of each of the phase outputs into direct and quadrature current components, wherein the direct components of each of the phase outputs are the components which are in phase with the phase frequency reference of the respective phase output and the quadrature components of each of the phase outputs are the components which are 90 degrees out of phase with the phase frequency reference of the respective phase output; error calculation means for comparing the direct and quadrature voltage components and the direct and quadrature current components of each of the phase outputs with target direct and quadrature components respectively to calculate direct and quadrature error values for each respective phase outputs; and control signal means for generating control signals for controlling each of the phase outputs of the power source, based upon said respective direct and quadrature error values.
12. A method of controlling an AC power source having one or more phase output, the method comprising: measuring the voltage of the or each phase output; measuring the current of the or each phase output; generating a phase frequency reference for the or each phase output; resolving the measured voltage of the or each phase output into direct and quadrature voltage components and resolving the measured current of the or each phase output into direct and quadrature current components, wherein the direct components of the or each phase output are the components which are in phase with the phase frequency reference of the respective phase output and the quadrature components of the or each phase output are the components which are 90 degrees out of phase with the phase frequency reference of the respective phase output; comparing the direct and quadrature voltage components and the direct and quadrature current components of the or each phase output with target direct and quadrature components respectively and calculating direct and quadrature error values for the or each respective phase output; and generating control signals for controlling the or each phase output of the power source, based upon said respective direct and quadrature error values.
13. A method as claimed in claim 12, wherein the desired droop characteristics of the power source determine the target direct and quadrature components.
14. A method as claimed in claim 13, wherein the droop characteristics are comprised of a direct droop constant and a quadrature droop constant, wherein: the direct droop constant is the desired change in the direct voltage component from an open circuit direct voltage component, per unit of direct current component; and the quadrature droop constant is the desired change in the quadrature voltage component from an open circuit quadrature voltage component, per unit of quadrature current component.
15. A method as claimed in any one of claims 12 to 14, wherein generating control signals further comprises summing the direct and quadrature error values for the or each phase output to produce cumulative direct and quadrature error values for the or each phase output.
16. A method as claimed in claim 15, wherein generating control signals further comprises converting the cumulative direct and quadrature error values of the or each phase output into control signals.
17. A method as claimed in claim 16, wherein converting the cumulative direct and quadrature error values of the or each phase output into realtime control signals comprises: producing a sinusoidal wave of frequency equal to the frequency of the phase frequency reference; producing a shifted sinusoidal wave which is shifted by 90 degrees relative to the sinusoidal wave; multiplying the sinusoidal wave and the shifted sinusoidal wave with the cumulative direct and quadrature error values respectively; and combining the multiplied outputs.
18. A method as claimed in any one of claims 12 to 17 where there are a plurality of phase outputs, wherein generating a phase frequency reference comprises averaging the phase of the voltages of the phase outputs to generate a primary phase frequency reference.
19. A method as claimed in claim 18, wherein generating a phase frequency reference for each phase outputs comprises applying a phase shift to the primary phase frequency reference, the respective phase shift depending on the desired phase spacing between the voltages of the phase outputs.
20. A carrier medium carrying computer readable instructions for controlling a computer to carry out the method of any one of claims 12 to 19.
21. A controller for controlling an AC output power source, substantially as hereinbefore described with reference to Figures 1 to 6 of the accompanying drawings.
22. A method of controlling an AC output power source, substantially as hereinbefore described with reference to Figures 1 to 6 of the accompanying drawings.