Multi-objective cooperative control method for offshore doubly-fed wind turbine considering variable coupling

Abstract

The application provides a kind of offshore doubly-fed wind turbine multi-objective collaborative control method considering variable coupling, first, a prediction model is established, the prediction object includes generator active and reactive amplitude and rate of change, grid-side converter active and reactive amplitude and rate of change and output voltage amplitude abrupt variable;Second, a multi-objective optimization problem is established, multiple optimization objectives include: (1) improve the accuracy of various power servo tracking of doubly-fed wind turbine;(2) improve the amplitude range of specific power in various powers of doubly-fed wind turbine;(3) improve the rate of change range of specific power in various powers of doubly-fed wind turbine;(4) reduce the grid-side converter output voltage amplitude abrupt variable;Finally, the model predictive controller solves the multi-objective optimization problem according to the fixed time period, obtains the optimal control amount and executes.The method realizes the multi-objective collaborative optimization control of various powers of doubly-fed wind turbine under coupling.

Description

Technical Field

[0001] This invention relates to the field of wind turbine power control, and specifically to a multi-objective cooperative control method for offshore doubly-fed induction generators that considers variable coupling. Background Technology

[0002] With the continuous depletion of non-renewable energy sources such as oil and coal, and the increasing emphasis on high-quality living environments, many countries and regions around the world are vigorously developing clean and renewable energy power generation technologies. Offshore wind power has experienced rapid development due to its advantages such as being clean and renewable, having great power generation potential, and being quick to invest and construct. As the installed capacity of offshore wind power continues to increase, its power output has a growing impact on the power system. To reduce the adverse effects on the power system after large-scale offshore wind power grid connection, higher requirements have been placed on the control of offshore wind turbines, and relevant standards have been formulated.

[0003] For example, to mitigate the negative impact of large-scale offshore wind power grid connection on power system power quality, researchers in the 2005 version of the national standard "Technical Regulations for Wind Turbine Units Connected to Power Systems" (GB / T19963—2005) proposed indicators such as voltage deviation, voltage fluctuation, flicker, and harmonics to evaluate the impact of wind farm output power on power system power quality, and imposed restrictions on the voltage and current output of offshore wind turbine units based on these indicators. To avoid more severe faults caused by large-scale off-grid disconnection of offshore wind farms during power system failures, researchers in the 2011 version of the national standard "Technical Regulations for Wind Turbine Units Connected to Power Systems" (GB / T19963—2011) required wind farms to possess low-voltage ride-through capability and the ability to regulate reactive power output during low-voltage ride-through to aid power system voltage recovery. Furthermore, reactive power regulation must meet the amplitude and response time requirements given in the standard. This necessitates that offshore wind turbine units can suppress the amplitude and rate of change of active power output during low-voltage ride-through, regulate reactive power output over a wide range, and regulate reactive power output with a relatively large rate of change. Furthermore, due to the large-scale presence of wind turbines that do not provide frequency regulation services crowding out the capacity of synchronous turbines that do provide such services, the equivalent inertia and frequency regulation capability of the power system are continuously declining. To reduce the impact of large-scale wind power grid connection on the power system's frequency regulation capability, the 2021 version of the national standard "Technical Specifications for Wind Turbine Units Connected to the Power System Part 1: Onshore Wind Power" (GB / T19963.1—2021) sets forth frequency regulation requirements for onshore wind farms, which can also be referenced for offshore wind farms. The standard states that when the power system frequency deviation exceeds the dead zone, the wind farm should change its active power output to participate in power system frequency regulation, and the active power regulation should meet the amplitude and response time requirements given in the standard. This necessitates that offshore wind turbines can adjust their active power output over a wide range and at a relatively large rate of change during frequency regulation.

[0004] It is evident that when controlling offshore wind turbines, in addition to the most basic requirement of ensuring that various power outputs servo-track reference values, it is also necessary to consider improving the output voltage quality and adaptively increasing the output range and rate of change range of a certain type of power under specific scenarios. As one of the mainstream turbine models, offshore doubly-fed induction generators have been the subject of numerous control methods proposed by researchers, but no control method or controller has yet considered the coordinated control of the aforementioned multiple control objectives.

[0005] In invention patent application number 202010903545.9, the inventors disclosed a model-predictive multi-objective optimization control method for wind turbines. This method uses current control accuracy, frequency control accuracy, and electromagnetic torque control accuracy as optimization objectives, effectively enhancing system frequency stability during system frequency disturbances while suppressing shaft oscillations caused by frequency disturbances. However, this method does not consider improving the voltage quality of the wind turbine or adaptively increasing the output range and rate of change of a specific power type of the wind turbine. Patent application number 201610312708.X discloses a multi-objective optimization control method for doubly-fed induction generators (DFIGs) under voltage asymmetry drops. This method uses minimizing active and reactive power fluctuations as the control objective and the total harmonic distortion rate of the stator current as a constraint to establish an optimization model for the rotor command current adjustment coefficient. A multi-objective fuzzy optimization algorithm is then used to solve the model. This reduces the fluctuations in active and reactive power of the wind turbine and limits the total harmonic distortion rate of the stator current to a safe range, achieving multi-objective optimization control of DFIGs under grid voltage imbalance drops. However, this method does not improve the voltage quality of the DFIG or adaptively increase the output range and rate of change of a certain type of power of the DFIG. Summary of the Invention

[0006] To address the shortcomings in existing research on offshore wind turbine control methods, this invention proposes a multi-objective cooperative control method for offshore doubly-fed induction generators (DFIGs) that considers variable coupling. The method employs model predictive control technology to perform multi-objective cooperative optimization control of the DFIGs.

[0007] This invention discloses a multi-objective cooperative control method for offshore doubly-fed induction generators considering variable coupling, the steps of which include:

[0008] Step 1: Establish the prediction model of the model predictive controller, including the prediction model of the generator active and reactive power amplitude and rate of change, the prediction model of the grid-side converter active and reactive power amplitude and rate of change, and the prediction model of the sudden change in the output voltage amplitude of the grid-side converter.

[0009] Step 2: Establish the optimization objective function and constraints of the model predictive controller. The model predictive controller needs to control multiple power sources, including generator active power, generator reactive power, grid-side converter active power, and grid-side converter reactive power. The multiple optimization objectives that the model predictive controller needs to coordinate include: a) improving the accuracy of the servo tracking of the reference values ​​of various power sources of the doubly-fed induction generator (DFIG); b) improving the allowable amplitude range of a specific power source among the various power sources of the DFIG; c) improving the allowable rate of change range of a specific power source among the various power sources of the DFIG; d) reducing the sudden change in the output voltage amplitude of the grid-side converter.

[0010] Step 3: The model predictive controller solves the optimization objective function and constraints in Step 2 at fixed time periods to obtain the optimal control quantity and execute it, thereby realizing multi-objective collaborative optimization control of various power types of doubly-fed wind turbine units.

[0011] In step 1, the prediction model refers to the relationship between the controlled variables (generator active and reactive power amplitude and rate of change, grid-side converter active and reactive power amplitude and rate of change, and grid-side converter output voltage amplitude abrupt change) and the control variables output by the model prediction controller.

[0012] In step 2, when establishing the optimization objective function and constraints for the model predictive controller, to improve the accuracy of the various power servo tracking reference values ​​of the doubly-fed induction generator (DFIG) wind turbine, the first optimization objective function module is set to minimize the deviation between the predicted values ​​of the various power values ​​of the DFIG wind turbine and their respective reference values, as follows:

[0013]

[0014] In the above formula, N is the prediction step size of the model predictive controller, T is the fixed time period of the model predictive controller, j = (k+1)T,...,,(k+N)T represents the starting time of different time periods, P(j) is the predicted value of generator active power at time j, Q(j) is the predicted value of generator reactive power at time j, p(j) is the predicted value of grid-side converter active power at time j, and q(j) is the predicted value of grid-side converter reactive power at time j. ref Let Q(j) be the reference value of the generator's active power at time j. ref Let p(j) be the reference value for the generator's reactive power at time j. ref Let q(j) be the reference value of the active power of the grid-side converter at time j. ref r is the reference value for reactive power of the grid-side converter at time j. P For generator active power deviation coefficient, r Q For generator reactive power deviation coefficient, r p For the active power deviation coefficient of the grid-side converter, r q This represents the reactive power deviation coefficient of the grid-side converter.

[0015] In step 2, when establishing the optimization objective function and constraints for the model predictive controller, to improve the allowable amplitude range of a specific power among the various power outputs of the doubly-fed induction generator (DFIG), the following amplitude coordination constraints for the various power outputs of the DFIG are first set to ensure that the total power of the DFIG does not exceed the rated total power S in each time period. N Secondly, the generator active power deviation coefficient r in the optimization objective function module J1 is set. P Generator reactive power deviation coefficient r Q Active power deviation coefficient r of grid-side converter p The reactive power deviation coefficient r of the grid-side converter q If the deviation coefficient of a specific power is increased, the amplitude of that specific power can be increased within the allowable range while ensuring that the total power does not exceed the rated value.

[0016]

[0017] In the above formula, S(j) is the predicted value of the total power of the doubly fed wind turbine at time j.

[0018] In step 2, when establishing the optimization objective function and constraints for the model predictive controller, to improve the allowable range of the rate of change of a specific power among the various power outputs of the doubly-fed induction generator (DFIG) wind turbine, the following coordination constraints on the rate of change of various power outputs of the DFIG wind turbine are first set to ensure that the rate of change of the total power output of the DFIG wind turbine in each time period does not exceed the rated value G of the total power rate of change. N Secondly, the generator active power deviation coefficient r in the optimization objective function module J1 is set. P Generator reactive power deviation coefficient r Q Active power deviation coefficient r of grid-side converter p The reactive power deviation coefficient r of the grid-side converter q At that time, increasing the deviation coefficient of a specific power can increase the allowable range of the rate of change of that specific power.

[0019]

[0020] In the above formula, dS(j) / dt is the predicted value of the rate of change of the total power of the doubly fed wind turbine at time j.

[0021] The model predictive controller (MMC) controls various power outputs by controlling the converter voltage; therefore, the converter voltage changes as the power output is altered. In step 2, when establishing the objective function and constraints for the MMC, to reduce the sudden changes in the output voltage amplitude of the grid-side converter during various power control processes, the second objective function module is set to minimize these sudden changes.

[0022]

[0023] In the above formula, Δu(j) is the predicted value of the sudden change in the output voltage amplitude of the grid-side converter at time j.

[0024] In step 2, when establishing the optimization objective function and constraints of the model predictive controller, the optimization objective function of the model predictive controller is a weighted combination of optimization objective function module one J1 and optimization objective function module two J2, as shown below, where r1 is the weight coefficient of optimization objective function module one and r2 is the weight coefficient of optimization objective function module two.

[0025] min J=r1*J1+r2*J2 (5)

[0026] Compared with the closest existing technology, the multi-objective cooperative control method for offshore doubly-fed induction generator (DFIG) wind turbines proposed in this invention considers the coupling between various power controls of the offshore DFIG wind turbine (including generator active power, generator reactive power, grid-side converter active power, and grid-side converter reactive power). Multiple cooperative optimization objectives (including improving the accuracy of the DFIG wind turbine's various power servo tracking of their respective reference values, improving the allowable amplitude range of a specific power among the various power types of the DFIG wind turbine, improving the allowable rate of change range of a specific power among the various power types of the DFIG wind turbine, and reducing the abrupt change in the output voltage amplitude of the grid-side converter) control various power types. Therefore, the technical solution provided by this invention has the following beneficial effects:

[0027] 1) The multi-objective cooperative control method for offshore doubly-fed wind turbines that considers variable coupling proposed in this invention can enable offshore doubly-fed wind turbines to adaptively suppress the range of active power output and the range of rate of change in scenarios such as low voltage ride-through where reactive power regulation is required. Under the premise of ensuring that the apparent power amplitude and rate of change do not exceed the upper limit, the reactive power output range and the range of rate of change are improved, so that offshore doubly-fed wind turbines can perform reactive power regulation better.

[0028] 2) The multi-objective cooperative control method for offshore doubly-fed wind turbines that considers variable coupling proposed in this invention can adaptively suppress the range and rate of change of reactive power output in scenarios where active power frequency regulation is required. Under the premise of ensuring that the apparent power amplitude and rate of change do not exceed the upper limit, the active power output range and rate of change range are improved, so that the offshore doubly-fed wind turbines can perform active power frequency regulation better.

[0029] 3) The multi-objective cooperative control method for offshore doubly-fed wind turbines that considers variable coupling proposed in this invention can limit the sudden change in voltage amplitude of the grid-side converter when the offshore doubly-fed wind turbine is adjusting its power, thereby improving the power quality of the offshore doubly-fed wind turbine. Attached Figure Description

[0030] Figure 1 It is the overall architecture of an offshore doubly fed wind turbine generator.

[0031] Figure 2 This is a flowchart illustrating the steps of the multi-objective cooperative control method for offshore doubly-fed wind turbines that considers variable coupling in this invention. Detailed Implementation

[0032] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0033] The overall architecture of the offshore doubly fed wind turbine generator of the present invention is as follows: Figure 1 As shown, an offshore doubly-fed induction generator (DFIG) wind turbine includes a wind turbine, a generator, a rotor-side converter, a grid-side converter, a model predictive controller (MMC), and two converter controllers. Both the generator and the grid-side converter are connected to the power grid and can output active and reactive power. The generator's active and reactive power are controlled by the rotor-side converter, and the grid-side converter's active and reactive power are controlled by the grid-side converter. The MMC calculates the optimal control variables (i.e., the rotor-side converter voltage command and the grid-side converter voltage command) based on reference and actual values ​​of various power parameters (including generator active power, generator reactive power, grid-side converter active power, and grid-side converter reactive power) and sends them to the rotor-side converter controller and the grid-side converter controller. Upon receiving the voltage control commands, the rotor-side converter controller and the grid-side converter controller perform voltage modulation, changing the output voltage of the rotor-side converter and the grid-side converter, thereby changing the generator's active and reactive power output and the grid-side converter's active and reactive power output.

[0034] like Figure 2 As shown, this invention discloses a multi-objective cooperative control method for offshore doubly-fed induction generators considering variable coupling, the steps of which include:

[0035] Step 1: Establish the prediction model of the model predictive controller, including the prediction model of the generator active and reactive power amplitude and rate of change, the prediction model of the grid-side converter active and reactive power amplitude and rate of change, and the prediction model of the sudden change in the output voltage amplitude of the grid-side converter.

[0036] Step 2: Establish the optimization objective function and constraints of the model predictive controller. The model predictive controller needs to control multiple power sources, including generator active power, generator reactive power, grid-side converter active power, and grid-side converter reactive power. The multiple optimization objectives that the model predictive controller needs to coordinate include: a) improving the accuracy of the servo tracking of the reference values ​​of various power sources of the doubly-fed induction generator (DFIG); b) improving the allowable amplitude range of a specific power source among the various power sources of the DFIG; c) improving the allowable rate of change range of a specific power source among the various power sources of the DFIG; d) reducing the sudden change in the output voltage amplitude of the grid-side converter.

[0037] Step 3: The model predictive controller solves the optimization objective function and constraints in Step 2 at fixed time periods to obtain the optimal control quantity and execute it, thereby realizing multi-objective collaborative optimization control of various power types of doubly-fed wind turbine units.

[0038] Step 1, which involves establishing the prediction model for the model prediction controller, includes the following steps:

[0039] Step 1.1: Establish a prediction model for the active and reactive power amplitudes and rates of change of the generator, as derived below;

[0040] Step 1.1.1: Establish the mathematical model of the generator in the dq rotating coordinate system;

[0041]

[0042] Among them, i dr Let i be the d-axis rotor winding current. qr U is the q-axis rotor winding current. dr U is the voltage of the d-axis rotor winding. qr ω is the voltage of the q-axis rotor winding. sl Let r2 be the difference between the electric angular velocity in the dq rotating coordinate system and the rotor electric angular velocity, and let r2 be the rotor winding resistance. s For the self-inductance of the stator winding, L r L is the self-inductance of the rotor winding. m For the mutual inductance of the stator and rotor windings, σ = (L s L r -L 2 m ) / L s For L s L r and L m The constants are P, Q, u1, and ψ1, where P is the generator active power, Q is the generator reactive power, u1 is the magnitude of the stator voltage vector, and ψ1 is the magnitude of the stator flux linkage vector.

[0043] Step 1.1.2: Design the generator inverse system to linearize the above generator mathematical model. The compensation law of the generator inverse system can be expressed as follows;

[0044]

[0045] Among them, v dr V is the voltage control quantity for the d-axis rotor winding. qr The voltage control quantity for the q-axis rotor winding is obtained by the model predictive controller and sent to the rotor-side converter controller.

[0046] Step 1.1.3: Substitute the generator inverter system model into the generator mathematical model to obtain the linearized generator mathematical model as follows;

[0047]

[0048] Step 1.1.4: Discretize the linearized generator mathematical model to obtain the prediction model of the generator's active and reactive power amplitudes and rates of change as follows;

[0049]

[0050] Among them, i dr (j) represents the predicted value of the d-axis rotor winding current at time j, i qr (j) represents the predicted value of the q-axis rotor winding current at time j, v dr (j) represents the value of the d-axis rotor winding voltage control quantity at time j, v qr (j) represents the value of the voltage control quantity of the q-axis rotor winding at time j, P(j) represents the predicted value of the generator active power at time j, Q(j) represents the predicted value of the generator reactive power at time j, dP(j) / dt represents the predicted value of the generator active power change rate at time j, and dQ(j) / dt represents the predicted value of the generator reactive power change rate at time j.

[0051] Step 1.2: Establish a prediction model for the active and reactive power amplitudes and rates of change of the grid-side converter, and derive it as follows;

[0052] Step 1.2.1: Establish the mathematical model of the grid-side converter in a two-phase rotating coordinate system;

[0053]

[0054] Among them, i dg For the d-axis grid-side converter current, i qg For the q-axis grid-side converter current, u dg For the d-axis grid-side converter voltage, u qg R is the voltage of the grid-side converter on the q-axis, L is the resistance of the filter of the grid-side converter, L is the inductance of the filter of the grid-side converter, ω is the synchronous electric angular velocity of the grid, e is the magnitude of the resultant vector of the grid voltage, p is the active power of the grid-side converter, and q is the reactive power of the grid-side converter.

[0055] Step 1.2.2: Linearize the above generator mathematical model by designing the grid-side converter inverse system. The compensation law of the grid-side converter inverse system can be expressed as follows;

[0056]

[0057] Among them, v dg For the voltage control quantity of the d-axis grid-side converter, v qg The voltage control quantity for the q-axis grid-side converter is obtained by the model predictive controller and sent to the grid-side converter controller.

[0058] Step 1.2.3: Substitute the grid-side converter inverter system model into the grid-side converter mathematical model to obtain the linearized grid-side converter mathematical model as follows;

[0059]

[0060] Step 1.2.4: Discretize the linearized mathematical model of the grid-side converter to obtain the prediction model of the active and reactive power amplitudes and rates of change of the grid-side converter as follows;

[0061]

[0062] Among them, i dg (j) represents the predicted value of the converter current on the d-axis grid side at time j, i qg (j) represents the predicted value of the converter current on the q-axis network side at time j, v dg The value of the voltage control quantity of the converter on the grid side of axis (j) at time j, v qg (j) represents the voltage control value of the q-axis grid-side converter at time j, p(j) represents the predicted value of the active power of the grid-side converter at time j, q(j) represents the predicted value of the reactive power of the grid-side converter at time j, dp(j) / dt represents the predicted value of the active power rate of the grid-side converter at time j, and dq(j) / dt represents the predicted value of the reactive power rate of the grid-side converter at time j.

[0063] Step 1.3: Establish a prediction model for the sudden change in the output voltage amplitude of the grid-side converter, and derive it as follows;

[0064] The Model Predictive Controller (MMC) algorithm is not executed continuously, but rather once within each fixed time period T. At the beginning of each time period (j = kT, ..., (k+N-1)T), the MMC solves for the objective function and constraints, and outputs the voltage control quantity (v) for that time period. dg (j) and v qg (j)). Then, the model predicts that the voltage control quantity output by the controller will remain constant until the next time cycle. Grid-side converter voltage (u dg and u qg) and grid-side converter voltage control quantity (V dg and v qg The relationship is shown in formula (16). Therefore, at the beginning of each time period, the model predicts that the controller will output a new round of grid-side converter voltage control quantity, and the grid-side converter voltage will also change abruptly accordingly. The prediction model for the magnitude change of the grid-side converter voltage can be expressed as follows:

[0065]

[0066] Where Δu(j) is the predicted value of the voltage amplitude change of the grid-side converter at time j, u dg (j) represents the predicted value of the d-axis grid-side converter voltage before it abruptly changes in the j-th time period, u qg (j) represents the predicted value of the q-axis grid-side converter voltage before it abruptly changes in the j-th time period, u dg (j+δ) is the predicted value of the d-axis grid-side converter voltage after a sudden change in the j-th time period, u qg (j+δ) is the predicted value of the converter voltage on the q-axis network side after a sudden change in the j-th time period, and δ represents the time spent by the model predictive controller to solve the optimization algorithm in each time period.

[0067] Step 2, which involves establishing the optimization objective function and constraints for the model predictive controller, includes the following steps:

[0068] Step 2.1: To improve the accuracy of the various power servo tracking reference values ​​of the doubly-fed wind turbine, the first optimization objective function module is set to minimize the deviation between the predicted values ​​of various power values ​​of the doubly-fed wind turbine and their respective reference values, as shown below;

[0069]

[0070] In the above formula, N is the prediction step size of the model predictive controller, T is the fixed time period of the model predictive controller, j = (k+1)T,...,,(k+N)T represents the starting time of different time periods, P(j) is the predicted value of generator active power at time j, Q(j) is the predicted value of generator reactive power at time j, p(j) is the predicted value of grid-side converter active power at time j, and q(j) is the predicted value of grid-side converter reactive power at time j. ref Let Q(j) be the reference value of the generator's active power at time j. ref Let p(j) be the reference value for the generator's reactive power at time j. ref Let q(j) be the reference value of the active power of the grid-side converter at time j. ref r is the reference value for reactive power of the grid-side converter at time j. P For generator active power deviation coefficient, r QFor generator reactive power deviation coefficient, r p For the active power deviation coefficient of the grid-side converter, r q This is the reactive power deviation coefficient of the grid-side converter.

[0071] Step 2.2: To improve the allowable amplitude range of a specific power among the various power outputs of the doubly-fed induction generator (DFIG), the following amplitude coordination constraints for the various power outputs of the DFIG are first set to ensure that the total power of the DFIG does not exceed the rated total power S in each time period. N Secondly, the generator active power deviation coefficient r in the optimization objective function module J1 is set. P Generator reactive power deviation coefficient r Q Active power deviation coefficient r of grid-side converter p The reactive power deviation coefficient r of the grid-side converter q At the same time, increasing the deviation coefficient of a specific power can increase the allowable range of the amplitude of that specific power while ensuring that the total power does not exceed the rated value.

[0072]

[0073] In the above formula, S(j) is the predicted value of the total power of the doubly fed wind turbine at time j.

[0074] Step 2.3: To improve the allowable range of the rate of change of a specific power among the various power outputs of the doubly-fed induction generator (DFIG) wind turbine, the following coordination constraint on the rate of change of various power outputs of the DFIG wind turbine is first set to ensure that the rate of change of the total power output of the DFIG wind turbine does not exceed the rated value G of the total power rate of change in each time period. N Secondly, the generator active power deviation coefficient r in the optimization objective function module J1 is set. P Generator reactive power deviation coefficient r Q Active power deviation coefficient r of grid-side converter p The reactive power deviation coefficient r of the grid-side converter q At that time, increasing the deviation coefficient of a specific power can increase the allowable range of the rate of change of that specific power.

[0075]

[0076] In the above formula, dS(j) / dt is the predicted value of the rate of change of the total power of the doubly fed wind turbine at time j.

[0077] Step 2.4: To reduce the sudden change in the output voltage amplitude of the grid-side converter during various power control processes, the second optimization objective function module is set to minimize the sudden change in the output voltage amplitude of the grid-side converter.

[0078]

[0079] In the above formula, Δu(j) is the predicted value of the sudden change in the output voltage amplitude of the grid-side converter at time j.

[0080] Step 2.5: The objective function of the model predictive controller is a weighted combination of objective function module one J1 and objective function module two J2, as shown below, where r1 is the weight coefficient of objective function module one and r2 is the weight coefficient of objective function module two.

[0081] min J=r1*J1+r2*J2 (5)

[0082] The embodiments described above are merely illustrative of the technical solutions of the present invention and should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A multi-objective cooperative control method for offshore doubly-fed induction generators considering variable coupling, characterized in that, The method describes a multi-objective coordinated control of the generator active power, generator reactive power, grid-side converter active power, and grid-side converter reactive power of an offshore doubly-fed induction generator (DFIG) by constructing a model predictive controller. The method includes the following steps: Step 1: Establish the prediction model of the model predictive controller, including the prediction model of the generator active and reactive power amplitude and rate of change, the prediction model of the grid-side converter active and reactive power amplitude and rate of change, and the prediction model of the sudden change in the output voltage amplitude of the grid-side converter. Step 2: Establish the optimization objective function and constraints of the model predictive controller. The model predictive controller needs to control multiple power sources, including generator active power, generator reactive power, grid-side converter active power, and grid-side converter reactive power. The multiple optimization objectives that the model predictive controller needs to coordinate include: a) improving the accuracy of the doubly-fed induction generator's various power servo tracking of their respective reference values. b) Improve the allowable amplitude range of a specific power among various power outputs in doubly-fed wind turbine generators. c) Improve the allowable range of the rate of change of a specific power among various power outputs in doubly-fed wind turbine units. d) Reduce the magnitude of sudden changes in the output voltage of the grid-side converter; Step 3: The model predictive controller solves the optimization objective function and constraints in Step 2 at fixed time periods to obtain the optimal control quantity and execute it, thereby realizing multi-objective collaborative optimization control of various power types of the doubly fed wind turbine. In step 2, when establishing the optimization objective function and constraints of the model predictive controller, the optimization objective function of the model predictive controller is optimization objective function module one. J1 With the optimization objective function module two J2 Weighted combination; （5） in, r1 To optimize the weight coefficients of module one of the objective function, r2 To optimize the weight coefficients of module two of the objective function; In step 2, when establishing the optimization objective function and constraints for the model predictive controller, to improve the accuracy of various power servo tracking reference values ​​of the doubly-fed induction generator (DFIG) wind turbine, an optimization objective function module 1 is set. J1 To minimize the deviation between the predicted values ​​of various power outputs of the doubly-fed wind turbine and their respective reference values, the following measures are taken: （1） in, N The prediction step size for the model predictive controller. T The model predicts the controller with a fixed time period. Representing the starting point of different time periods. The generator is of great benefit j Predicted value at time, For generator reactive power j Predicted value at time, For grid-side converter active power j Predicted value at time, For grid-side converter reactive power j The predicted value at any given time; The generator is of great benefit j Reference value of time, For generator reactive power j Reference value at any time, For grid-side converter active power j Reference value of time, For grid-side converter reactive power j Reference value at any given time; For generator active power deviation coefficient, For generator reactive power deviation coefficient, For the active power deviation coefficient of the grid-side converter, This is the reactive power deviation coefficient of the grid-side converter.

2. The multi-objective cooperative control method for offshore doubly-fed induction generators considering variable coupling as described in claim 1, characterized in that, In step 2, when establishing the optimization objective function and constraints for the model predictive controller, to improve the allowable amplitude range of a specific power among the various power outputs of the doubly-fed induction generator (DFIG), the following amplitude coordination constraints for the various power outputs of the DFIG are first set to ensure that the total power of the DFIG does not exceed the rated total power value in each time period. Secondly, set up the optimization objective function module one. J1 Generator active power deviation coefficient Generator reactive power deviation coefficient Active power deviation coefficient of grid-side converter Reactive power deviation coefficient of grid-side converter When the deviation coefficient of a specific power is increased, the amplitude of that specific power can be increased within the allowable range, provided that the total power does not exceed the rated value. （2） in, For the total power of the doubly fed wind turbine unit at j The predicted value at any given time.

3. The multi-objective cooperative control method for offshore doubly-fed induction generators considering variable coupling as described in claim 1, characterized in that, In step 2, when establishing the optimization objective function and constraints for the model predictive controller, to improve the allowable range of the rate of change of a specific power among the various power outputs of the doubly-fed induction generator (DFIG) wind turbine, the following coordination constraints on the rate of change of various power outputs of the DFIG wind turbine are first set to ensure that the rate of change of the total power output of the DFIG wind turbine does not exceed the rated value of the rate of change of total power output in each time period. Secondly, set up the optimization objective function module one. J1 Generator active power deviation coefficient Generator reactive power deviation coefficient Active power deviation coefficient of grid-side converter Reactive power deviation coefficient of grid-side converter At the same time, increasing the deviation coefficient of a specific power can increase the allowable range of the rate of change of that specific power; （3） in, The rate of change of the total power of the doubly fed wind turbine unit j The predicted value at any given time.

4. The multi-objective cooperative control method for offshore doubly-fed induction generators considering variable coupling as described in claim 1, characterized in that, In step 2, when establishing the optimization objective function and constraints for the model predictive controller, in order to reduce the sudden changes in the output voltage amplitude of the grid-side converter during various power control processes, optimization objective function module two is set. J2 To minimize the amplitude abrupt change in the output voltage of the grid-side converter; （4） in, For the sudden change in the output voltage amplitude of the grid-side converter j The predicted value at any given time.

Images