Wind motor driven current generator of variable RPM

(12) PATENT DESCRIPTION (21) Filing number: 338639 (22) Filing date: August 7, 1998 (19) PL (11) 196763 (13) B1 (86) Date and number of international filing:

August 7, 1998, PCT / US98 / 16512 (87) International application publication date and number:

February 18, 1999, WO99 / ​​07996 PCT Gazette No. 07/99 (51) Int.Cl.

F03D 7/04 (2006.01)

F03D 7/02 (2006.01)

F03D 9/00 (2006.01)

H02P 9/00 (2006.01

The patent description was reprinted due to noticed errors (54)

Variable speed wind turbine system (73) Patent holder:

General Electric Company, Schenectady, US (30) Priority:

08/08/1997, US, 08 / 907,513 (43) Application announced:

November 6, 2000 BUP 23/00 (45) The grant of the patent was announced:

31.01.2008 WUP 01/08 (72) Inventor (s):

Amir S. Mikhail, Tehachapi, US Craig L. Christenson, Tehachapi, US Kevin L. Cousineau, Tehachapi, US William L. Erdman, Brentwood, US William E. Holley, Pleasanton, US (74) Agent:

Plewa Elżbieta, PATPOL Sp. z o. o

⁽⁵⁷⁾ A variable speed wind turbine system comprising an annular induction generator characterized in that the annular induction generator (620) is coupled to the generator torque regulator (623) using oriented excitation current control and a blade pitch regulator (609) based on o generator rotor speed.

<img file="PL196763B1_D0001.tif" />

PL 196 763 B1

Description of the invention

The present invention relates to a variable speed wind turbine system.

Wind turbines are currently attracting considerable interest as they are environmentally safe and represent a relatively inexpensive alternative energy source. And because of this growing interest, much effort has gone into designing turbines of this type that are both reliable and efficient.

Typically, wind turbines include a rotor having a plurality of blades. The rotor itself is housed in a housing that sits on top of the truss or tubular tower. The turbine blades convert wind energy into a torque or force that drives one or more generators that are rotatably coupled to the rotor via a gearbox. The gearbox gradually increases the inherently low speed of the generator turbine rotor to efficiently convert the mechanical energy into electricity that is fed to the utility grid.

Many different generators are used in wind turbines.

Many of the known wind turbines rotate at a constant speed, which allows a current of a constant frequency, for example sixty cycles per second (60 Hz), which corresponds to the American standard for alternating current, or 50 Hz, which in turn is the European standard. However, as wind speeds are constantly changing, wind turbines use either active (blade pitch control) or passive (stall control) aerodynamic control in conjunction with the characteristics of typical squirrel-cage induction generators to keep the turbine rotor speed constant.

Some turbines operate at variable speed by using a power converter to set their power output. If the speed of the turbine rotor fluctuates, the frequency of the alternating current flowing from the generator also changes. A power converter, placed between the generator and the grid, converts the AC of variable frequency to DC and then converts it back to AC of constant frequency. The total output of the generator depends on the operation of the converter (total conversion).

In US Patent No. 4,994,684, a wind turbine is disclosed which includes a bi-fed annular induction generator. A WRIG ring induction generator (Wound Rotor Induction Generator) usually consists of four main components: stator, rotor, slip rings and caps with bearings.

A variable speed wind turbine is known from US 5,083,039.

The use of variable speed wind turbines for electrical power has a number of benefits, including higher propeller efficiency compared to fixed speed wind turbines, with reactive power VAR and power factor control, and reduced loads.

Some known variable speed wind turbines are total conversion systems that use a power converter to completely straighten the entire output voltage of the wind turbine. In other words, a variable frequency wind turbine produces a variable frequency output power and then converts it to a fixed frequency to tune to the grid. Such total conversion systems are very expensive. Therefore, less expensive solutions are sought, such as, for example, a ring induction generator that uses partial conversion, where only a part of the wind turbine output is rectified and converted by the power converter.

Problems are currently encountered with the various control algorithms used by power converters to control the partial conversion process. For example, some systems have a problem with stability, which means considerable fluctuations in power and torque. Other systems cannot generate enough power without overheating essential components. It is not easy to refine them to provide a cost-effective solution for series production.

A variable speed wind turbine system comprising an annular induction generator according to the invention is characterized in that the annular induction generator is coupled to the generator torque regulator using oriented excitation current control and a blade pitch regulator based on the rotor speed of the generator.

Preferably, the blade pitch controller is a PID Proportional Integral / Differential Controller.

PL 196 763 B1

Preferably, the blade pitch controller is a PI controller.

Preferably, the blade pitch controller is a proportional differential PD controller.

Preferably, the blade pitch controller is a differential integral controller.

Preferably, the blade pitch controller is an integral-differential controller.

Preferably, the blade pitch controller is a differential member loop controller.

Preferably, the ring induction generator is a skid-free ring induction generator.

Preferably, the torque controller includes a filter to dampen the torque resulting from the oscillating motion between the turbine blades and the generator.

An advantage of the wind turbine system according to the invention is its stability and the fact that it produces considerable amounts of power at a relatively low cost. At the same time, it does not generate excessive heat.

Known solutions for wind turbine systems are illustrated in Fig. Fig. I is a cross-sectional view of a known bipolar 3-phase generator; Pos. II - the known WRIG ring induction generator system, in the block diagram; Pos. III - the relationship between the torque and the vector product of the current and flux, in the coordinate system; Pos. IV - DC ring motor, and Pos. V shows a cross-section of a bipolar 3-phase generator in Fig. I, with the flux direction plotted when only the A phase is energized.

In the cross-section (Fig. I) of a known bipolar 3-phase generator, for the sake of simplicity, the windings are shown as a pair of conductors. The generator 100 includes a stator 101, a rotor 102, a rotor phase A winding 103, and a stator phase A winding 104. Also shown is a shaft 105 engaging wind turbine blades with generator 100 via a gearbox.

On Pos. II of the drawing, showing the known WRIG induction generator ring system, the stator winding 104 is connected to a 3-phase power grid, such as a 3-phase 480V grid 201, and the rotor winding 103 is connected to the generator side converter 202 via slip rings (not shown in the figure). The stator winding 104 is also connected to a 480V 3-phase network 201, in parallel with the line side converter 203. The line side converter 203 and the generator side converter 202 are coupled to each other by a DC bus 204. Shown in Fig. In Figure II, the circuit allows the power to flow to or from the rotor winding 103. Both converters are controlled by the digital signal processor DSP 205.

The variable speed system includes a wind turbine generator with a specified power and torque which is coupled to the power grid and feeds it. The generator is a WRIG ring induction generator or a double-fed DFG (Doubly Fed Generator) generator. The rotor uses a blade pitch adjustment and variable speed operation to achieve optimal output power at any wind speed.

The ability of an induction generator to generate power is synonymous with its ability to generate torque at rotational speeds. When torque is exerted on the generator rotor in the opposite direction to its rotation, the mechanical energy of the rotor is converted into electrical energy. In an induction generator, the torque is the effect of the interaction between the current and the flux, as illustrated in Fig. Figure III, more precisely, the torque is the vector product of the current and the flux. To obtain the maximum torque for a given flux, the direction of the rotor current vector must be exactly 90 degrees to the flux direction. In the case of a DC motor, this perpendicular positioning between the flux and the armature current is maintained by means of commutators.

In the figure Fig. IV shows the mechanical structure of the ring DC motor. Due to the separation of the armature and field windings, the DC motor can be controlled by adjusting the armature current to obtain the desired output torque and also by adjusting the field winding current to obtain the desired flux.

The generation of torque in an induction generator is based on the same principle as in a DC motor. The main difference between them is that in the case of a DC motor both the armature current and flux directions are constant, while in the case of an induction generator, both of these vectors are constantly rotating.

PL 196 763 B1

Field Oriented Control (FOC) is an algorithm that identifies the flux vector and controls the torque generation according to the current.

In the figure Fig. V is the direction of the flux when only phase A of the stator winding is energized. According to the arrangement shown in Fig. Pos. II, the stator phases are fed sequentially from a 3-phase voltage source and this results in a rotating flux vector.

It should be noted here that both the flux vector and the 3-phase current vector have a specific modulus and angle, and that when the rotor current is zero, the dependence of the flux vector (Ψ) on the stator current vector (Is) can be written as the following equation:

(1) Ψ = Ls * Is where Ls is the stator inductance.

If the rotor winding is not energized, the generator behaves like an inductor, i.e. the stator current delays the stator voltage by 90 degrees, and in particular, (2) Vs = d / dt Ψ = Ls dIs / dt where Vs is the stator voltage.

An important element in FOC-oriented field control is the flux model. The stream model is used to identify the stream vector. Equation (1) is a very simple form of the flux model of the inductive system of a WRIG ring generator and indicates that the flux vector can be determined simply by measuring the stator current (Is) and the stator inductance (Ls). Using a flux model, the flux vector can be determined so that the torque can be adjusted to generate power.

Such a generator can be used in systems that have other sources that rotate the shaft coupled to the rotor of the generator, such as hydroelectric turbines, gas turbines, and common drive motor systems.

The invention is illustrated in the drawing in which Fig. 1A shows a flowchart of a wind turbine system; Fig. 1B shows a ring induction generator and a torque regulator in a block diagram; Fig. 1C shows the relationship between the flux vector and the rotor current vector in the coordinate system; Fig. 1D shows the components of the rotor current in the coordinate system; fig. 2 - a flowchart of the wind turbine governor illustrating the active / inactive sequence for the power / torque controller as well as the different operating modes of the blade pitch controller; Fig. 3 is a flowchart of a blade pitch adjustment process; Fig. 4 is a flowchart of the RPM control process; Fig. 5A is a block diagram of the blade pitch control and Fig. 5B is a PID proportional-integral-differential blade pitch controller in a block diagram.

Figure 1A shows a flowchart of a system according to the invention. The generator torque controller 603 on the variable speed converter receives the calculated torque 601, based on the measured RPM 607 and the preselected maximum torque command value 602. Calculated torque 601 is a function of the RPM of the generator based on table 640. power-speed search / plot. The output of the table 640 is divided by the measured RPM 607 by a divider 641.

The maximum torque of 602 is approximately 5250 Nm, and its selection depends on the maximum allowable current from the converter ratings. In other words, the selection is based on a velocity characteristic plot of the calculated torque for a particular turbine rotor design. In one embodiment, this selection corresponds to an excitation of 290 A.

In response to such an input, the torque regulator 603 produces a torque control signal to a generator rotor 604. The torque regulator 603 also receives a power factor control signal 642.

The rotor of the generator 604 receives the torque control signal from the torque regulator 603 and supplies power through the flux gap to the stator of the generator 605. Feedback 612 is provided from the stator output of the generator 605 to the rotor of the generator 604. The outputs of the generator rotor 604 and the stator of the generator 605 are applied. to the power grid 606.

The rotor of the generator 604 is also coupled to a measurement device 607 which gives a measured number of revolutions per minute of a generator rotor 604. Such a measurement device 607 may include an optical encoder to determine the position and rotational speed of the generator rotor 604.

The PID proportional-integral-differential controller and blade pitch speed limitation block 609 receive the measured RPM from the measuring device 607

PL 196 763 B1 and a set point of 608 operating rpm. The setpoint 608 may be set based on the same torque velocity plot that was used to establish the maximum torque setpoint. In one embodiment, the operating speed setpoint is based on a maximum value of torque and power. In this case, the operating speed setpoint 608 is 1423 revolutions per minute. In response to such an output, the PID blade pitch controller and the blade pitch rate limitation block 609 produce a voltage output.

The Variable Pitch Control 610 receives a PID speed signal from the PID controller and pitch speed limitation from block 609. The VPC 610 is coupled to the rotor of the blades 611 to control the speed of the generator 604 rotor through an input control. the aerodynamic torque of the blade rotor 611 by the blade setting act. The PID controller and blade tilt rate limitation block 609 produce the desired blade tilt rate which is converted to voltage using a table, as detailed below. The variable output voltage is brought to proportional values ​​in a hydraulic system that controls the pitch of the blades by actuating the gear at a variable speed. Thus, the variable pitch control of the vane controls the RPM by controlling the aerodynamic torque.

PID controller and blade pitch limiting block 609 together with RPM measuring device 607 and operating RPM setpoint 608, VPC 610 Variable Pitch Control and 611 blade impeller form a 650 blade pitch system, while the measuring device The RPM 607 and the remainder of the circuit shown in Figure 1A are part of the power converter and generator circuit 651. It should be noted that the measured RPM in the metering device 607 is used simultaneously in the blade pitching system 650 and in the power converter / generator 651.

The power converter controls the annular induction generator according to a predetermined power-speed curve. Thereby, the variable speed system allows the turbine to operate at the highest power factor Cp, from cut-in to rated speed, as is the case for Area II, where it is ensured that the maximum aerodynamic energy capture is achieved. Note that the power-speed curve is related to the torque-speed curve because P = Τω, where P is power, T is torque and ω is angular velocity.

The power-speed curve is encoded in the power converter in the form of a look-up table (LUT) of the power and the corresponding generator speeds. The LUT can exist as part of the hardware or software. To control the torque, the power converter measures the generator rotor speed, interpolates the LUT to determine the target turbine output, and calculates the desired generator torque from T = P / ω using the generator rotor speed. The torque can also be produced by determining the desired current vector and generating it using well known pulse width modulation techniques.

Due to slight differences between the theoretical and actual values, the power converter of the invention uses a P1 feedback-loop controller that compares the actual turbine output power with the target or desired power and makes small adjustments to the calculated torque to achieve and maintain the desired value at turbine outlet.

The power converter torque controller uses oriented FOC field control to produce generator torque as a function of generator rotor speed. Using the stator current, rotor current, and rotor angle as inputs, the torque controller identifies the flux vector and requests a desired rotor current vector which, together with the stator flux vector, produces the desired generator torque. The rotor current is produced by appropriate switching of Insulated Gate Bipolar Transistors (IGBTs) using known PWM (Pulse Width Modulation) techniques.

It should be noted that the values ​​of the power / torque-speed tables are based on the aerodynamics of, in particular, the specific wind turbine rotor and the geometry of the wind turbine rotor. Therefore, the set of table values ​​may vary for different turbine rotors.

PL 196 763 B1

An exemplary embodiment of a torque controller and the corresponding parts of an annular induction generator are shown in Fig. 1B. The torque can be expressed in the form of the relationship:

(3) Td = k * ^ * Irq where k is the generator parameter. From the point of view of the controller, equation (3) takes the following form.

(4) Irq = Td / (k * T)

Equation (4) gives the absolute value of the rotor current for "desired torque Td, which is the output of the torque regulator 623.

In Figure 1B, torque controller 623 includes a power table 623a, a PI controller 623b, a divider 623c, a switch 629, and comparators 623d and 623e, which may be part of hardware or software to produce different values, and a forward coupled damping filter. 623f. Power table 623a is a LUT that receives a generator speed signal from measuring device 607 and outputs a target power value corresponding to the generator speed. An embodiment of the power board 623A is shown in Table 1 below.

Table 1

<td>Generator speed (revolutions per minute)</td><td>Electric power (kW)</td>

<td> 750</td><td> -177</td>

<td> 800</td><td> -177</td>

<td> 850</td><td> 135</td>

<td> 900</td><td> 167</td>

<td> 950</td><td> 203</td>

<td> 1000</td><td> 247</td>

<td> 1050</td><td> 287</td>

<td> 1100</td><td> 335</td>

<td> 1150</td><td> 388</td>

<td> 1200</td><td> 450</td>

<td> 1250</td><td> 507</td>

<td> 1300</td><td> 575</td>

<td> 1350</td><td> 647</td>

<td> 1400</td><td> 743</td>

<td> 1450</td><td> 750</td>

<td> 1500</td><td> 750</td>

<td> 1550</td><td> 750</td>

<td> 1600</td><td> 750</td>

The target output power is compared by a comparator 623d to generate a difference between the target output power and the actual output power. The difference thus obtained is an input to PI controller 623b, which controls the power as described above. The divider 623c receives the adjusted power from the PI controller 623b and the generator speed signal from the metering device 607, and the output is the command torque signal.

This commanded torque may be corrected by a torque value generated by the forward coupled damping filter 623f. This filter detects oscillations (at resonance)

A non-stiffened (flexible) shaft (not shown in the drawing, so as not to obscure the invention), caused by its coupling between two independent inertias - one is the gearbox and the generator, the other is the turbine blades. When an oscillation is detected, damping filter 623f applies a counter-torque to reduce the relative motion between the one inertia and the other. The attenuation filter 623f includes a bandpass filter whose center of the passband is at the resonant frequency of the two inertia and the shaft.

The commanded torque thus obtained is applied to the wound rotor of the induction generator.

In response to a deceleration indication (e.g., signal (s)), switch 629 is triggered to shift the commanded torque to maximum constant torque 660, as described below.

To generate the torque, the rotor current component, Irq, is controlled to follow a direction perpendicular to the flux direction (Fig. 1D). The absolute value of Irq is determined by the relationship:

Irq = Τΰ / (Κ * Ψ) where k is the generator parameter. The rotor current, Irq, produces the generator flux but has no effect on the torque production.

The rotor current component block 622 receives the commanded torque and the flux vector scalar component from the rectangular to polar coordinate converter 626, which converts the flux vector from flux model 621 to polar coordinates. In response to these input signals, rotor current component block 622 generates a rotor current torque component, Irq.

The stream model 621 identifies the stream vector. To this end, the current converter blocks 621a and 621b receive the stator current vector and the rotor current vector. It should be noted that since the current vector can be determined by measuring two of the three phase currents, only two current sensors (not shown) are necessary. The stator current vector with rotor angle 621b of generator 620 is input to frame convert 621c. This block converts the stator current into a fixed rotor frame. Based on the outputs of frame convert 621c, the stator inductance Ls at block 621d is determined. From the rotor current vector, the rotor inductance Lr can be obtained at block 621f. A flux vector is generated from the stator inductance Ls and the rotor inductance Lr.

After the flux vector is determined, the rotor current vector from inverter 624 is aligned in the perpendicular flux direction so as to generate the torque. Since the rotor current is determined with respect to the rotor assembly, the rotor current control signal depends on both the flux angle and the rotor angle. In particular, the flux angle is first converted to a fixed rotor reference frame, and in this reference frame, the rotor current command direction is a direction perpendicular to the flux direction. This procedure is illustrated in Fig. 1C.

By using the rotor current component, Irq, in conjunction with the inductive portion of the output from the square-to-polar converter block 626, a reference current is generated at the input of the converter 624. The drawing also shows a converter 630 connected to the stator side (line side) of the generator 620 and, via a DC bus 631, to the converter 624.

When this rotor current is forced to flow through the rotor winding, a desired torque Td is produced and a power Td * f &lt;is generated, where ω is the rotor speed. This power is generated in the form of stator current that flows back into the grid. This power-carrying stator current is in phase with the stator voltage.

When the power is produced by the generator, the flux model described by equation (1) is no longer valid because the stator current Is now has two components: flux generation component and power transmission component. The power transfer component does not affect the flux production since the current component has the same absolute value (when normalized by the winding ratio) as the torque producing the rotor current, but in the opposite direction. In other words, the flux produced by both of these current vectors (ie, the current stands against the power carrying and the torque producing rotor current) adds up to zero. To remove the power transfer component from the stator current measurements, the rotor current, Ir, was added to equation (1), and thus:

Ψ = Ls * Is + Lr * Ir

Where Lr is the rotor inductance. Ls and Lr differ in the ratio of the windings.

It should be noted that in the operation described herein, while the power carrying stator current component is in phase with the stator voltage, the flux generating component delays the stator voltage by 90 degrees. This flux generating component results in a non-uniform stator power factor. Since the flux-generating current automatically lags the voltage by 90 degrees, flux is produced by the rotor winding to achieve a uniform factor on the stator side.

To produce flux through the rotor winding, an additional component, Ird, of the rotor current must be commanded. This additional component should extend along the flux direction as shown in Figure 1D.

As the flux-producing component of the rotor current, Ird, increases, the flux-producing stator current decreases. This is because the absolute value of the flux is kept constant due to the constant stator voltage (as seen from equation (2)). The component Ird producing the rotor current can be controlled in such a way that the produced flux induces a voltage equal to the mains voltage. That is, the induced voltage is in phase with the mains voltage and has the same absolute value as the mains voltage. In this case, the induced voltage counteracts the mains voltage and therefore the stator winding does not consume any stator current. It is therefore a case of a uniform power factor system.

It should be noted that reactive power VAR / power factor control 670 may be included in the VAR generation control system. The product of the stator voltage Vs and the stator current vector Is (when no torque is produced) represents the magnetizing reactive power VAR required by the generator.

The power converter only works when it is activated. The turbine controller turns the power converter on and off, as shown in Figure 2, block 705. The turbine controller may be implemented as hardware or software, or a combination thereof. The solution according to the present invention uses binary logic voltage signals to turn the power converter on and off, referred to herein as the converter start signal.

When the turbine controller is in normal operation, referred to herein as auto mode, the turbine controller tilts the turbine to the wind and tilts the turbine blades to the full power position. Such a position is fully understandable to a skilled person. With sufficient wind force, the blades begin to rotate and the speed of the generator increases. When this speed reaches the pre-selected converter start speed, the turbine controller sends a converter start signal. In an exemplary embodiment, this speed is 820 revolutions per minute.

In response to the start signal, the converter starts a startup sequence. The system initially closes an AC line contactor (in converter 630), which involves connecting the line matrix (in converter 630) to the network. A predetermined delay allows the contactor to close and calm the transients. This delay is 1.5 seconds. An embodiment of the actuation sequence is described in detail with reference to Fig. 2 and blocks 714, 715, 716, and 717.

After the contactor closes, a bus regeneration cycle takes place to ensure that the bus is fully regenerated and allow instant torque adjustment. In this case, the DC bus voltage is regulated to a predetermined value. The preferred voltage is 750 V DC. Another delay can be used to ensure that the bus is regenerated enough to regulate properly. This delay may be up to 5 seconds. If the bus fails to regulate, the generated voltage is too low or too high, a converter failure signal is sent to the turbine controller.

When the generator speed reaches or above the preselected speed and the predetermined bus delay time has elapsed (i.e. after the bus has fully recovered for 5 seconds), the stator contactor closes (block 714), energizing the stator windings and producing a rotating flux. the stator. The stator windings are energized only. Due to the inductance of the stator windings, the current surge is very low and amounts to only 75% of the maximum operating current. The speed selected is 900 rpm. The delay can be used to close the stator contactor and calm transients. This delay is 3 seconds.

When the generator speed reaches or exceeds the preselected value and the rotor voltage is determined to be lower than the preset peak voltage value, the contactor

The rotor will be closed (block 715), attaching the generator matrix to the rotor of the annular induction generator. The pre-set speed is 1000 rpm and the pre-set peak voltage is 318 V. The delay can be used to close the contactor. This delay is 1/2 second. Before this time has elapsed, the IGBT-gate bipolar transistors on the rotor side (in converter 624) will not be switched. There is no current flowing and no transients, and no power is produced. Since there is no real power (only reactive power), no torque pulses are produced.

Power generation begins with the gating of the rotor-side IGBT insulated gate bipolar transistors, which produces a vector of the current (both absolute and position) required to generate the desired torque. The current vector is produced in response to a command from a torque controller (e.g., a processor). Initially, this torque increases from 0 to a value determined by the Optimum Power / Torque-Speed ​​curve. This steady rise (Block 716) eliminates sudden spikes in power and torque and also allows the turbine to turn on smoothly directly.

The timing of the present invention differs from the traditional "timing" process used in synchronous or cage induction machines. With the present invention, there are no sudden surges, transients, or power oscillations accompanying the starting of the turbine.

Once synchronized, the power converter behaves according to the power-speed relationship described above (block 717) until it is turned off by the turbine controller.

It should be noted that the delays discussed above regarding the converter starting sequence may be adjusted depending on the components used in the system and the ambient conditions of the turbine. These adjustments can be made using appropriate software, devices, or both.

The power to the turbine is supplied by the wind. If its speed changes, so does the power at the turbine input. In order to compensate for changes in the input power, the present invention provides a process for updating the torque of the generator. As the generator torque is (momentarily) determined by the power converter, the speed of the generator increases according to the formula for power P = Tω. The power converter, which continuously samples the generator speed, recognizes that the speed has changed and identifies the new speed, and updates the requested power from the LUT. The power converter determines a new torque from the requested power and, based on the FOC oriented field control system, calculates a new current vector that is applied to the generator rotor. The update process takes place every 33 milliseconds or every 2 cycles for a line at a frequency of 60 Hz. This causes the turbine to run smoothly and accurately in accordance with the relationship between power and speed. Note that the update rate may vary or may change dynamically as you run.

At below-rated wind speeds (eg, Area II), the blades are held at an angle assumed for a preselected power, and the resulting generator / turbine speed is dependent on the commanded torque and wind power input. This ensures that the power-speed curve has been selected correctly. The angle used for the preselected power is the angle used for maximum power (ie, 0, 1, or 2 degree blade pitch). The number of degree changes varies as a function of the wind speed.

Rated power occurs at a predetermined generator rotor speed. In an embodiment of the invention, this speed is 1423 revolutions per minute. Above the rated wind speed, the generator rotor speed is controlled by a PID blade pitch controller which deflects the blades in response to an indication of the generator rotor speed. These indications can take various forms, including a signal or speed values ​​stored in the recording equipment. It is important that the PID controller works independently of the power converter. If the converter fails, the PID controller will keep the generator speed (for example, 1423 rpm) by issuing commands to increase the blade pitch angles. Thus, the system has a fail-safe operation built in.

For a generator speed equal to or greater than the speed at which the rated power is achieved (such as 1423 revolutions per minute or more), the power-speed relationship is such that the power converter maintains the power constant without significant fluctuations. Hence, rated wind gusts, which tend to increase turbine speed, have little effect on generator power, as the PID controller responds by regulating the generator rotor speed.

PL 196 763 B1. The answer of the PID controller is that it is able to effectively control the rotor speed, and hence power spikes in the range of about 5%, ensuring that approximately constant power is produced for a wind speed equal to or greater than the rated wind speed.

Rated power spikes have no effect on the line voltage if the power overload is caused by the rotor of the annular induction generator, as the stator power remains constant. The rotor current (and the stator current) are kept constant during the spikes mentioned by the power converter and the torque is kept constant (the torque is proportional to the current). Since the rotor current is constant during the gusts of wind, the increase in rotor power depends on the increase in rotor voltage. But this voltage surge has no effect on the grid as the power converter, positioned between the generator rotor and the grid, electronically translates this varying rotor voltage (and frequency) into a constant AC waveform (e.g., 60 cycles, 460 V).

The VPC blade pitch control system works in real time, is separated into a system with an automatic pitch control servo and a wind turbine rotor speed control system. The VPC system monitors and controls the pitch position of the blades, pitch speed and the rotation speed of the generator.

A pitch position encoder provides an analog signal which is proportional to the pitch position of the blades and which is later converted into a digital signal to identify the current position of the turbine blades. The blade actuator coupled to the blades is used to mechanically change the pitch of the blades.

Figure 2 is a flowchart of an embodiment of a blade pitch adjustment system. The control or processing logic in this system performs some operations in conjunction with the electrical / mechanical tooling in the system. The control / processing logic may be implemented by devices, software, or a combination of them, such as a computer or controller chip.

Referring to Fig. 2, the pitch adjustment process begins by measuring the rotor speed (block 701). At the same time, the system determines its operating status (block 702). The test determines whether the control system is in auto mode (block 703). If not, the test determines that the generator rotor speed (in revolutions per minute) is less than a predetermined speed (block 704). In an embodiment, this predetermined speed is 1035 rpm. If the system is not in auto mode and the rotor speed is less than a predetermined value, the power converter gets an input in the shutdown sequence (block 705), otherwise the system remains as it is.

If the system is in auto mode, processing continues at block 706 where it is determined whether the generator rotor speed is increasing. If not, the generator rotor speed is checked to be less than a predetermined setpoint (block 707). This value is 835 revolutions per minute. If the generator rotor speed does not increase and is less than 835 rpm, the power converter receives a signal to enter the disable state (block 705), otherwise the system remains as it is.

The disabling process includes decreasing the rotor current (block 708), disconnecting the generator rotor (block 709), and disconnecting the generator stator (block 710).

If the rotor speed is increasing as found at block 706, the test indicates if the speed is greater than 100 rpm (block 711). If so, the slope is set to a predetermined set point (processing block 713). In an embodiment, this predetermined set point is zero degrees. In other examples, the tilt can be set to any number of degrees, such as one, two, or three degrees. In another embodiment, the predetermined amount is variable.

If the rotor speed is greater than 100 rpm, the test determines if the rotor speed is greater than a predetermined speed (block 712). This speed is 820 revolutions per minute. If the rotor speed is greater than this predetermined speed, the converter receives a signal to start the start-up sequence (processing block 705). Therefore, in this solution, the power converter will be activated when the rotor speed is greater than 820 rpm.

In one embodiment, the startup sequence includes the following steps. First, the generator stator becomes networked (block 714). After it is attached, the generator rotor (block 715) is attached. After the generator rotor is connected, the generator rotor current component, Ird, steadily increases (block 716) and then the torque is regulated (717). Yeah

The starting sequence is a passive synchronization technique that connects the generator to enter directly with a rotor current of 0. This allows vector control in conjunction with an annular induction generator.

If the test shows that the generator rotor speed is increasing but has not yet exceeded 100 rpm (block 711), the slope is set to a predetermined number of degrees (block 718). Here, the slope covers a set of 25 degrees. Note that this set of slopes may vary. The slope should be selected so as to obtain additional thrust to aid faster acceleration of the turbine.

The present invention also relates to the tilt position portion of the system. First, it is necessary to measure the position of the slope using well known measurement techniques (block 720). The error is then computed between the actual slope and the predetermined set slope (block 721).

After calculating said error, it is amplified (block 722). After the error amplification and velocity measurement (block 701), the rate of change of the dynamic pitch of the blades is limited (block 723).

After limiting the rate of change to a predetermined amount, it is determined whether the generator rotor speed is greater than the predetermined value. In an embodiment, this predetermined value is 1423 revolutions per minute. If the generator rotor speed does not exceed the preset value, the pitch control system enters the pitch fixed position mode (block 726), otherwise the pitch control system enters the RPM control mode (block 727).

Tilt adjustment is concerned with maintaining the blade pitch angle in the designed operating position for operation below rated power. For example, the position is 0 degrees. However, other positions are also possible. The VPC blade pitch control system adjusts the pitch of the blades by applying a negative voltage which causes the index roller to move from its original position (i.e. 90 degrees) or the bias position of the propeller into a sick note, with a constant speed of several degrees per second (e.g. 1.0) towards its nominal setpoint of zero degrees.

In the present invention, a position control voltage is applied to the error amplifier to produce an error output that is proportional to the difference between the command position Pc and the return position Pf. In one embodiment, the error amplifier is software implemented. However, such amplification may be implemented in hardware.

The error output signal is amplified and sent to the proportional valve. The position rate limiter is used to limit the slope rate initially to one degree per second. This limits the acceleration of the rotor in both low and high winds and allows a smooth transition to generation without problems with excessive rotational speed.

When the turbine reaches its zero degree position, the proportional amplifier helps to hold that position by creating a voltage that is proportional to any error that might occur due to a drop in hydraulic system pressure. If, during the initial pitch to the working pitch angle, the speed of the generator does not exceed a predetermined rate (e.g., 100 rpm), the system tilts the blades to a predetermined amount (e.g., 25 degrees). This helps to start turning the impeller in very light winds. If the speed of the generator is greater than the predetermined speed, the system tilts the blades to a nominal position of zero degrees.

The slope adjustment is at and above rated power (ie, region II) when the generator speed is less than its nominal set value (eg 1423 rpm). During the transition from under-nominal to over-nominal size, the PID system begins to pitch the blades in the lug direction of the propeller before the generator speed reaches the nominal steady value (e.g. 1423 revolutions per minute), depending on the acceleration of the generator rotor speed signal (via, e.g., block 607).

Adjusting the slope below the assumed power does not require a full PID system due to the variation of the travel speed, which is limited to only one degree per second.

Figure 3 shows the slope positioning mode. In block 800, a slope position error amount is calculated that is proportional to the difference between the command position Pc and the feedback position Pf. It then determines (block 801) whether the slope error is positive. If the pitch error is not positive, it is determined (block 803) whether the rotor speed is greater than the first predetermined speed command. In one embodiment, a predetermined set point

The speed is 1,200 rpm as measured at block 802. If the pitch error is not positive and the rotor speed is not greater than the first predetermined speed set point, the process continues with block 804 where a ramp rate limitation is set. equal to Y1 and is input to the dynamic pitch limiter 805.

If the rotor speed is greater than the first predetermined speed set point, the test determines (block 806) whether the rotor speed is greater than the second higher predetermined speed set point. In one embodiment, the second higher predetermined speed set point is 1250 rpm. If the rotor speed is greater than the second higher predetermined speed set point, the process continues with block 807, where the pitch rate value Y is set to -Y2 and is input to the dynamic pitch rate limiter 805. If the rotor speed is not greater than the second predetermined speed command, the Y pitch limit is set as a function of the rotor speed (block 808), which is between -Y1 and -Y2, ​​and the Y pitch limit is sent to the limiter. dynamic slope rate (block 805). The function is a linear function of the slope rate limiter that increases linearly from minimum to maximum. If the pitch error is positive, it is determined whether the rotor speed is greater than a third predetermined speed set point (block 809). In an embodiment, the third predetermined speed set point is 1100 rpm. If the slope error is positive and the generator rotor speed is not greater than a third predetermined speed set point, the process continues at block 810, where the slope rate threshold Y is equal to Y1 and is input to the dynamic pitch rate limiter (block 805).

If the rotor speed is greater than a third predetermined speed set point, the test determines whether the rotor speed is greater than a fourth predetermined speed set point (block 811). In one embodiment, the fourth predetermined speed set point is 1150 rpm. If the rotor speed is greater than a fourth predetermined speed set point, the process continues at block 812 with the slope rate limit Y set as Y2 and input to the dynamic pitch rate limiter (block 805). If the rotor speed is not greater than a fourth predetermined speed setpoint, then the slope rate limit Y is set as a function of the rotor speed (block 813), which is between Y1 and Y2, and is sent to the dynamic pitch rate limiter (block 805) . Thus, this function takes the opposite direction to that of block 808 described above. In one embodiment, the function is a linear ramp rate limiter function that is uniformly sloped between Y1 and Y2 between the maximum and minimum, respectively.

The slope position error value determined at block 800 is amplified (block 814) and is input to the dynamic pitch rate limiter (block 805). In response to the pitch rate limit Y and the amplified pitch position error value, pitch rate variation is initially limited to one degree per second to limit the rotor acceleration in both low and high winds and to allow a smooth transition to generation without the hassle of with excessive speed.

The test determines whether the rotor speed measured at block 802 is greater than a fifth predetermined speed set point (block 815). In one embodiment, the fifth predetermined speed set point is 1423 rpm. If the measured rotor speed is greater than a fifth predetermined speed set point, the system enters the RPM control mode (block 816). On the other hand, if the measured rotor speed is not greater than a fifth predetermined speed set point, the pitch rate is set to a programmed amount (block 817), which may be a binary voltage, and the process continues at block 818.

In block 818, it is determined whether the system is in the auto run mode. In an embodiment, this test is performed by checking that the system is in the standby / fault state in the event that a fault is detected at block 819. If the system is not in auto run mode, process continues at block 820 where pitch control is reset to manual to shut down the system. In one, the system is turned off by tilting the blades 90 °. If the system is in auto mode, the binary voltage representing the programmed quantities is converted to analog (block 821) and drives the proportional hydraulic valve (block 822).

A single D / A D / A converter produces the voltage required by a proportional hydraulic valve. The voltage is directly proportional to the speed of the hydraulic tilt cylinder, i.e. the rate of change of the pitch of the blades. Positive voltage causes the blades to tilt towards the lug position (blade offset), while negative voltage causes the blades to tilt towards the power (slope towards the power). The slope rate is controlled by the amplitude of the D / A converter output voltage. In an embodiment, the sampling rate of the D / A converter output is fixed at 10 Hz.

The VPC system regulates the speed of the generator. In one embodiment, the speed of the generator is controlled by a proportional-integral-differential PID controller for the pitch angle of the turbine blades. The VPC system calculates and then amplifies the error signal, via software, to produce an error output signal that is proportional to the difference between the reference rate (e.g. 1423 revolutions per minute), denoted herein as Rc, and a feedback rate, denoted as Rf. The present invention uses this output to produce the PID quantities required to properly control the speed of the proportional valve, and thus, the blade pitch angle.

As the rotor speed approaches a predetermined setpoint speed (e.g., 1423 rpm), the PID controller generates a voltage that tilts the blades in the direction of the lug propeller. Conversely, when the rotor speed drops below a predetermined set speed (e.g. 1423 revolutions per minute) the PID controller produces a voltage that tilts the blades in the power direction until it again reaches the nominal pitch setting or exceeds a nominal predetermined setpoint speed (e.g., 1423 rpm).

In one embodiment, a table is used to change the value of the slope rate produced by the PID controller logic to a specific voltage applied to the proportional valve. An example of a table is shown in Table 2. The maximum pitch-to-pinning speed of the propeller is 12 degrees per second, while the maximum pitch-to-speed (during speed control) is 8 degrees per second. This corresponds to the output voltage of the D / A converter of 5.1 and 4.1, respectively.

Table 2

Table for converting the inclination rate into driving voltage

<td>Voltage [V]</td><td>Speed ​​[degrees / s]</td>

<td> 1</td><td> 2</td>

<td> -8,25</td><td> -20</td>

<td> -7,90</td><td> -19</td>

<td> -7,55</td><td> -18</td>

<td> -7,20</td><td> -17</td>

<td> -6,85</td><td> -16</td>

<td> -6,50</td><td> -15</td>

<td> -6,15</td><td> -14</td>

<td> -5,80</td><td> -13</td>

<td> -5,45</td><td> -12</td>

<td> -5,10</td><td> -11</td>

<td> -4,75</td><td> -10</td>

PL 196 763 B1 cont. table 2

<td> 1</td><td> 2</td>

<td> -4,40</td><td> -09</td>

<td> -4,05</td><td> -08</td>

<td> -3,41</td><td> -07</td>

<td> -3,12</td><td> -06</td>

<td> -2,88</td><td> -05</td>

<td> -2,67</td><td> -04</td>

<td> -2,34</td><td> -03</td>

<td> -1,96</td><td> -02</td>

<td> -1,45</td><td> -01</td>

<td> 0,00</td><td> 00</td>

<td> 1,83</td><td> 01</td>

<td> 2,33</td><td> 02</td>

<td> 2,71</td><td> 03</td>

<td> 3,12</td><td> 04</td>

<td> 3,46</td><td> 05</td>

<td> 3,79</td><td> 06</td>

<td> 4,08</td><td> 07</td>

<td> 4,25</td><td> 08</td>

<td> 4,45</td><td> 09</td>

<td> 4,65</td><td> 10</td>

<td> 4,85</td><td> 11</td>

<td> 5,05</td><td> 12</td>

<td> 5,25</td><td> 13</td>

<td> 5,45</td><td> 14</td>

<td> 5,65</td><td> 15</td>

<td> 5,85</td><td> 16</td>

<td> 6,05</td><td> 17</td>

<td> 6,25</td><td> 18</td>

<td> 6,45</td><td> 19</td>

<td> 6,65</td><td> 20</td>

Note that in Table 2, negative pitch rate is pitch toward the power direction, while the zero position or positive pitch rate is pitch towards the lug.

The valve control switch disengages the proportional valve during STOP and STANDBY modes as commanded.

Figure 4 shows one embodiment of the method for regulating the RPMs per minute. In block 900, a rate deviation value is calculated that is proportional to the difference between the set speed Pc at block 930 and the measured Pf at block 902.

PL 196 763 B1

The rpm deviation is checked to be positive (block 901). If the speed deviation is not positive, it is checked to see if the rotor speed is greater than the first predetermined speed set point (block 903). In one embodiment, the first predetermined speed set point is 1200 rpm. If the rpm deviation is not positive and the generator rotor speed is not greater than the first predetermined speed set point, the process continues at block 904 where the pitch rate limit is -Y1 and is sent to the dynamic pitch rate limiter 905.

If the rotor speed is greater than the first predetermined speed set point, the test determines whether the rotor speed is greater than the second higher predetermined speed set point (block 906). In one embodiment, the second predetermined set speed is 1250 rpm. If the generator rotor speed is greater than the second predetermined setpoint speed, the process continues at block 907 where the pitch rate threshold is -Y2 and is input to the pitch rate limiter 905.

If the generator rotor speed is not greater than the second higher predetermined speed set point, the pitch rate limit Y is set as a function of the rotor speed (block 908). In one embodiment, this function is a linear function of the pitch rate limiter that is uniformly sloped between -Y1 and -Y2. The pitch rate value Y is sent to the pitch rate dynamics limiter (block 905).

If the rate deviation is positive, the pitch rate threshold Y is set to the value Y2 (block 912) and is input to the pitch rate dynamics limiter (block 905).

After calculating the rate of the rate error, the PID system determines if the acceleration is too high and sets the slope accordingly (block 940). In response to the pitch cutoff Y and the output of PID loop 940, the pitch rate is initially limited to one degree per second (block 905).

It is then determined whether the measured rotor speed (block 902) is greater than a third predetermined speed set point (block 915). In one embodiment, the third predetermined speed set point is 1423 rpm. If the measured rotor speed is less than a third predetermined speed set point, the system enters the tilt position mode (block 916). On the other hand, if the measured rotor speed is greater than a third predetermined set speed, the pitch rate is converted, using the pitch rate, into the driving voltage conversion table described above (block 917) and the process continues at block 918.

In block 918, it is determined whether the system is in the auto run mode. This test is performed by verifying that the system is in the EMERGENCY / STOP mode for failure in the event that a failure is detected at block 919. If the system is not in auto run mode, process continues at block 920 where the pitch control is reset to manual to shut down the system. In one embodiment, the system is turned off by tilting the blades 90 °. If the system is in auto mode, the voltage representing the programmed quantities is converted to analog (block 921) and drives a proportional hydraulic valve (block 922) to initiate tilt action (block 922).

Fig. 5A shows an embodiment of a tilting system. It includes a PID loop driver 1010 and nonlinear arrays 1011 for converting the slope rate input signal to a voltage output signal. The slope rate values ​​obtained from the table 1011 are generated by the PID controller 1010 in response to the difference in the output speed and the set speed as determined by comparing the logic or software. The output voltage signals from table 1011 are applied to a proportional valve which tilts the blades.

A diagram of the signal flow in an embodiment of the PID controller is shown in figure 5B of the drawing. The difference between the feedback position value Pf and the position reference Pc is determined by logical comparison (e.g., a subtractor) or by software 1001. The difference represents a position deviation. The position error is amplified by the scaling factor K of the amplifier 1002 to produce the quantity yc. The K factor is set to a value of 0.5. The value of yc is applied as an input to the limiter 1005 which is controlled by the limiter driver 1004. The limiter 1005 limits the pitch rate of the blade16.

When moving the tilt position. This speed is slow. The controller 1004 receives the generator speed feedback signal and in response to this signal changes the limiter 1005 based on the generator speed (in revolutions per minute). The limiter controller (block 1004) changes the maximum pitch to a lug propeller or pitch to the power position using a linear function of the measured RPM, Rf.

The PID controller also includes a comparator logic (e.g., a subtractor) or software 1003 to produce a difference between a predetermined generator speed Rc and an actual generator speed Rf. The output of comparer 1003 is the amount of rate deviation x that is received by the PID algorithm blocks 1006 and 1007. The PID algorithm (blocks 1006 and 1007) calculate the desired slope rate based on the proportional, integral, and differential functions of the rate deviation value. The slope rate output as a function of the rate deviation input may also include a gain selection that adjusts the gain as a function of the slope position. The speed scheduler (block 1012) provides a multiplier E based on the pitch position feedback and two setpoint parameters E1 and E2. In one embodiment, the two parameters are -0.85 and 0.0028, respectively. The output of block 1005 is coupled to output 1006 and yf and applied to block 1008. A limiter 1005 limits the maximum pitch rate to pinning and sliding to the power position during the speed control mode.

The output of the limiter 1008 provides a voltage input to the voltage generator 1009 and feedback to the PID algorithm block 1007. The output of the voltage generator 1009 is coupled to the input of the switch 1010, which is controlled to turn off the proportional valve in response to a turbine stop command. The output from switch 1010 is coupled to a D / A converter 1011 that provides a voltage output to the system that powers the proportional valve to control the pitching operation.

In order to achieve dynamic braking, the torque-speed curve in the present invention may be deliberately sloped. The power converter applies a maximum constant torque. This maximum constant torque is applied to the system in response to the failure conditions, causing the turbine speed to decrease. Figure 6B shows maximum constant torque power converter and switch 629.

The safety system initially applies gentle braking and tilts the blades up to 90 degrees. Then it is determined whether it was a defect. Dynamic braking is only used in response to severe damage.

In response to a determination that severe damage is occurring, the blades are adjusted to 90 degrees and the maximum constant torque is applied. Torque is exerted on the generator rotor, which causes the turbine speed to decrease. The turbine is slowed down to a predetermined speed. Upon reaching this predetermined speed, the braking may be released either automatically or manually (e.g. manually set by the operator).

Since the power converter controls the rotor current directly, the overall power factor system can be dynamically adjusted from lag 0.90 to lead 0.90 depending on the turbine output level. In the solution according to the invention, the reactive power drivers VAR are supplied to the second inductive generator. Thus, the power converter may act as a VAR compensator. This is done by the control system that applies a specific number of kilovoltamperes kVAR from each turbine through the SCADA system. In Fig. 6B, an input 670 for controlling the VAR is shown. By matching the supply of the VARs to the secondary supply, the entire VAR system can be selected dynamically.

The desired power factor may be set to any nominal value between 0.9 lag and 0.9 lead, or vary in response to voltage oscillations in the network. Thus, a power converter operating via SCADA can operate in constant power factor mode, constant VAR mode, or voltage regulation mode.

One benefit of power conditioning according to the present invention is that it provides maximum energy capture, torque control, voltage flicker elimination, and power factor control. Additionally, it is possible to dynamically adjust the power factor. Moreover, the variable speed ensures a smoothing of the torque pulses. Torque transients that cause voltage flicker and damage to train propulsion components are suppressed by allowing an increase in speed

Of the rotor, "thus storing additional wind gust energy in the rotational inertia of the rotor blades. This energy can be recovered and fed back into the grid by reducing the rotor speed as a reflection of gusts of wind, or it can be "dampened by pitching the blades downwind." Thus, variable speed operation can significantly reduce torque transients, which translates into lower costs and longer life of wind turbine driving parts.