METHOD OF CONTROLLING A GRID SIDE CONVERTER OF A WIND TURBINE AND SYSTEM SUITABLE THEREFORE

Abstract

According to an embodiment of the present invention, a method of controlling a grid side converter of a wind turbine is provided, wherein an output of the grid side converter is connected or connectable via a power line to an input of a grid transformer, the method comprising: determining a converter volt-sec occurring at the output of the grid side converter based on a converter voltage occurring at the output of the grid side converter; determining a volt-sec error between the determined converter volt-sec and a converter volt-sec reference, wherein the converter volt-sec reference is determined based on active power reference, reactive power reference, line current and a line voltage occurring at the input of the grid transformer (or at wind turbine terminals); and controlling, based on the determined volt-sec error, the grid side converter such that the volt-sec error is partly or fully compensated. The method is capable of supporting the demanded power quality even during abnormal voltage conditions, and is easy to realize. The method can seamlessly transit from linear to non-linear region of converter operation.

Description

TECHNICAL FIELD

The present invention relates generally to a method of controlling a grid side converter of a wind turbine and a system suitable therefore. Further, the present invention relates to a wind turbine.

BACKGROUND

In order to electrically connect a wind turbine generator to a power grid, several power connection arrangements may be used. A typical power connection arrangement (e.g. for a full scale converter based turbine) comprises a generator side converter, a DC (direct current) link, a grid side converter and a grid transformer which are placed between the wind turbine generator and the power grid in this order. The generator side converter controls the generator power and transfers it to the grid through the DC-link and the grid side converter. The control of the grid side converter should be carried out such that the quality of the electrical power supplied to the grid is kept within predetermined tolerance ranges even during abnormal grid voltage conditions like symmetric/asymmetric voltage dip and swell conditions. In case of abnormal grid voltage conditions, support of reactive power may be needed to fulfil the grid requirements. That is, the power which is supplied to the grid needs to have, besides an active power component, a reactive power component. During normal working conditions, the reactive power component might be required depending on the grid support requirement. The reactive power support is getting increasingly important due to the grid code requirements that are getting stricter along the increase of wind power penetration.

It is therefore an object to provide a method of controlling a wind turbine which is capable of supporting the demanded power quality even during abnormal voltage conditions. Further, this method should be easy to realize.

SUMMARY

According to an embodiment of the present invention, a method of controlling a grid side converter of a wind turbine is provided, wherein an output of the grid side converter is connected or connectable via a power line to an input of a grid transformer, the method comprising: a) determining a converter volt-sec occurring at the output of the grid side converter based on a converter voltage occurring at the output of the grid side converter; b) determining a volt-sec error between the determined/actual converter volt-sec and a converter volt-sec reference, wherein the converter volt-sec reference is determined based on an active power reference, a reactive power reference, a line current and a line voltage occurring at the input of the grid transformer (or at wind turbine terminals); and c) controlling, based on the determined volt-sec error, the grid side converter such that the volt-sec error is (partially or fully) compensated. The term ‘volt-sec’ stands for volt-seconds. It is a vector entity and is equivalent to flux in AC machines. According to an embodiment of the present invention, this process may be supported by volt-sec prediction, as will become apparent later on.

According to an embodiment of the present invention, the converter volt-sec reference is determined based on an active power volt-sec reference, a reactive power volt-sec reference, a line current and a line voltage.

According to an embodiment of the present invention, the active power volt-sec reference is determined based on a difference between an active power reference demanded at the input of the grid transformer (or at wind turbine terminals) and an active power occurring at the input of the grid transformer (or at wind turbine terminals).

According to an embodiment of the present invention, the reactive power volt-sec reference is determined based on a difference between a reactive power reference demanded at the input of the grid transformer (or at wind turbine terminals) and a reactive power occurring at the input of the grid transformer (or at wind turbine terminals).

According to an embodiment of the present invention, the active power volt-sec reference and the reactive power volt-sec reference are respectively frozen if the amplitude of the line current occurring at the input of the grid transformer exceeds a predetermined threshold value.

According to an embodiment of the present invention, the active power and the reactive power are determined based on the line current and the line voltage occurring at the input of the grid transformer.

According to an embodiment of the present invention, determining the converter volt-sec reference comprises generating a first volt-sec reference in dependence on the active power volt-sec reference and the reactive power volt-sec reference.

According to an embodiment of the present invention, the first volt-sec reference is generated as a αβ frame volt-sec vector.

According to an embodiment of the present invention, determining the converter volt-sec reference comprises generating a second volt-sec reference by adding a grid volt-sec to the first volt-sec reference.

According to an embodiment of the present invention, the grid volt-sec is determined based on at least one of the line current and the line voltage occurring at the input of the grid transformer.

According to an embodiment of the present invention, determining the converter volt-sec reference comprises predicting a converter volt-sec which should occur in the next sampling cycle based on the second volt-sec reference, an angular velocity of the line voltage occurring at the input of the grid transformer and a sampling period of the grid side converter, and using the predicted converter volt-sec as converter volt-sec reference.

According to an embodiment of the present invention, controlling of the grid side converter is carried out by supplying, in dependence on the determined volt-sec error, pulse width modulation signals to the grid side converter which adjust the output voltage of the grid side converter accordingly.

According to an embodiment of the present invention, the compensation of the volt-sec error (Δφ_c) is carried out using a PWM switching technique in the linear region of grid side converter operation, and using an over-modulation technique in the non-linear region of the grid side converter operation. Therefore unlike in conventional approaches of current controls, this embodiment naturally extends the grid side converter operation to better utilize the DC-link. Moreover, volt-sec error (Δφ_c) compensation technique naturally fits very well to space vector modulation.

According to an embodiment of the present invention, a controlling system for controlling a grid side converter of a wind turbine is provided, an output of the grid side converter being connected or connectable via a power line to an input of a grid transformer. The controlling system comprises an inner control loop and an outer control loop. The inner control loop comprises: a converter volt-sec determining unit which determines a converter volt-sec occurring at the output of the grid side converter based on a converter voltage occurring at the output of the grid side converter; a volt-sec error determining unit determining a volt-sec error between the determined converter volt-sec and a converter volt-sec reference; and a controlling unit which controls, based on the determined volt-sec error, the grid side converter such that the volt-sec error is compensated, wherein the outer control loop is configured to determine the converter volt-sec reference based on active power demand and reactive power demand (active and reactive power reference), a line voltage and a line current occurring at the input of the grid transformer (or at wind turbine terminals).

According to an embodiment of the present invention, the outer control loop comprises an active power volt-sec reference determining unit and a reactive power volt-sec reference determining unit which determines an active power volt-sec reference and a reactive power volt-sec reference based on (1) an active power reference and active power occurring at the input of the grid transformer (or at wind turbine terminals) (2) a reactive power reference and reactive power occurring at the input of the grid transformer (or at wind turbine terminals).

According to an embodiment of the present invention, the active power volt-sec reference determining unit determines the active power volt-sec reference based on a difference between an active power reference and an active power occurring at the input of the grid transformer (or at wind turbine terminals).

According to an embodiment of the present invention, the reactive power volt-sec reference determining unit determines the reactive power volt-sec reference based on a difference between a reactive power reference and a reactive power occurring at the input of the grid transformer (or at wind turbine terminals).

According to an embodiment of the present invention, the outer control loop comprises a current limiting unit which causes the active power volt-sec reference and the reactive power volt-sec reference to be respectively frozen if the amplitude of the line current occurring at the input of the grid transformer exceeds a predetermined threshold value.

According to an embodiment of the present invention, the outer control loop comprises an active power determining unit and a reactive power determining unit which determine the active power and reactive power based on the line current and the line voltage occurring at the input of the grid transformer, respectively.

According to an embodiment of the present invention, the outer control loop comprises a αβ frame volt-sec vector generating unit which generates a first volt-sec reference as a αβ frame volt-sec vector in dependence on the active power volt-sec reference, the reactive power volt-sec reference and an angle of the line voltage.

According to an embodiment of the present invention, the outer control loop comprises a second volt-sec reference generating unit which generates a second volt-sec reference by adding a grid volt-sec to the first volt-sec reference.

According to an embodiment of the present invention, the outer control loop comprises a grid volt-sec determining unit which determines the grid volt-sec based on at least one of the line current and the line voltage occurring at the input of the grid transformer.

According to an embodiment of the present invention, the outer control loop comprises a converter volt-sec reference predicting unit which predicts a converter volt-sec for the next sampling cycle based on the second volt-sec reference, an angular velocity of the line voltage occurring at the input of the grid transformer and a sampling period of the grid side converter, and outputs the predicted converter volt-sec as converter volt-sec reference.

According to an embodiment of the present invention, the inner control loop comprises a PWM unit which controls, in dependence on the determined volt-sec error, the output voltage of the grid side converter by supplying pulse width modulation signals to the grid side converter.

According to an embodiment of the present invention, a wind turbine is provided comprising a power generator (AC machine) and a grid side converter, wherein the power generator is connected to the grid side converter via a generator side converter and a DC-link connecting the generator side converter with the grid side converter, and wherein the wind turbine comprises a controlling system for controlling the grid side converter of the wind turbine.

According to an embodiment of the present invention, in order to compensate the volt-sec error (i.e. to compensate the volt-sec error between the determined converter volt-sec and a converter volt-sec reference, also referred to as “volt-sec (volt-seconds) compensation”), any PWM technique can be used. Among various PWM techniques, space vector modulation (SVM) is known for its various advantages. The volt-sec compensation approach naturally fits with the SVM technique.

According to an embodiment of the present invention, the outer control loop may be operated at low bandwidth, whereas the inner control loop may be operated at high bandwidth which helps to obtain good dynamic response. The inner control loop may simply comprise or consist of a volt-sec compensation block which can be executed very fast.

According to an embodiment of the present invention, almost all currents, voltages and volt-secs which are used by the embodiments of the present invention are respectively vectors.

According to an embodiment of the present invention, a PLL (phase locked loop) is used with positive and negative sequence separator. It is also referred to as PLL with positive and negative sequence calculator (PNSC).

According to an embodiment of the present invention, an active power p and reactive power q are calculated using the expressions p=1.5(v_αi_α+v_βi_β) and q=1.5(v_αi_β−v_βi_α) where [v_α,v_β] are actual or positive sequence grid voltages and [i_α, i_β] are actual or positive sequence line currents. Here, α and β indicate the α and β components of the current and voltage signals in the stationary αβ frame, respectively.

In the scope of the embodiments of the present invention, the term “volt-sec” stands for “volt-seconds”. “Volt-sec” means a vector entity, the unit of which being volt*second. In AC power generators, the term “volt-sec” is equivalent to the term “flux”.

According to an embodiment of the present invention, depending on regulation and system requirements, controlling the active power p to the active power reference p* is given priority over controlling the reactive power q to the reactive power reference q*, or vice versa. That is, according to an embodiment of the present invention, depending on the situations, the active power reference p* and the reactive power reference q* may take priority one over the other. For example, when grid support is required, q* may take priority over p*, and in conditions where active power control is required to reduce mechanical stress on the turbine drivetrain/structure, p* may take priority over q*. Of course, it is also possible to give the same priority to both controlling the active power and the reactive power. Many other situations are possible.

In the following, further aspects of embodiments of the present invention will be described.

As has become apparent in the preceding description, embodiments of the present invention provide a grid side converter control philosophy to directly control the real and reactive power transacted with the grid. Due to the simplicity of the proposed close loop control, it can be easily implemented on two-level converters as well as on multi-level converters. The proposed close loop control is capable of riding through grid symmetric/asymmetric voltage dip and swell conditions. This ensures a better connectivity of the turbine with the grid under fault conditions. Further, the approach is computationally simple to realize. With respect to conventional controls, the proposed control requires lesser controllers.

According to an embodiment of the present invention, reactive power needed during a low voltage grid fault ride through event is generated by setting a reactive power reference to a desired value and by limiting an active power reference based on grid side converter current limits. In the low voltage condition, a chopper resistor located in a DC-link may also be used such that active power coming from the generator can be dissipated in the chopper resistor. The same can be applied to high voltage condition.

According to an embodiment of the present invention, a PLL algorithm may be used to estimate the grid angle (γ) and the angular speed (ω) of the grid voltage. The PLL algorithm may be implemented using synchronous rotating frame (SRF) technique and may use moving average (MA) technique to eliminate negative sequence and harmonics components from the grid voltage. Thus, the grid angle (γ) and the angular speed (ω) may be calculated based on the positive sequence voltage. However, the grid angle (γ) and the angular speed (ω) may also be calculated based on the negative sequence voltage. There are many other methods of estimating the grid angle (γ) and the angular speed (ω) of the grid voltage which may be used.

According to embodiments of the present invention, the grid side converter controlling method is able to easily handle symmetric/asymmetric grid voltage swell and dip conditions. Further, the hardware/software structure for realizing the grid side converter controlling method is the same for balanced/unbalanced grid voltages.

According to an embodiment of the present invention, SVM technique may be used to control the grid side converter which can be seamlessly integrated into techniques such as bus-clamping, three phase symmetry and half-wave symmetry PWM. These techniques can help to significantly improve converter performance.

According to an embodiment of the present invention, the switching frequency of the grid side converter is kept constant during operation.

According to an embodiment of the present invention, the proposed algorithm can be easily applied to multilevel inverters as grid side converters. With the use of multilevel inverters, the line filter size and weight may be reduced and cost saving can be achieved.

The following advantages may be achieved by using embodiments of the present invention:

a) A tight control of power and voltage can be achieved due to the use of volt-second based approach.b) The proposed control method performs well in LVRT (low voltage ride through) and HVRT (high voltage ride through) in both symmetric and asymmetric fault conditions. This is one of the main advantages, compared to conventional DPC (direct power control) approaches. Further, the hardware/software structure for realizing the proposed grid side converter control method is the same for balanced/unbalanced grid conditions.c) A good dynamic response of the grid side control (fast controlling speed) can be achieved due to an inherently fast inner loop.d) The proposed control method can be easily operated up to six-step operation which can help to achieve faster dynamic response due to better utilization of the DC-link. This is due to the fact that in the normal modulation range up to 90.7% of the installed DC-link capacity can be utilized. However, through operation in overmodulation region till six-step operation 100% of the installed DC-link capacity can be used. The volt-sec error is significantly large in a dynamic grid condition. Thus, a fast dynamic performance of the grid side controller is needed. This fast dynamic performance can be achieved in view of the simplicity of the proposed control method. The proposed algorithm ensures that the volt-sec balance is maintained even in a dynamic condition e.g. grid faults. As far as details of overmodulation and six-step techniques which may be employed by embodiments of the present invention are concerned, the following references should be cited: (1) N. Mohan, T. M. Undeland, and W. P. Robbins, Power electronics: converters, applications, and design. Hoboken, N.J.: John Wiley and Sons, third ed., 2003. (2) J. Holtz, “Pulsewidth modulation-a survey,” Industrial Electronics, IEEE Transactions on, vol. 39, pp. 410-420, October 1992. (3) J. Holtz, W. Lotzkat, and A. M. Khambadkone, “On continuous control of pwm inverters in overmodulation range including six-step,” IEEE Transaction on Power Electronics, vol. 8, pp. 546-553, 1993.e) Due to the use of volt-second compensation, the proposed algorithm is inherently suitable to space vector modulation (SVM) which can be easily applied to two-level or multilevel based grid side converters.f) The switching frequency of the grid side converter can be kept constant, in contrast to some DPC based control algorithms where the switching frequency is not constant. Constant switching frequency leads to simplicity of a grid filter design. This also helps to limit and reduce the switching loss occurring in power converters.g) The proposed approach is computationally simple.h) The proposed concept can easily be used together with synchronous SVM technique which helps to eliminate the subharmonics and interharmonics included within the voltage signal output by the grid side converter. These harmonics are difficult to be filtered by a grid side filter and may create a number of problems for grid control and grid connected applications.i) Since the proposed concept uses volt-sec as basis, it is inherently suitable for the SVM. The implementation of bus-clamping is very easy for SVM. With bus-clamped SVM, the proposed concept can achieve 33% switching loss reduction. In bus clamping one of the phase legs is not switched at all for a given switching period while still forming the given reference vector with discrete switching states. As far as details of SVM based bus clamping techniques which may be employed by embodiments of the present invention are concerned, the following references should be cited: (1) A. Beig, G. Narayanan, and V. Ranganathan, “Space vector based synchronized pwm algorithm for three level voltage source inverters: principles and application to v/f drives,” IECON 02 [Industrial Electronics Society, IEEE 2002 28th Annual Conference of the], vol. 2, pp. 1249-1254, November 2002. (2) G. Narayanan and V. T. Ranganathan, “Synchronised bus-clamping pwm strategies based on space vector approach for modulation up to six-step mode,” Power Electronic Drives and Energy Systems for Industrial Growth, 1998. Proceedings. 1998 International Conference on, vol. 2, pp. 996-1001, December 1998. (3) G. Narayanan and V. T. Ranganathan, “Synchronised pwm strategies based on space vector approach. i. principles of waveform generation,” Electric Power Applications, IEE Proceedings, vol. 146, pp. 267-275, May 1999. (4) G. Narayanan and V. T. Ranganathan, “Two novel synchronized busclamping pwm strategies based on space vector approach for high power drives,” Power Electronics, IEEE Transactions on, vol. 17, pp. 84-93, January 2002.j) Due to absence of cross coupling term, the proposed controller is more robust especially considering grid choke saturation and frequency variations.

According to an embodiment of the present invention, the compensation of the volt-sec error (volt-sec error Δψ) is understood as follows: According the faraday's law, the voltage vector equals the change of the volt-sec vector over time, i.e. v=dψ/dt. It can be equivalently written as Δψ=vΔt, i.e. the volt-sec error Δψcan be compensated by holding the voltage vector v for the small time period Δt. This explains the principle of volt-sec error based implementation of SVM.

BRIEF DESCRIPTION OF THE DRAWINGS

In the drawings, like reference characters generally refer to the same parts throughout the different views. The drawings are not necessarily to scale, emphasis instead generally being placed upon illustrating the principles of the invention. In the following description, various embodiments of the invention are described with reference to the following drawings, in which:

FIG. 1 illustrates a schematic drawing of a common setup of a conventional wind turbine.

FIG. 2 shows a schematic drawing of an arrangement comprising a grid side converter, a grid transformer, and a power line connecting the grid side converter to the grid transformer as well as corresponding voltages and currents occurring in this arrangement.

FIG. 3 shows a schematic drawing of a grid converter control system according to an embodiment of the present invention.

FIGS. 4A and 4B show schematic drawings of examples of grid side converters which may be used within a control system according to an embodiment of the present invention.

FIG. 5 shows the relationship between voltages, currents and volt-secs occurring in the control system shown in FIG. 6.

FIG. 6 shows a schematic drawing of an example of a control system according to an embodiment of the present invention.

FIG. 7 shows a flowchart of a method of controlling a grid side converter of a wind turbine according to an embodiment of the present invention.

FIG. 8 shows a schematic drawing of a PLL algorithm usable within the control system shown in FIG. 6 according to an embodiment of the present invention.

FIG. 9 shows a schematic drawing of an error detection function usable within the PLL algorithm shown in FIG. 8 according to an embodiment of the present invention.

DETAILED DESCRIPTION

FIG. 1 illustrates a common setup of a conventional wind turbine 100. The wind turbine 100 is mounted on a base 102. The wind turbine 100 includes a tower 104 having a number of tower sections, such as tower rings. A wind turbine nacelle 106 is placed on top of the tower 104. The wind turbine rotor includes a hub 108 and at least one rotor blade 110, e.g. three rotor blades 110. The rotor blades 110 are connected to the hub 108 which in turn is connected to the nacelle 106 through a low speed shaft which extends out of the front of the nacelle 106.

FIG. 2 shows an arrangement 200 comprising a grid side converter 202, a low voltage side 204 of a grid transformer (not shown), and a power line 206 connecting the grid side converter 202 to the grid 204 as well as corresponding voltages and currents occurring in this arrangement. In FIG. 2, u_g denotes the voltage occurring at the input or low voltage side 204 of the grid transformer (also referred to as “grid voltage”), u_e denotes the voltage occurring at the output of the grid side converter 202 (also referred to as “converter voltage”), and i_g denotes the line current flowing through the power line 206. x_s denotes an reactance of the power line 206, and u_x denotes the voltage drop over the reactance x_s. The major component of x_s is the grid choke reactance.

In normal condition, it can be assumed that the grid voltage u_g is constant. In general, for any grid condition u_g can be considered as the grid reference voltage for the grid converter control. The voltage u_x changes if the line current i_g changes. The line current i_g itself changes with respect to the demanded power which has to be supplied by the grid side converter 202 to the low voltage side 204 of the grid transformer. Thus, the voltage u_e has to be adapted due to a change of the demanded power, i_g and therefore u_x changes.

In order to adapt the voltage u_e, the grid side converter 202 has to be controlled accordingly. In the following, methods/systems for controlling the grid side converter 202 are described which are capable of supporting the demanded power quality even during abnormal voltage conditions, and which are easy to realize.

FIG. 7 shows an embodiment of a method 700 of controlling a grid side converter of a wind turbine. An output of the grid side converter is connected or connectable via a power line to an input of a grid transformer. At 702, a converter volt-sec occurring at the output of the grid side converter is determined based on a converter voltage occurring at the output of the grid side converter. At 704, a volt-sec error (i.e. a difference between the converter volt-sec and a converter volt-sec reference) is determined, wherein the converter volt-sec reference is determined based on an active power reference, a reactive power reference, a line current and a line voltage occurring at the input of the grid transformer (or at wind turbine terminals). At 706, based on the determined volt-sec error, the grid side converter is controlled such that the volt-sec error is compensated.

FIG. 3 shows a controlling system 300 according to an embodiment of the present invention which is an example of realizing the method as indicated in FIG. 7. The controlling system 300 controls a grid side converter 302 of a wind turbine. An output 306 of the grid side converter 302 is connected or is connectable via a power line 304 to an input 308 of a grid transformer 310. The controlling system 300 comprises an inner control loop arrangement 312 and an outer control loop arrangement 314. The inner control loop arrangement 312 comprises a converter volt-sec determining unit 316 which determines a converter volt-sec φ_c occurring at the output 306 of the grid side converter 302 based on a converter voltage u_c occurring at the output 306 of the grid side converter 302. The inner control loop 312 further comprises a volt-sec error determining unit 318 determining a volt-sec error Δφ_c between the determined converter volt-sec φ_c and a converter volt-sec reference φ_pc*, and a controlling unit 320 which controls, based on the determined volt-sec error Δφ_c, the grid side converter 302 such that the volt-sec error Δφ_c is compensated. The outer control loop arrangement 314 comprises a unit 322 for determining the converter volt-sec reference φ_pc* based on an active power reference p*, a reactive power reference q*, a line current i_g and a line voltage u_g occurring at the input 308 of the grid transformer 310.

The unit 322 for determining the converter volt-sec reference φ_pc* may comprise several subunits, as will become apparent below in conjunction with FIG. 6 where a more detailed example of a realization of the embodiment as shown in FIG. 3 will be explained. It should also be noted that the outer control loop arrangement 314 may comprise not only one, but several feedback lines, as will become apparent below in conjunction with FIG. 6.

As can be derived from FIG. 3, the controlling system 300 is a volt-sec based control system. That is, the controlling system 300 controls the power output by the grid side converter 302 not based on direct voltage control or direct current control, but on volt-sec control. This means that not the voltage or current output by the grid side converter 302 is controlled, but the volt-sec (also referred to as “volt seconds” (“volt-sec”)) output by the grid side converter 302 is controlled.

FIG. 4A shows an example 400 of a grid side converter 302 which may be employed in embodiments of the present invention. In this embodiment, the grid side converter 400 is a three-level converter which has three legs i.e. 402, 404, and 406 (since it is assumed that a three phase power generating system is used). The present invention is not restricted to the grid side converter 400; instead converters having two-levels or more than three-levels may be used, depending on the power generating system used. For example, FIG. 4B shows a two-level converter.

FIG. 6 shows a controlling system 600 for controlling a grid side converter 602 of a wind turbine 100. An output 606 of the grid side converter 602 is connected or connectable via a power line 604 (which includes a grid choke 634) to an input 608 of a grid transformer 640. The grid side converter 602 is connected via a DC-link (not shown) to an AC power generator (not shown) through a generator side converter. A DC link voltage 636 is measured across the DC-link connected to the grid side converter 602. An output 638 of the grid transformer 640 (which may be an output of the wind turbine nacelle) is connected to a power grid (not shown).

The controlling system 600 comprises an inner control loop arrangement 610 and an outer control loop arrangement 612. The inner control loop arrangement 610 comprises: a converter volt-sec determining unit 614 which determines a converter volt-sec φ_e occurring at the output 606 of the grid side converter 602 based on a converter voltage u_c occurring at the output 606 of the grid side converter 602. The converter volt-sec φ_c may for example be determined using the following formula:

φ_c=∫u_cdt

That is, the converter volt-sec φ_c may be obtained by integrating the measured converter voltage u_c occurring at the output 606 of the grid side converter 602 over the time. The converter volt-sec φ_c may also be estimated through converter duty ratios and the DC-link voltage. The converter volt-sec φ_c may also be calculated with different approaches.

The inner control loop arrangement 610 further comprises a volt-sec error determining unit 616 determining a volt-sec error Δφ_c (i.e. a difference between the converter volt-sec φ_c and a converter volt-sec reference φ_pc* (φ_pc* is the target volt-sec to be output by the grid side converter 602)). The volt-sec difference Δφ_c is also referred to as “volt-sec error”. Part of the inner control loop arrangement 610 is also a controlling unit 618 which controls, based on the determined volt-sec error Δφ_c, the grid side converter 602 such that the volt-sec error Δφ_c is compensated. The controlling unit 618 is essentially a PWM unit which determines the duty ratios for the power switches of the grid side converter 602.

The outer control loop arrangement 612 comprises several units for determining the converter volt-sec reference φ_pc* based on active power reference p*, reactive power reference q*, line current i_g and a line voltage u_g occurring at the input 608 of the grid transformer 640. These units will be described in the following.

The outer control loop arrangement 612 comprises an active power volt-sec reference determining unit 620 and a reactive power volt-sec reference determining unit 622 which determine an active power volt-sec reference φ_xq*(projection of the volt-sec φ_x* on the q-axis) and a reactive power volt-sec reference φ_xd* (projection of the volt-sec φ_x* on the d-axis) based on an active power reference p* and a reactive power reference q* demanded at the input 608 of the grid transformer 640 (or at wind turbine terminals), respectively.

The active power volt-sec reference determining unit 620 determines the active power volt-sec reference φ_xq* based on a difference between an active power reference p* and an active power p occurring at the input 608 of the grid transformer 640 (or at wind turbine terminals). The difference between an active power reference p* and an active power p is calculated by an active power difference calculating unit 642. The active power reference p* is generated by a DC link voltage controller unit 646 and supplied to the active power difference calculating unit 642. The DC link voltage controller unit 646 is itself driven by a difference between a DC link voltage reference V_dc* and the DC link voltage V_dc determined by a DC voltage difference calculating unit 648. The DC-link controller can be implemented in many different ways but the output of the controller is p*. The active power reference p* can be reduced under certain conditions e.g. grid support. This can be done through a modification block (not shown).

The reactive power volt-sec reference determining unit 622 determines reactive power volt-sec reference φ_xd* based on a difference between a reactive power reference q* and a reactive power q occurring at the input 608 of the grid transformer 640 (or at wind turbine terminals). The difference between a reactive power reference q* and a reactive power q is calculated by a reactive power difference calculating unit 644. The reactive power reference q* is mainly determined based on grid support requirements.

The outer control loop arrangement 612 comprises a current limiting unit 624 which causes the active power volt-sec reference φ_xq* and the reactive power volt-sec reference φ_xd* to be respectively frozen if the amplitude of the line current i_g occurring at the input 608 of the grid transformer 640 exceeds a predetermined threshold value.

The outer control loop arrangement 612 further comprises a unit 626 comprising an active power determining subunit and a reactive power determining subunit which determine the active power p and reactive power q based on the line current i_g and the line voltage u_g occurring at the input 608 of the grid transformer 640, respectively. The unit 626 further comprises an amplitude determining subunit which determines the amplitude of the line current i_g which is supplied to the current limiting unit 624.

The amplitude of the line current i_g may for example be calculated based on the following equation:

|i_g|√{square root over (i_gα²+i_gβ²)},

wherein i_gα is the α-component of the line current, and i_gβ is the β-component of the line current. However, the line current i_g may also be calculated in a different manner.

The active power p may for example be calculated based on the following equation:

p=1.5*(v_gα⁺i_gα⁺+v_gβ⁺i_gβ⁺),

wherein v_gα⁺ is the positive sequence α-component of the line voltage u_g, v_gβ⁺ is the positive sequence β-component of the line voltage u_g, i_gα⁺ is the positive sequence α-component of the line current i_g, and v_gβ⁺ is the positive sequence β-component of the line current i_g. In general ‘+’ indicates the positive sequence component of the respective currents/voltages. The active power p may also be calculated in a different manner.

The reactive power q may for example be calculated based on the following equation:

q=1.5*(v_gα⁺i_gβ⁺−v_gβ⁺i_gα⁺),

wherein v_gα⁺ is the positive sequence α-component of the line voltage u_g, v_gβ⁺ is the positive sequence β-component of the line voltage u_g, i_gα⁺ is the positive sequence α-component of the line current i_g, and v_gβ⁺ is the positive sequence β-component of the line current i_g. The reactive power q may also be calculated in a different manner.

In the above equations, it is shown above that only positive sequence voltages and currents are used to calculate active and reactive power. However, the embodiments of the present invention are not restricted thereto: it is also possible to use the negative sequence voltage and current to calculate the active and reactive power. Hence, either positive sequence or negative sequence or the combination can be used to calculate the active and reactive power.

The outer control loop arrangement 612 further comprises an αβ frame volt-sec vector generating unit 628 which generates a first volt-sec reference φ_x* as an αβ frame volt-sec vector in dependence on the active power volt-sec reference φ_xq* and the reactive power volt-sec reference φ_xd*.

The outer control loop arrangement 612 comprises a second volt-sec reference generating unit 630 which generates a second volt-sec reference φ_c* by adding a grid volt-sec φ_g to the first volt-sec reference φ_x*. The unit 626 further comprises a grid volt-sec determining subunit which determines the grid volt-sec φ_g based on at least one of the line current i_g and the line voltage u_g occurring at the input 608 of the grid transformer 640.

The grid volt-sec is for example determined based on the following equation:

φ_gβ∫u_gdt

That is, the grid volt-sec φ_g may be obtained by integrating the measured converter voltage u_g occurring at the input 608 of the grid transformer 640 over the time. However, the grid volt-sec φ_g may also be calculated in a different manner.

The outer control loop arrangement 612 comprises a converter volt-sec reference predicting unit 632 which predicts a next sampling cycle converter volt-sec φ_pc* based on the second volt-sec reference φ_c* and an angle difference Δθ_s. Δθ_s is calculated based on an angular velocity ω_g of the line voltage u_g occurring at the input 608 of the grid transformer 640 according to the equation Δθ_s=ω_g*T_s, wherein T_s is the sampling period/cycle of the grid side converter 602. The volt-sec reference predicting unit 632 outputs the predicted converter volt-sec φ_pc* as converter volt-sec reference. The angular velocity ω_g of the line voltage u_g is determined by a PLL (Phase Locked Loop) unit 650.

The PLL unit 650 also determines, based on the line voltage u_g occurring at the input 608 of the grid transformer 640, a grid angle γ_g⁺ which is the angle of the positive sequence voltage. The grid angle γ_g⁺ is used to transform the volt-sec reference from the dq frame representation (the active power volt-sec reference φ_xq* and the reactive power volt-sec reference φ_xd*) into the αβ frame representation. If the power control with negative sequence is desired as well, a grid angle γ_g⁻ which is the angle of the negative sequence voltage can also be easily calculated and used to transform the negative sequence volt-sec reference from the dq frame representation (the respective negative sequence active power volt-sec reference φ_xq* and the reactive power volt-sec reference φ_xd*) into the αβ frame representation.

Use of the volt-sec reference predicting unit 632 leads to more accurate grid side converter response especially when the switching frequency is low; it also helps to obtain better dynamic response. However, it should be noted that the volt-sec reference predicting unit 632 is not absolutely necessary but its presence improve the performance.

In the following description, making reference to FIGS. 8 and 9, a possible embodiment of the PLL unit 650 will be described. It is to be understood that also other types of PLL (PNSC) algorithms may be used.

The PLL unit 650 receives three phase voltages U_a, U_b, U_c extracted from the line voltage u_g occurring at the input 608 of the grid transformer 640. The three phase voltages U_a, U_b, U_c are transformed into the αβ frame in order to obtain an α-axis component U_α of the line voltage u_g, and a β-axis component U_β of the line voltage u_g using an abc/αβ transforming unit 800.

A first αβ/qd transforming unit 802 transforms the line voltage components U_α and U_β into the dq frame using the voltage vector phase angle γ_g⁺, e.g. in accordance with

$\left[\begin{array}{c}\mathrm{ULQ}\\ \mathrm{ULD}\end{array}\right]=\left[\begin{array}{cc}\cos {\gamma }_g^{+}& \sin {\gamma }_g^{+}\\ \sin {\gamma }_g^{+}& -\cos {\gamma }_g^{+}\end{array}\right]·\left[\begin{array}{c}U\alpha \\ U\beta \end{array}\right]$

wherein ULD is the d-axis component of the positive sequence line voltage in the d/q frame, and ULQ is the q-axis component of the positive sequence line voltage in the d/q frame.

Further, a second αβ/qd transforming block 804 transforms the line voltage components U_α and U_β into the dq frame using the negative voltage vector phase angle −γ_g⁺, e.g. in accordance with

$\left[\begin{array}{c}\mathrm{ULQ\_NEG}\\ \mathrm{ULD\_NEG}\end{array}\right]=\left[\begin{array}{cc}\cos \left(-{\gamma }_g^{+}\right)& \sin \left(-{\gamma }_g^{+}\right)\\ \sin \left(-{\gamma }_g^{+}\right)& -\cos \left(-{\gamma }_g^{+}\right)\end{array}\right]·\left[\begin{array}{c}U\alpha \\ U\beta \end{array}\right].$

where ULD_NEG is the d-axis component of the negative sequence line voltage in the d/q frame, and ULQ_NEG is the q-axis component of the negative sequence line voltage in the d/q frame.

The line voltage components ULD, ULQ are fed into a first half cycle moving average function unit 806, and the line voltage components ULD_NEG, ULQ_NEG are fed into a second half cycle moving average function unit 808. The half cycle moving average function units 806, 808 respectively process the input signals ULD, ULQ/ULD_NEG, ULQ_NEG according to the following equation

$y=f\left(x\right)=\frac{1}{T}{\int }_{t-T}^tx\left(t\right)*t,$

wherein y is the moving average of the respective input signal, x(t) is the input signal, T is the time of a sampling cycle, and t is the integration parameter (time). The half cycle moving average function unit 806 serves for cancelling negative sequence effects and harmonics effects.

In order to implement the half cycle moving average function units 806, 808, a buffer may respectively be used which holds data of a half cycle, wherein the output of the half cycle moving average function units 806, 808 is the average of the half cycle data stored within the buffer. The signal ULD_MA output by the half cycle moving average function unit 806 represents the d-axis component of the positive sequence line voltage of a half cycle moving average. The signal ULQ_MA output by the half cycle moving average function unit 806 represents the q-axis component of the positive sequence line voltage of a half cycle moving average. The signal ULD_MA_NEG output by the half cycle moving average function unit 808 represents the d-axis component of the negative sequence line voltage of a half cycle moving average. The signal ULQ_MA_NEG output by the half cycle moving average function unit 808 represents the q-axis component of the negative sequence line voltage of a half cycle moving average.

A normalization block 810 normalizes the signals ULD_MA and ULQ_MA. in accordance to

ULD _— NOM=ULD _— MA/√{square root over ((ULD_—MA²+ULQ_—MA²))},

ULQ _— NOM=ULQ _— MA/√{square root over ((ULD_—MA²+ULQ_—MA²))}

in order to generate the normalized output signals ULD_NOM and ULQ_NOM. The normalization process may be used to eliminate the influence of the voltage amplitude on the phase lock loop control. Thus, the PLL unit 650 has the same performance for different line voltages u_g on different platforms.

The normalized output signals ULD_NOM and ULQ_NOM are fed into a mapping unit 812 which maps these signals to a function value of a function F which may be seen as function of the phase error Δθ between the actual voltage phase angle (i.e. the phase angle of the current line voltage as represented by ULD_NOM and ULQ_NOM) and the estimated voltage vector phase angle (i.e. the fed back estimate for the line voltage phase angle γ_g⁺) according to

F(Δθ)=tan(Δθ),−Δθ1<Δθ<Δθ1

F(Δθ)=tan(Δθ1)=ω_Lim,Δθ1<Δθ<π

F(Δθ)=tan(−Δθ1)=−ω_Lim,−π<Δθ<−Δθ1

where Δθ1 is a pre-defined threshold value. The phase error Δθ is for example determined according to tan(Δθ)=ULD_NOM/ULQ_NOM.

The function F is illustrated in FIG. 9. FIG. 9 shows a graph indicating the output of a function F(d,q) versus the phase error according to one embodiment. A first axis 902 (x-axis) of the graph 900 corresponds to the possible values of the phase error Δθ, and a second axis 904 (y-axis) of the graph 900 corresponds to the function values of F. It can be seen from the graph 900 that the output of this mapping unit 812 is following a tangent function when the angle difference is less than Δθ1; and when the angle difference is larger than Δθ1, the output is clamped at the max value of ω_Lim. This mechanism will adjust the tuning velocity depending on the angle difference, and when the angle difference is large, it will yield a faster tuning speed.

The output of the mapping unit 812 is compared to a reference value ULD_REF (d-axis reference value; it is normally zero if the voltage sensor delay is not considered) by a comparing unit 816, and a corresponding comparison result is fed into a PI controller unit 814.

An adding unit 818 adds to the output of the PI controller unit 814 the nominal voltage frequency (50/60 Hz) in order to obtain the estimated angular speed ω_g of the voltage vector in the dq frame. The PI controller unit 814 changes (by acceleration or deceleration) the estimated angular speed ω_g such that, in dependence on the phase error Δθ, the estimated phase angle γ_g⁺ (in the dq frame) meets up with the voltage vector angle.

An integrating unit 820 generates an estimated phase angle γ_g⁺ of the voltage vector by integration of the angular speed ω_g in the dq frame.

In the above description of the PLL unit 650, the positive sequence voltage has been used in order to determine the estimated phase angle γ_g⁺ and the angular speed ω_g. That is, the signals ULD_NOM and ULQ_NOM have been used to determine γ_g⁺ and ω_g. However, it should be mentioned that, in order to determine γ_g⁺ and ω_g, also the signals ULD_MA_NEG and ULQ_MA_NEG could be used. In other words, the estimated phase angle γ_g⁺ and the angular speed ω_g can also be determined using negative sequence voltages.

In other words, according to an embodiment of the present invention, the working principle of the embodiment of FIG. 6 may be summarized as follows: FIG. 6 shows an implementation of a grid converter direct power control scheme using a volt-sec compensation based method. A DC-link voltage controller 646 generates an active power reference p*. The active power reference p* may be modified during grid/system faults. A reactive power reference q* is set according to a reactive power demand. An active power feedback control is used to generate a q-axis volt-sec reference φ_xq*, whereas a reactive power feedback control is used to generate a d-axis volt-sec reference φ_xd*. The vector sum of φ_xq* and φ_xd* and subsequent dq→αβ transformation provides a total volt-sec φ_x* associated with a grid choke 634. A line current limiting is realized by freezing reactive and active power controller outputs when a measured line current i_g hits a maximum current. A simple PLL 650 is used to estimate grid angle, angular speed of the grid voltage and the grid voltage. The PLL 650 may be implemented in synchronous rotating frame (SRF) and may use moving average (MA) to eliminate negative sequence and harmonics components from the grid voltage u_g. Grid voltage angular frequency omega ω_g and grid voltage angle gamma γ_g⁺ are calculated based on a positive sequence of voltages. However, use of negative sequence voltage is also possible. A reference volt-sec vector φ_c* for the converter is the sum of grid volt-sec vector φ_g and volt-sec vector φ_x* associated with the grid choke. Subsequently volt-sec prediction is used to implement a deadbeat controller. The converter volt-sec error vector Δφ_c is utilized to generate a converter voltage reference vector. Using Δφ_e, the final PWM signals are produced for the grid side converter 602.

While embodiments of the invention have been particularly shown and described with reference to specific embodiments, it should be understood by those skilled in the art that various changes in form and detail may be made therein without departing from the spirit and scope of the invention as defined by the appended claims. The scope of the invention is thus indicated by the appended claims and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced.

Claims

1. A method of controlling a grid side converter of a wind turbine, wherein an output of the grid side converter is connected or connectable via a power line to an input of a grid transformer, comprising: computing a converter volt-sec occurring at the output of the grid side converter based on a converter voltage occurring at the output of the grid side converter;determining a volt-sec error between the determined converter volt-sec and a converter volt-sec reference, wherein the converter volt-sec reference is determined based on an active power reference, a reactive power reference, a line current and a line voltage occurring at the input of the grid transformer; andcontrolling, based on the determined volt-sec error, the grid side converter such that the volt-sec error is partially or fully compensated.

2. The method according to claim 1, wherein the converter volt-sec reference is determined based on an active power volt-sec reference and a reactive power volt-sec reference.

3. The method according to claim 2, wherein the active power volt-sec reference is determined based on a difference between an active power reference demanded at the input of the grid transformer and an active power occurring at the input of the grid transformer.

4. The method according to claim 2, wherein the reactive power volt-sec reference is determined based on a difference between a reactive power reference demanded at the input of the grid transformer and a reactive power occurring at the input of the grid transformer.

5. The method according to claim 2, wherein the active power volt-sec reference and the reactive power volt-sec reference are respectively frozen if the amplitude of the line current occurring at the input of the grid transformer exceeds a predetermined threshold value.

6. The method according to claim 3, wherein the active power and the reactive power are determined based on the line current and the line voltage occurring at the input of the grid transformer.

7. The method according to claim 2, wherein determining the converter volt-sec reference comprises generating a first volt-sec reference in a dq frame in dependence on the active power volt-sec reference and the reactive power volt-sec reference.

8. The method according to claim 7, wherein the first volt-sec reference is generated as an αβ frame volt-sec vector.

9. The method according to claim 7, wherein determining the converter volt-sec reference comprises generating a second volt-sec reference by adding a grid volt-sec to the first volt-sec reference.

10. The method according to claim 9, wherein the grid volt-sec is determined based on at least one of the line current and the line voltage occurring at the input of the grid transformer.

11. The method according to claim 9, wherein determining the converter volt-sec reference comprises predicting a next sampling cycle converter volt-sec based on the second volt-sec reference, an angular velocity of the line voltage occurring at the input of the grid transformer and a sampling period of the grid side converter, and using the predicted converter volt-sec as converter volt-sec reference.

12. The method according to claim 1, wherein the controlling of the grid side converter is carried out by supplying, in dependence on the determined volt-sec error, pulse width modulation signals to the grid side converter which adjust the output voltage of the grid side converter accordingly.

13. The method according to claim 1, wherein the compensation of the volt-sec error is carried out by using a PWM switching technique in the linear region of grid side converter operation, and by using an over-modulation technique in the non-linear region of the grid side converter operation.

14. The method according to claim 1, wherein positive sequence voltages and currents are used to calculate the active power and the reactive power, or wherein negative sequence voltages and currents are used to calculate the active power and the reactive power, or wherein a combination of positive sequence voltages and currents and negative sequence voltages and currents are used to calculate the active power and the reactive power.

15. The method according to claim 1, wherein, depending on regulation and system requirements, controlling the active power to the active power reference is given priority over controlling the reactive power to the reactive power reference, or vice versa.

16. A controlling system for controlling a grid side converter of a wind turbine, comprising: an inner control loop and an outer control loop;wherein the inner control loop comprises:a converter volt-sec determining unit which determines a converter volt-sec occurring at an output of a grid side converter based on a converter voltage occurring at the output of the grid side converter, wherein the output of the grid side converter is connectable via a power line to an input of a grid transformer;a volt-sec error determining unit determining a volt-sec error between the determined converter volt-sec and a converter volt-sec reference; anda controlling unit which controls, based on the determined volt-sec error, the grid side converter such that the volt-sec error is partially or fully compensated; andwherein the outer control loop is configured to determine the converter volt-sec reference based on an active power reference, a reactive power reference, a line current and a line voltage occurring at the input of the grid transformer.

17. The controlling system according to claim 16, wherein the outer control loop comprises an active power volt-sec reference determining unit and a reactive power volt-sec reference determining unit which determines an active power volt-sec reference and a reactive power volt-sec reference in a dq frame based on an active power reference and a reactive power reference requested at the input of the grid transformer.

18. The controlling system according to claim 17, wherein the active power volt-sec reference determining unit determines the active power volt-sec reference based on a difference between an active power reference and an active power occurring at the input of the grid transformer.

19. The controlling system according to claim 17, wherein the reactive power volt-sec reference determining unit determines reactive power volt-sec reference based on a difference between a reactive power reference and a reactive power occurring at the input of the grid transformer.

20. The controlling system according to claim 17, wherein the outer control loop comprises a current limiting unit which causes the active power volt-sec reference and the reactive power volt-sec reference to be respectively frozen if the amplitude of the line current occurring at the input of the grid transformer exceeds a predetermined threshold value.

21. The controlling system according to claim 16, wherein the outer control loop comprises an active power determining unit and a reactive power determining unit which determine the active power and reactive power based on the line current and the line voltage occurring at the input of the grid transformer, respectively.

22. The controlling system according to claim 17, wherein the outer control loop comprises an αβ frame volt-sec vector generating unit which generates a first volt-sec reference as an αβ frame volt-sec vector in dependence on the active power volt-sec reference, the reactive power volt-sec reference and an angle of the line voltage.

23. The controlling system according to claim 22, wherein the outer control loop comprises a second volt-sec reference generating unit which generates a second volt-sec reference by adding a grid volt-sec to the first volt-sec reference.

24. The controlling system according to claim 16, wherein the outer control loop comprises a grid volt-sec determining unit which determines the grid volt-sec based on at least one of the line current and the line voltage occurring at the input of the grid transformer.

25. The controlling system according to claim 23, wherein the outer control loop comprises a converter volt-sec reference predicting unit which predicts next sampling cycle converter volt-sec based on the second volt-sec reference, an angular velocity of the line voltage occurring at the input of the grid transformer and a sampling period of the grid side converter, and outputs the predicted converter volt-sec as converter volt-sec reference.

26. The controlling system according to claim 16, wherein the inner control loop comprises a PWM unit which controls, in dependence on the determined volt-sec error, the output voltage of the grid side converter by supplying pulse width modulation signals to the grid side converter.

27. A wind turbine, comprising: a power generator;a grid side converter, wherein the power generator is connected to the grid side converter via a generator side converter and a DC-link connecting the generator side converter with the grid side converter; anda controlling system, wherein an output of the grid side converter is connected or connectable via a power line to an input of a grid transformer, wherein the controlling system comprises an inner control loop and an outer control loop; wherein the inner control loop comprises: a converter volt-sec determining unit which determines a converter volt-sec occurring at the output of the grid side converter based on a converter voltage occurring at the output of the grid side converter;a volt-sec error determining unit determining a volt-sec error between the determined converter volt-sec and a converter volt-sec reference; anda controlling unit which controls, based on the determined volt-sec error, the grid side converter such that the volt-sec error is partially or fully compensated; andwherein the outer control loop is configured to determine the converter volt-sec reference based on an active power reference, a reactive power reference, a line current and a line voltage occurring at the input of the grid transformer.