Multi-objective optimization method for control parameters of grid-side converter of medium-frequency grid-connected wind turbine.

Abstract

This invention relates to the field of grid-connected technology for medium-frequency grid-connected wind turbines, and particularly to a multi-objective optimization method for the control parameters of the grid-side converter of a medium-frequency grid-connected wind turbine. The method includes: determining the grid-connected model of the grid-side converter of the medium-frequency grid-connected wind turbine, and calculating the range of virtual inertia and virtual damping; establishing a multi-objective optimization function with the optimization objectives of minimizing the overshoot of the power response of the grid-side converter, minimizing the maximum frequency fluctuation of the frequency response, and maximizing the damping ratio of the dominant oscillation mode; constructing a multi-objective optimization model using the parameter combination of virtual inertia and virtual damping as population parameters and the multi-objective optimization function as the applicability function; and solving the multi-objective optimization model using a particle swarm optimization algorithm to obtain the optimal solution for the parameter combination of virtual inertia and virtual damping. This method can improve the control parameters of the grid-side converter, thereby achieving better stability, and can also serve as an initial value selection method for adaptive control.

Description

Technical Field

[0001] This invention relates to the field of grid connection technology for medium-frequency grid-connected wind turbines, and in particular to a multi-objective optimization method for the control parameters of the grid-side converter of medium-frequency grid-connected wind turbines. Background Technology

[0002] The development of nearshore wind power is nearing saturation, and offshore wind technology is evolving towards "deep-sea" technology. Deep-sea wind technology is characterized by long distances from shore, large transmission capacity, and high construction costs. Traditional industrial frequency AC collection and transmission technology is no longer suitable for deep-sea scenarios due to the skin effect of cables. Furthermore, with the development of new power electronic switching technology, flexible DC transmission technology has become the inevitable choice for long-distance power transmission. Since flexible DC transmission requires an AC-DC-AC conversion, the collection head of the offshore wind farm needs to be converted to DC transmission via a converter platform. Therefore, the internal frequency of the wind farm does not need to be maintained at 50Hz. Considering factors such as the cost of the converter platform and the cost of the step-up collection transformer, the frequency can be increased to the mid-frequency range; this technology is called mid-frequency collection technology. If the operating frequency of the offshore wind farm system is increased to the mid-frequency range, according to the principle of electromagnetic induction, increasing the operating frequency can reduce the size and weight of transformers and reactors, and also enable the miniaturization and lightweighting of the offshore converter station platform, thereby achieving low-cost DC transmission of deep-sea wind power.

[0003] Mid-frequency convergence technology has solved the problem of difficult transmission of deep-sea wind power due to long distances from the shore, but it still faces the problem of easy oscillation in weak grid scenarios of traditional grid-connected offshore wind turbines. Therefore, the grid connection of new mid-frequency grid-connected offshore wind turbines can improve the grid connection stability of wind farms in weak grid scenarios. At the same time, it can also provide more construction solutions for offshore converter platforms. New uncontrolled rectifier converter equipment can be used to replace traditional MMC converter platforms, further reducing construction investment costs.

[0004] When the system frequency is increased to the medium frequency, the control parameters of the grid-connected wind turbine in the medium frequency scenario need to be further optimized, especially the virtual inertia and virtual damping parameters of the power outer loop of the virtual synchronous generator technology. These parameters essentially determine the oscillation characteristics, response speed and steady-state deviation of the grid-connected offshore wind turbine when connected to the grid.

[0005] Because different intermediate frequency control parameters have different small-signal stability, the common method for designing intermediate frequency power outer loop parameters is to first calculate the range of control parameters using formulas, and then manually optimize them within this range. However, for intermediate frequency network control with a three-loop structure containing a power outer loop and a voltage and current double closed-loop inner loop, manual adjustment is not only time-consuming, but also makes it difficult to achieve precise optimization of control parameters. Summary of the Invention

[0006] This invention provides a multi-objective optimization method for the control parameters of the grid-side converter of a medium-frequency grid-connected wind turbine, which can effectively solve the problems in the background art.

[0007] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A multi-objective optimization method for control parameters of grid-side converters in medium-frequency grid-connected wind turbines includes the following steps: Determine the grid-connected model of the grid-side converter of the medium-frequency grid-type wind turbine, and calculate the range of virtual inertia and virtual damping; A multi-objective optimization function is established with the optimization objectives of minimizing the overshoot of the power response of the grid-side converter of the medium-frequency grid-type wind turbine, minimizing the maximum frequency fluctuation of the frequency response, and maximizing the damping ratio of the dominant oscillation mode. Using the parameter combination of the virtual inertia and virtual damping as population parameters and the multi-objective optimization function as the applicability function, a multi-objective optimization model is constructed. The multi-objective optimization model is solved using the particle swarm optimization algorithm to obtain the optimal solution for the combination of virtual inertia and virtual damping parameters.

[0008] Furthermore, the determination of the grid-side converter grid-connected model for medium-frequency grid-type wind turbines includes an LC-type filter circuit and the equivalent impedance of the power grid transmission line; The LC-type filter circuit is based on a mathematical model of state-space equations: In the formula: For filtering inductors and resistors; , These are the filter inductor and filter capacitor; The d-axis component of the output current of the grid-side converter of the medium-frequency grid-type wind turbine unit; This refers to the q-axis component of the output current of the grid-side converter of a medium-frequency grid-connected wind turbine. This represents the d-axis component of the voltage across the filter capacitor. This represents the q-axis component of the voltage across the filter capacitor. The actual angular velocity of the grid-side converter of the medium-frequency grid-type wind turbine unit; The d-axis component of the current after filtering by the grid-side converter of the medium-frequency grid-type wind turbine; The q-axis component of the current after filtering by the grid-side converter of the medium-frequency grid-connected wind turbine unit; The mathematical model of the equivalent impedance of the power grid transmission line based on the state-space equation is as follows: In the formula: The equivalent resistance of the power grid transmission line; The equivalent inductance of the power grid transmission line; The d-axis component of the intermediate frequency equivalent grid-connected voltage source; This is the q-axis component of the medium-frequency equivalent grid-connected voltage source.

[0009] Furthermore, the grid-side converter adopts dual closed-loop control of power outer loop and voltage and current; The power outer loop is controlled by a virtual synchronous generator. The active and reactive power outputs of the grid-side converter of the medium-frequency grid-connected wind turbine are calculated using the inductor current and capacitor voltage of the AC-side filter, as shown in the following formula: In the formula, The active power output of the grid-side converter for medium-frequency grid-type wind turbine units; The reactive power output of the grid-side converter for medium-frequency grid-type wind turbines; The active power outer loop uses virtual synchronous generator control, and the reactive power outer loop uses droop control. The mathematical model based on the state-space equations is as follows: In the formula, This is virtual inertia; For virtual damping; The actual angular velocity of the grid-side converter of the medium-frequency grid-type wind turbine unit; The reference angular velocity for the grid-side converter of a medium-frequency grid-type wind turbine; The reference active power output of the grid-side converter for medium-frequency grid-type wind turbine units; The reference reactive power output of the grid-side converter for medium-frequency grid-type wind turbine units; This is the active power droop factor; This refers to the reactive power droop factor. Generate d-axis reference voltage for the grid-side converter of medium-frequency grid-connected wind turbine units; This is the reference voltage for the grid-side converter of a medium-frequency grid-connected wind turbine. The power inner loop of the grid-side converter of the medium-frequency grid-type wind turbine adopts a dual closed-loop design for voltage and current. The mathematical expression for the inner voltage loop is: In the formula: This is an intermediate variable for the d-axis PI integrator in the inner voltage loop; This is an intermediate variable for the q-axis PI integrator in the voltage inner loop; The proportional gain of the voltage inner-loop PI controller; The integral coefficient of the voltage inner-loop PI controller; Generate d-axis reference voltage for the grid-side converter of medium-frequency grid-connected wind turbine units; Generate q-axis reference voltage for the grid-side converter of medium-frequency grid-connected wind turbine units; This is the d-axis reference current value generated by the voltage inner loop; This is the q-axis reference current value generated by the voltage inner loop; The mathematical expression for the inner current loop is: In the formula, For the intermediate variable of the d-axis PI integrator in the inner current loop; This is an intermediate variable for the q-axis PI integrator in the inner current loop; The proportional gain of the inner-loop PI controller; The integral coefficient of the current inner loop PI controller.

[0010] Furthermore, after determining the grid-connected model of the grid-side converter of the medium-frequency grid-type wind turbine, the process also includes constructing a full-order small-signal linearization model: ; In the formula, x represents the system's state variables, which have 12 orders: In the formula, Let be the system angular velocity. , The dq coordinate system components of the grid-side output current of the medium-frequency grid-type converter; , For the dq coordinate system components of the line inductance current; u d u q The dq coordinate system components of the grid-side output voltage of the medium-frequency grid-type converter; , The dq coordinate system components of the filter capacitor voltage in a medium-frequency grid-type converter; , This is an intermediate variable for the voltage inner-loop PI integrator; , This is an intermediate variable for the inner loop PI integrator of the current; For the angle of attack.

[0011] Furthermore, the establishment of the multi-objective optimization function includes: Constructing the power response overshoot of the grid-side converter for medium-frequency grid-connected wind turbine units The objective function, with overshoot The overshoot is minimized as the objective. The calculation method is as follows: In the formula: This refers to the power response overshoot of the grid-side converter in a medium-frequency grid-connected wind turbine. This represents the maximum power response of the grid-side converter in a medium-frequency grid-connected wind turbine unit. This represents the steady-state power response of the grid-side converter in a medium-frequency grid-connected wind turbine generator set. Maximum frequency fluctuation in the frequency response of the grid-side converter for medium-frequency grid-connected wind turbines The objective function is to maximize the frequency fluctuation. Minimize the maximum frequency fluctuation amount The calculation method is as follows: In the formula: t1 represents the maximum frequency fluctuation; t2 represents the start time of the disturbance; and t2 represents the end time of the recovery to stability. Constructing the dominant mode damping ratio of oscillation in the grid-side converter of a medium-frequency grid-connected wind turbine The objective function is to maximize the damping ratio, where the damping ratio of the dominant oscillation mode is... The calculation method is as follows: In the formula: the subscript i indicates that the i-th eigenvalue is the dominant mode; Damping ratio for the dominant mode; The real part of the dominant characteristic root; The imaginary part of the dominant characteristic root; Overshoot of grid-side converter power response in medium-frequency grid-connected wind turbine units Maximum frequency fluctuation in frequency response oscillation-dominant mode damping ratio Based on the objective function, the three objective functions are normalized, and then the total objective function F is obtained by combining the weighted coefficients. obj (J,D), calculated as follows: In the formula: , , The weighting coefficients are, in order, the overshoot, the maximum frequency deviation, and the dominant mode damping ratio; The normalized overshoot value; The maximum frequency fluctuation after normalization value; The normalized damping ratio of the dominant oscillation mode value; The normalization calculation method is as follows: In the formula, This is the maximum permissible overshoot. This is the maximum permissible frequency fluctuation. This is the minimum permissible dominant mode damping ratio.

[0012] Furthermore, the multi-objective optimization function also includes a penalty function, the construction of which includes: In the formula: This is the penalty coefficient; The maximum rate of change of frequency; This refers to the active power adjustment time.

[0013] Furthermore, based on the objective function and the penalty function, the final fitness objective function is constructed, and its mathematical expression is as follows: .

[0014] Furthermore, the step of solving the multi-objective optimization model using the particle swarm optimization algorithm includes the following steps: Initialize the particle swarm's number, position, and velocity; The parameter combination of the virtual inertia J and virtual damping D is substituted into the multi-objective optimization model as population parameters, and the multi-objective optimization model is solved to obtain the overshoot of the grid-side converter power response of the medium-frequency grid-type wind turbine. Maximum frequency fluctuation in frequency response oscillation-dominant mode damping ratio Weighted final fitness objective function value ; The obtained objective function value is compared with the historical best value of the objective function. If the obtained objective function value is better than the historical best value of the objective function, the obtained objective function value is stored in the optimal objective function value, the obtained parameter combination is stored in the optimal parameter value, and then the next step is executed. If the obtained objective function value is not better than the historical best value of the objective function, the next step is executed directly. Determine if the maximum number of iterations has been reached. If it has, output the current optimal parameters. If not, update the population position and velocity, repeat the substitution of parameters, and solve for the final fitness objective function value Z.

[0015] Furthermore, the rate of updating the population location includes: The MOPSO algorithm is used to update the population position and velocity. The velocity and position of particles in the MOPSO algorithm are shown in the following equation: In the formula: Inertial weight; For individual learning factors; As a social learning factor; , These are random values ​​generated between 0 and 1; Let t be the velocity of the t-th generation; For the t-th generation position; The optimal particle value for an individual; This represents the optimal value for all particles.

[0016] Furthermore, in the process of updating the population position and velocity using the MOPSO algorithm, a linear weighting method is used to adjust the inertia weights. Optimization is performed, where in the k-th iteration... value for: In the formula: G is the maximum number of iterations; The maximum weight value; This is the minimum weight value.

[0017] This invention also provides a multi-objective optimization device for control parameters of grid-side converters in medium-frequency grid-connected wind turbines, using the multi-objective optimization method for control parameters of grid-side converters in medium-frequency grid-connected wind turbines as described above. The device includes: The range unit is used to determine the grid-connected model of the grid-side converter of the medium-frequency grid-type wind turbine and to calculate the range of virtual inertia and virtual damping. Target unit for overshoot of grid-side converter power response in medium-frequency grid-connected wind turbine units. Minimize the maximum frequency fluctuation of the frequency response Minimize the damping ratio of the dominant oscillation mode. To maximize the optimization objective, a multi-objective optimization function is established. The modeling unit is used to build a multi-objective optimization model using a combination of virtual inertia and virtual damping parameters as population parameters and a multi-objective optimization function as the applicability function. The solution unit is used to solve the multi-objective optimization model using the particle swarm optimization algorithm to obtain the optimal solution for the combination of virtual inertia and virtual damping parameters.

[0018] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements the method as described in any one of claims 1-10.

[0019] The present invention also provides a storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described above.

[0020] The technical solution of this invention can achieve the following technical effects: This invention focuses on the overshoot δ of the grid-side converter power response of a medium-frequency grid-connected wind turbine. P%, the maximum frequency fluctuation of the frequency response Δω max The oscillation-dominant mode damping ratio ζ is used as the optimization objective, and a multi-objective optimization function is established. The combination of virtual inertia and virtual damping parameters is used as the population parameters, and the overall objective function in the multi-objective optimization model is used as the applicability function. The particle swarm optimization algorithm is employed to solve the multi-objective optimization model, thereby obtaining the optimal solution for the combination of virtual inertia and virtual damping parameters of the controller. This allows for more accurate control parameters of the grid-side converter in medium-frequency grid-connected wind turbines, resulting in better stability under different short-circuit ratios. Simultaneously, this invention can also provide an initial value calculation method for adaptive parameter control of grid-side converters in medium-frequency grid-connected wind turbines. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a flowchart of the method in Example 1; Figure 2 This is a schematic diagram of the device in Example 1; Figure 3 This is a diagram of the grid-side converter control strategy in Example 1; Figure 4 The flowchart is shown in Example 2. Figure 5 This is a graph showing the trend of the root locus of the dominant mode as a function of the short-circuit ratio before optimization in Example 2. Figure 6 This is a graph showing the trend of the root locus of the dominant mode with short-circuit ratio in Example 2, where only the damping ratio is optimized. Figure 7 This is a graph showing the trend of the root locus of the dominant mode as a function of the short-circuit ratio in Example 2. Figure 8 The diagram shows the power response of the three schemes under small q-axis voltage disturbances in Example 2. Figure 9 The above are the frequency response diagrams of the three schemes under small q-axis voltage perturbations in Example 2. Figure 10 The diagram shows the power response of the three schemes when the wind speed changes abruptly in Example 2. Figure 11 The frequency response diagrams for the three schemes when the wind speed changes abruptly in Example 2 are shown. Figure 12 This is a schematic diagram of the structure of a computer device. Detailed Implementation

[0023] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0024] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0025] Example 1: like Figure 1 As shown, this application provides a multi-objective optimization method for the control parameters of the grid-side converter of a medium-frequency grid-connected wind turbine, including the following steps: Determine the grid-connected model of the grid-side converter of the medium-frequency grid-type wind turbine, and calculate the range of virtual inertia and virtual damping; Overshoot of grid-side converter power response in medium-frequency grid-connected wind turbine units Minimize the maximum frequency fluctuation of the frequency response Minimize the damping ratio of the dominant oscillation mode. To maximize the optimization objective, a multi-objective optimization function is established. Using the parameter combination of the virtual inertia and virtual damping as population parameters and the multi-objective optimization function as the applicability function, a multi-objective optimization model is constructed. The multi-objective optimization model is solved using the particle swarm optimization algorithm to obtain the optimal solution for the combination of virtual inertia and virtual damping parameters.

[0026] Furthermore, the grid-connection model for the grid-side converter of the medium-frequency grid-type wind turbine is determined as follows: like Figure 3 As shown, the grid-side converter of the medium-frequency grid-connected wind turbine adopts an LC-type filter circuit and is connected to the medium-frequency ideal grid through the grid equivalent resistance and inductance; the main circuit topology is based on a mathematical model of state-space equations, including an LC-type filter circuit and the equivalent impedance of the grid transmission line; The LC-type filter circuit is based on a mathematical model of state-space equations: In the formula: R f For filter inductors and resistors; L f C f For filter inductors and filter capacitors; i d i represents the d-axis component of the output current of the grid-side converter of a medium-frequency grid-connected wind turbine.q The q-axis component of the output current of the grid-side converter of a medium-frequency grid-connected wind turbine; u od The d-axis component of the voltage across the filter capacitor; u oq ω is the q-axis component of the voltage across the filter capacitor; ω is the actual angular velocity of the grid-side converter of the medium-frequency grid-connected wind turbine; i od i represents the d-axis component of the filtered current from the grid-side converter of a medium-frequency grid-connected wind turbine. oq The q-axis component of the current after filtering by the grid-side converter of the medium-frequency grid-connected wind turbine unit; The mathematical model of the equivalent impedance of the power grid transmission line based on the state-space equation is as follows: In the formula: R g L is the equivalent resistance of the power grid transmission line; g The equivalent inductance of the power grid transmission line; u gd The d-axis component of the intermediate frequency equivalent grid-connected voltage source; u gq This is the q-axis component of the medium-frequency equivalent grid-connected voltage source.

[0027] In this scheme, the grid-side converter adopts dual closed-loop control of power outer loop and voltage and current. The power outer loop adopts virtual synchronous generator control, and the mathematical expression of the second-order motion equation of the rotor of a traditional synchronous generator is as follows: In the formula, J is the virtual inertia; D is the virtual damping; ω n P is the reference angular velocity of the grid-side converter of the medium-frequency grid-connected wind turbine. ref The reference active power output of the grid-side converter for medium-frequency grid-connected wind turbine units; Q ref K is the reference reactive power output of the grid-side converter for medium-frequency grid-connected wind turbine units. d K is the active power droop factor. q The reactive power droop factor; u VSG_d Generate d-axis reference voltage for the grid-side converter of the medium-frequency grid-connected wind turbine; u ref P is the reference voltage for the grid-side converter of a medium-frequency grid-connected wind turbine; e For the grid-side converter of medium-frequency grid-connected wind turbine units, the output active power is Q. e It outputs reactive power to the grid-side converter of the medium-frequency grid-type wind turbine.

[0028] The active and reactive power outputs of the grid-side converter of a medium-frequency grid-connected wind turbine are calculated using the inductor current and capacitor voltage of the AC-side filter, as shown in the following formula: In the formula, P e For the grid-side converter of medium-frequency grid-connected wind turbine units, the output active power is Q.e The reactive power output of the grid-side converter for medium-frequency grid-type wind turbines; The phase angle of the grid-connected voltage is controlled by active power, and the amplitude of the grid-connected voltage is controlled by reactive power. The grid-side converter of a medium-frequency grid-connected wind turbine adopts a dual closed-loop power system with both voltage and current inputs. The mathematical expression for the voltage inner loop is as follows: In the formula: x ud x is the intermediate variable of the d-axis PI integrator in the inner voltage loop; uq K is the intermediate variable of the q-axis PI integrator in the inner voltage loop; up K is the proportional gain of the voltage inner-loop PI controller; ui The integral coefficient of the voltage inner-loop PI controller; u VSG_d Generate d-axis reference voltage for the grid-side converter of the medium-frequency grid-connected wind turbine; u VSG_q Generate q-axis reference voltage for the grid-side converter of the medium-frequency grid-connected wind turbine; ud The d-axis reference current value generated by the voltage inner loop; i uq This is the q-axis reference current value generated by the inner voltage loop.

[0029] The mathematical expression for the inner current loop is: In the formula, x id x is an intermediate variable for the d-axis PI integrator of the inner current loop; iq K is the intermediate variable of the q-axis PI integrator in the inner current loop; ip K is the proportional gain of the inner-loop PI controller; ii The integral coefficient of the current inner loop PI controller.

[0030] The specific control strategy of the grid-side converter is based on the outer loop power reference value P. ref With reactive power reference value Q ref The voltage dq reference value u is generated autonomously through a virtual synchronous generator. dref u qref The corresponding control voltage U is then obtained through dual closed-loop PI control of voltage and current. d* U q* And add the cross-coupling voltage compensation term ΔU d and ΔU q This allows us to obtain the final d-axis and q-axis control voltage components U. d and U q Then, the drive signal required by the grid-side converter is obtained through PWM control to realize power transmission.

[0031] Furthermore, after determining the grid-connected model of the grid-side converter of the medium-frequency grid-type wind turbine, the process also includes constructing a full-order small-signal linearization model: In the formula, x represents the system's state variables, which have 12 orders: In the formula, ω is the angular velocity of the system, and i d i q For the grid-side output current dq coordinate system components of the medium-frequency grid-type converter; i od i oq For the dq coordinate system components of the line inductance current; u d u q For the grid-side output voltage components of the medium-frequency grid-type converter in the dq coordinate system; u od u oq For the dq coordinate system components of the filter capacitor voltage in a medium-frequency grid-type converter; x ud x uq x is the intermediate variable of the voltage inner-loop PI integrator; id x iq δ is the intermediate variable of the PI integrator in the inner current loop; δ is the power angle.

[0032] The multi-objective optimization function established in this invention includes: Constructing the power response overshoot of the grid-side converter for medium-frequency grid-connected wind turbine units The objective function, with overshoot The overshoot is minimized as the objective. The calculation method is as follows: In the formula: For the power response overshoot of the grid-side converter of a medium-frequency grid-connected wind turbine, P max P represents the maximum power response of the grid-side converter of a medium-frequency grid-connected wind turbine. ∞ This represents the steady-state power response of the grid-side converter in a medium-frequency grid-connected wind turbine generator set. Maximum frequency fluctuation in the frequency response of the grid-side converter for medium-frequency grid-connected wind turbines The objective function is to maximize the frequency fluctuation. Minimize the maximum frequency fluctuation amount The calculation method is as follows: In the formula: t1 represents the maximum frequency fluctuation; t2 represents the start time of the disturbance; and t2 represents the end time of the recovery to stability. Constructing the dominant mode damping ratio of oscillation in the grid-side converter of a medium-frequency grid-connected wind turbine The objective function is to maximize the damping ratio, where the damping ratio of the dominant oscillation mode is... The calculation method is as follows: In the formula: the subscript i indicates that the i-th eigenvalue is the dominant mode; Damping ratio for the dominant mode; The real part of the dominant characteristic root; The imaginary part of the dominant characteristic root; Overshoot of grid-side converter power response in medium-frequency grid-connected wind turbine units Maximum frequency fluctuation in frequency response oscillation-dominant mode damping ratio Based on the objective function, the three objective functions are normalized, and then the total objective function F is obtained by combining the weighted coefficients. obj (J,D), calculated as follows: In the formula: , , The weighting coefficients are, in order, the overshoot, the maximum frequency deviation, and the dominant mode damping ratio; The normalized overshoot value; The maximum frequency fluctuation after normalization value; The normalized damping ratio of the dominant oscillation mode value; The normalization calculation method is as follows: In the formula, This is the maximum permissible overshoot. This is the maximum permissible frequency fluctuation. This is the minimum permissible dominant mode damping ratio.

[0033] In a preferred embodiment, the multi-objective optimization function further includes a penalty function, which is constructed as follows: Establish the maximum frequency change rate RoCoF max The penalty function, the maximum rate of change of frequency RoCoF max The penalty function is calculated as follows: Where: RoCoF max ω represents the maximum frequency change rate; ω represents the output angular velocity of the grid-side converter in a medium-frequency grid-connected wind turbine; RoCoF lim This represents the maximum permissible rate of frequency change.

[0034] Establish active power regulation time t P The penalty function, the active power adjustment time t P The penalty function is calculated as follows: In the formula: t P For active power regulation time; ε P The allowable active power deviation; t lim This is the maximum allowable adjustment time.

[0035] With the maximum frequency change rate RoCoF max and active power regulation time t P Based on the constraints, construct the penalty function F. pen (J,D), penalty function F pen The calculation method for (J,D) is as follows: In the formula: κ is the penalty coefficient; The maximum rate of change of frequency; t P This refers to the active power adjustment time.

[0036] Furthermore, based on the objective function and the penalty function, the final fitness function is constructed, and its mathematical expression is as follows: .

[0037] A penalty function is introduced to supplement the objective function. The penalty term consists of two constraints: the maximum rate of change of frequency and the active power adjustment time. Finally, the final fitness objective function is formed by fusing the weighted objective function and the penalty function.

[0038] The present invention employs a particle swarm optimization algorithm to solve the multi-objective optimization model, including the following steps: Initialize the particle swarm's number, position, and velocity; The parameter combination of the virtual inertia J and virtual damping D is substituted into the multi-objective optimization model as population parameters, and the multi-objective optimization model is solved to obtain the overshoot δ of the grid-side converter power response of the medium-frequency grid-connected wind turbine. P %, the maximum frequency fluctuation of the frequency response Δω max The objective function value Z after weighting the damping ratio ζ of the dominant oscillation mode; The obtained objective function value is compared with the historical best value of the objective function. If the obtained objective function value is better than the historical best value of the objective function, the obtained objective function value is stored in the optimal objective function value, the obtained parameter combination is stored in the optimal parameter value, and then the next step is executed. If the obtained objective function value is not better than the historical best value of the objective function, the next step is executed directly. Determine if the maximum number of iterations has been reached. If it has, output the current optimal parameters. If not, update the population position and velocity, repeat the substitution of parameters, and solve for the final fitness objective function value Z.

[0039] Furthermore, the rate of updating population location includes: The MOPSO algorithm is used to update the population position and velocity. The velocity and position of particles in the MOPSO algorithm are shown in the following equation: In the formula: ω is the inertia weight; c1 is the individual learning factor; c2 is the social learning factor; r1 and r2 are random values ​​generated between 0 and 1; V i (t) represents the velocity in generation t; X i (t) represents the position of the t-th generation; P best For the individual optimal particle value; G best This represents the optimal value for all particles.

[0040] Furthermore, in updating the population position and velocity using the MOPSO algorithm, a linear weighting method is employed to adjust the inertia weights. Optimization is performed, where in the k-th iteration... value for: In the formula: G is the maximum number of iterations; The maximum weight value; This is the minimum weight value.

[0041] To prevent the MOPSO algorithm from getting trapped in local optima, a linear weighting method is used to optimize the inertia weight ω. By improving the setting of the inertia weight ω, the convergence speed of the MOPSO algorithm is improved.

[0042] like Figure 2 As shown, the present invention also provides a multi-objective optimization device for control parameters of grid-side converters of medium-frequency grid-connected wind turbines, using the above-described multi-objective optimization method for control parameters of grid-side converters of medium-frequency grid-connected wind turbines. The device includes: The range unit is used to determine the grid-connected model of the grid-side converter of the medium-frequency grid-type wind turbine and to calculate the range of virtual inertia and virtual damping. Target unit for overshoot of grid-side converter power response in medium-frequency grid-connected wind turbine units. Minimize the maximum frequency fluctuation of the frequency response Minimize the damping ratio of the dominant oscillation mode. To maximize the optimization objective, a multi-objective optimization function is established. The modeling unit is used to build a multi-objective optimization model using the parameter combination of the virtual inertia and virtual damping of the controller as population parameters and the multi-objective optimization function as the applicability function. The solution unit is used to solve the multi-objective optimization model using the particle swarm optimization algorithm to obtain the optimal solution of the combination of virtual inertia and virtual damping parameters of the controller.

[0043] Example 2: This embodiment includes a multi-objective optimization method for control parameters of grid-side converters in mid-frequency grid-connected wind turbines based on particle swarm optimization. This method uses the converter's power response overshoot and maximum frequency response fluctuation as links, and adjusts virtual inertia and virtual damping parameters in the controller to achieve better stability. The specific steps are as follows: A. First, determine the grid-connected model of the grid-side converter of the medium-frequency grid-type wind turbine, then calculate the range of the virtual inertia J and virtual damping D of the controller. With the overshoot of the power response, the maximum frequency fluctuation of the frequency response, and the damping ratio of the dominant oscillation mode of the medium-frequency grid-type wind turbine as optimization objectives, establish a multi-objective optimization function. B. Using the combination of virtual inertia J and virtual damping D in the control circuit as the population parameters, and the total objective function in the multi-objective optimization model as the applicability function, a multi-objective optimization model is constructed. The particle swarm optimization algorithm is used to solve the multi-objective optimization model, thereby obtaining the optimal solution of the combination of capacitor and inductor parameters.

[0044] The solution steps for the multi-objective optimization model in step A are as follows: A1. Determine the simulation model and calculate the range of virtual inertia J and virtual damping D; A2. Constructing the power response overshoot δ of the grid-side converter for a medium-frequency grid-type wind turbine. P % of the objective function; A3. Constructing the maximum frequency deviation Δω of the grid-side converter frequency response of a medium-frequency grid-type wind turbine. max The objective function; A4. Construct the objective function of the damping ratio ζ of the dominant oscillation mode of the grid-side converter of the medium-frequency grid-type wind turbine; A5. Overshoot δ of the grid-side converter power response of a medium-frequency grid-connected wind turbine. P %, the maximum frequency fluctuation of the frequency response Δω max Based on the objective function of the damping ratio ζ of the dominant oscillation mode, the overall objective function is constructed. A6. Using the maximum frequency change rate RoCoF under the constraint condition max and active power regulation time t P Construct the penalty function.

[0045] A7. Construct the final applicability function based on the objective function and the penalty function.

[0046] The solution steps for the multi-objective optimization model in step B are as follows: B1. Construct a multi-objective optimization model based on the objective function; B2. Initialize the particle swarm size, position, and velocity; B3. Substitute the filter parameters into the model and solve the model to obtain the weighted objective function value Z of T, ΔP, and ζ; B4. Compare the obtained objective function value with the historical best value of the objective function. If the obtained objective function value is better than the historical best value of the objective function, store the obtained objective function value in the optimal value of the objective function, store the obtained parameter combination in the optimal parameters, and then proceed to the next step. If the obtained objective function value is not better than the historical best value of the objective function, proceed directly to the next step. B5. Determine if the maximum number of iterations has been reached. If it has, output the current optimal parameters. If not, update the population position and velocity, and return to step B3.

[0047] The following description, in conjunction with a preferred embodiment, illustrates the content involved in the above embodiments.

[0048] A grid-connected model of a mid-frequency grid-connected wind turbine converter was built in Simulink. The converter adopted a grid-connected control strategy based on a virtual synchronous generator. The mid-frequency was selected as 200Hz, the filter was an LC filter, and the grid-connected power supply was a mid-frequency power supply. The grid equivalent resistance and inductance were calculated based on the short-circuit ratio. Small-signal full-order matrix analysis code based on state-space equations was written in Matlab to calculate the eigenvalues ​​under different virtual inertia J and virtual damping D.

[0049] Its system parameters are shown in Table 1: Table 1 System Parameters A grid-based control strategy based on virtual synchronous generators is adopted. First, the ranges of virtual inertia J and virtual damping D are calculated. The range of virtual inertia J is 1.8 ≤ J ≤ 7.6 kg / m². 2 The range of virtual damping D is 40≤D≤140.

[0050] Set the initial value of the virtual inertia J to 6.5 and the initial value of the virtual damping D to 50; set the algorithm model with ω1=0.2, ω2=0.2, ω3=0.6; c1=0.5, c2=0.5; number of particles n=10; number of iterations k=100. Initialize the particle position and velocity.

[0051] Run the model, the model is as follows Figure 4The algorithm flow shown is executed, and the optimal virtual inertia J is found to be 7.35 kg / m. 2 The optimal virtual damping D is 127.

[0052] Example Analysis: The initial values ​​of virtual inertia J and virtual damping D are set to 6.5 and 50 respectively. These values ​​are then substituted into the grid-connected model of the grid-side converter of the medium-frequency grid-type wind turbine. To verify the performance of the proposed method, it is compared with the scheme that only optimizes the damping ratio of the dominant eigenvalue.

[0053] The data before and after optimization are shown in Table 2: Table 2 Comparison before and after optimization Figures 5-7 The dominant mode root locus of the system is determined by continuously changing the short-circuit ratio (SCR) from 1.5 to 8 under three combinations of virtual inertia J and virtual damping D. (Comparison) Figure 5 , Figure 6 , Figure 7 It can be seen that when the short-circuit ratio before optimization is changed to 7.3, the system loses stability; only the damping ratio optimization scheme and the scheme proposed in this paper remain stable as the SCR changes from 1.5 to 8, demonstrating good stability capabilities.

[0054] Figure 8 The power response of the scheme before optimization, the scheme that only optimizes the damping ratio, and the scheme proposed in this paper are compared when there is a small disturbance in the q-axis voltage. The power response before optimization has a large overshoot and requires oscillation for many cycles to finally converge to the steady state value. The scheme proposed in this paper has a slightly larger overshoot than the scheme that only optimizes the damping ratio, but it is maintained within a good fluctuation range.

[0055] Figure 9 This paper compares the frequency responses of the unoptimized scheme, the scheme that only optimizes the damping ratio, and the scheme proposed in this paper when there is a small disturbance in the q-axis voltage. The unoptimized scheme has a large maximum deviation in frequency response and requires multiple oscillations to finally converge to the steady-state value. The proposed scheme can reduce the frequency fluctuation under disturbance conditions and improve the frequency stability of the system compared with the scheme that only optimizes the damping ratio.

[0056] Figure 10 The power responses of the unoptimized scheme, the scheme that only optimizes the damping ratio, and the scheme proposed in this paper are compared under the condition of a step wind speed. The unoptimized scheme has a large overshoot and requires multiple oscillations to finally converge to the steady-state value. The scheme proposed in this paper has a slightly larger overshoot than the scheme that only optimizes the damping ratio, but it is maintained within a good fluctuation range.

[0057] Figure 11The frequency responses of the unoptimized scheme, the scheme that only optimizes the damping ratio, and the scheme proposed in this paper are compared under the condition of a step wind speed. The unoptimized scheme has a large maximum deviation in frequency response and requires multiple oscillations to finally converge to the steady-state value. The scheme proposed in this paper can reduce the frequency fluctuation under disturbance conditions and improve the frequency stability of the system compared with the scheme that only optimizes the damping ratio.

[0058] Please see Figure 12 The diagram shows a structural schematic of a computer device provided in an embodiment of this application. An embodiment of this application provides a computer device 400, including a processor 410 and a memory 420. The memory 420 stores a computer program executable by the processor 410. When the computer program is executed by the processor 410, it performs the method described above.

[0059] This application embodiment also provides a storage medium 430, on which a computer program is stored, and the computer program is executed by a processor 410 to perform the above method.

[0060] The storage medium 430 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0061] Although this application has been described in conjunction with specific features and embodiments, it is obvious that various modifications and combinations can be made thereto without departing from the spirit and scope of this application. Accordingly, this specification and drawings are merely exemplary illustrations of the application as defined herein, and are to be considered as covering any and all modifications, variations, combinations, or equivalents within the scope of this application. Clearly, those skilled in the art can make various alterations and modifications to this application without departing from its scope. Thus, if such modifications and modifications fall within the scope of this application and its equivalents, this application intends to include such modifications and modifications.

Claims

1. A multi-objective optimization method for control parameters of grid-side converters in medium-frequency grid-connected wind turbine units, characterized in that, Includes the following steps: Determine the grid-connected model of the grid-side converter of the medium-frequency grid-type wind turbine, and calculate the range of virtual inertia and virtual damping; A multi-objective optimization function is established with the optimization objectives of minimizing the overshoot of the power response of the grid-side converter of the medium-frequency grid-type wind turbine, minimizing the maximum frequency fluctuation of the frequency response, and maximizing the damping ratio of the dominant oscillation mode. Using the parameter combination of the virtual inertia and virtual damping as population parameters and the multi-objective optimization function as the applicability function, a multi-objective optimization model is constructed. The multi-objective optimization model is solved using the particle swarm optimization algorithm to obtain the optimal solution for the combination of virtual inertia and virtual damping parameters.

2. The multi-objective optimization method for control parameters of the grid-side converter of a medium-frequency grid-connected wind turbine as described in claim 1, characterized in that, The determination of the grid-side converter grid-connection model for medium-frequency grid-type wind turbines includes LC-type filter circuits and the equivalent impedance of power grid transmission lines. The LC-type filter circuit is based on a mathematical model of state-space equations: In the formula: For filtering inductors and resistors; , These are the filter inductor and filter capacitor; The d-axis component of the output current of the grid-side converter of the medium-frequency grid-type wind turbine unit; This refers to the q-axis component of the output current of the grid-side converter of a medium-frequency grid-connected wind turbine. This represents the d-axis component of the voltage across the filter capacitor. This represents the q-axis component of the voltage across the filter capacitor. The actual angular velocity of the grid-side converter of the medium-frequency grid-type wind turbine unit; The d-axis component of the current after filtering by the grid-side converter of the medium-frequency grid-type wind turbine; The q-axis component of the current after filtering by the grid-side converter of the medium-frequency grid-connected wind turbine unit; The mathematical model of the equivalent impedance of the power grid transmission line based on the state-space equation is as follows: In the formula: The equivalent resistance of the power grid transmission line; The equivalent inductance of the power grid transmission line; The d-axis component of the intermediate frequency equivalent grid-connected voltage source; This is the q-axis component of the medium-frequency equivalent grid-connected voltage source.

3. The multi-objective optimization method for control parameters of the grid-side converter of a medium-frequency grid-connected wind turbine as described in claim 1, characterized in that, The grid-side converter adopts power outer loop and voltage and current dual closed loop control; The power outer loop is controlled by a virtual synchronous generator. The active and reactive power outputs of the grid-side converter of the medium-frequency grid-connected wind turbine are calculated using the inductor current and capacitor voltage of the AC-side filter, as shown in the following formula: In the formula, The active power output of the grid-side converter for medium-frequency grid-type wind turbine units; The reactive power output of the grid-side converter for medium-frequency grid-type wind turbines; The active power outer loop uses virtual synchronous generator control, and the reactive power outer loop uses droop control. The mathematical model based on the state-space equations is as follows: In the formula, This is virtual inertia; For virtual damping; The actual angular velocity of the grid-side converter of the medium-frequency grid-type wind turbine unit; The reference angular velocity for the grid-side converter of a medium-frequency grid-type wind turbine; The reference active power output of the grid-side converter for medium-frequency grid-type wind turbine units; The reference reactive power output of the grid-side converter for medium-frequency grid-type wind turbine units; This is the active power droop factor; This refers to the reactive power droop factor. Generate d-axis reference voltage for the grid-side converter of medium-frequency grid-connected wind turbine units; This is the reference voltage for the grid-side converter of a medium-frequency grid-connected wind turbine. The power inner loop of the grid-side converter of the medium-frequency grid-type wind turbine adopts a dual closed-loop design for voltage and current. The mathematical expression for the inner voltage loop is: In the formula: This is an intermediate variable for the d-axis PI integrator in the inner voltage loop; This is an intermediate variable for the q-axis PI integrator in the voltage inner loop; The proportional gain of the voltage inner-loop PI controller; The integral coefficient of the voltage inner-loop PI controller; Generate d-axis reference voltage for the grid-side converter of medium-frequency grid-connected wind turbine units; Generate q-axis reference voltage for the grid-side converter of medium-frequency grid-connected wind turbine units; This is the d-axis reference current value generated by the voltage inner loop; This is the q-axis reference current value generated by the voltage inner loop; The mathematical expression for the inner current loop is: In the formula, For the intermediate variable of the d-axis PI integrator in the inner current loop; This is an intermediate variable for the q-axis PI integrator in the inner current loop; The proportional gain of the inner-loop PI controller; The integral coefficient of the current inner loop PI controller.

4. The multi-objective optimization method for control parameters of the grid-side converter of a medium-frequency grid-connected wind turbine as described in claim 2, is characterized in that, After determining the grid-connection model of the grid-side converter of the medium-frequency grid-type wind turbine, the process also includes constructing a full-order small-signal linearization model: ； In the formula, x represents the system's state variables, which have 12 orders: In the formula, Let be the system angular velocity. , The dq coordinate system components of the grid-side output current of the medium-frequency grid-type converter; , For the dq coordinate system components of the line inductance current; u d u q The dq coordinate system components of the grid-side output voltage of the medium-frequency grid-type converter; , The dq coordinate system components of the filter capacitor voltage in a medium-frequency grid-type converter; , This is an intermediate variable for the voltage inner-loop PI integrator; , This is an intermediate variable for the inner loop PI integrator of the current; For the angle of attack.

5. The multi-objective optimization method for control parameters of the grid-side converter of a medium-frequency grid-connected wind turbine as described in claim 1, characterized in that, The establishment of the multi-objective optimization function includes: Constructing the power response overshoot of the grid-side converter for medium-frequency grid-connected wind turbine units The objective function, with overshoot The overshoot is minimized as the objective. The calculation method is as follows: In the formula: This refers to the power response overshoot of the grid-side converter in a medium-frequency grid-connected wind turbine. This represents the maximum power response of the grid-side converter in a medium-frequency grid-connected wind turbine unit. This represents the steady-state power response of the grid-side converter in a medium-frequency grid-connected wind turbine generator set. Maximum frequency fluctuation in the frequency response of the grid-side converter for medium-frequency grid-connected wind turbines The objective function is to maximize the frequency fluctuation. Minimize the maximum frequency fluctuation amount The calculation method is as follows: In the formula: t1 represents the maximum frequency fluctuation; t2 represents the start time of the disturbance; and t2 represents the end time of the recovery to stability. Constructing the dominant mode damping ratio of oscillation in the grid-side converter of a medium-frequency grid-connected wind turbine The objective function is to maximize the damping ratio, where the damping ratio of the dominant oscillation mode is... The calculation method is as follows: In the formula: the subscript i indicates that the i-th eigenvalue is the dominant mode; Damping ratio for the dominant mode; The real part of the dominant characteristic root; The imaginary part of the dominant characteristic root; Overshoot of grid-side converter power response in medium-frequency grid-connected wind turbine units Maximum frequency fluctuation in frequency response oscillation-dominant mode damping ratio Based on the objective function, the three objective functions are normalized, and then the total objective function F is obtained by combining the weighted coefficients. obj (J,D), calculated as follows: In the formula: , , The weighting coefficients are, in order, the overshoot, the maximum frequency deviation, and the dominant mode damping ratio; The normalized overshoot value; The maximum frequency fluctuation after normalization value; The normalized damping ratio of the dominant oscillation mode value; The normalization calculation method is as follows: In the formula, This is the maximum permissible overshoot. This is the maximum permissible frequency fluctuation. This is the minimum permissible dominant mode damping ratio.

6. The multi-objective optimization method for control parameters of the grid-side converter of a medium-frequency grid-connected wind turbine as described in claim 5, is characterized in that, The multi-objective optimization function also includes a penalty function, the construction of which includes: In the formula: This is the penalty coefficient; The maximum rate of change of frequency; This refers to the active power adjustment time.

7. The multi-objective optimization method for control parameters of the grid-side converter of a medium-frequency grid-connected wind turbine as described in claim 6, characterized in that, Based on the objective function and the penalty function, the final fitness objective function is constructed, and its mathematical expression is as follows: 。 8. The multi-objective optimization method for control parameters of the grid-side converter of a medium-frequency grid-connected wind turbine as described in claim 1, characterized in that, The step of solving the multi-objective optimization model using the particle swarm optimization algorithm includes the following steps: Initialize the particle swarm's number, position, and velocity; The parameter combination of the virtual inertia J and virtual damping D is substituted into the multi-objective optimization model as population parameters, and the multi-objective optimization model is solved to obtain the overshoot of the grid-side converter power response of the medium-frequency grid-type wind turbine. Maximum frequency fluctuation in frequency response oscillation-dominant mode damping ratio Weighted final fitness objective function value ; The obtained objective function value is compared with the historical best value of the objective function. If the obtained objective function value is better than the historical best value of the objective function, the obtained objective function value is stored in the optimal objective function value, the obtained parameter combination is stored in the optimal parameter value, and then the next step is executed. If the obtained objective function value is not better than the historical best value of the objective function, the next step is executed directly. Determine if the maximum number of iterations has been reached. If it has, output the current optimal parameters. If not, update the population position and velocity, repeat the substitution of parameters, and solve for the final fitness objective function value Z.

9. The multi-objective optimization method for control parameters of the grid-side converter of a medium-frequency grid-connected wind turbine as described in claim 8, characterized in that, The rate of updating population location includes: The MOPSO algorithm is used to update the population position and velocity. The velocity and position of particles in the MOPSO algorithm are shown in the following equation: In the formula: Inertial weight; For individual learning factors; As a social learning factor; , These are random values ​​generated between 0 and 1; Let t be the velocity of the t-th generation; For the t-th generation position; The optimal particle value for an individual; This represents the optimal value for all particles.

10. The multi-objective optimization method for control parameters of the grid-side converter of a medium-frequency grid-connected wind turbine as described in claim 9, characterized in that, In the process of updating the population position and velocity using the MOPSO algorithm, a linear weighting method is used to adjust the inertia weights. Optimization is performed, where in the k-th iteration... value for: In the formula: G is the maximum number of iterations; The maximum weight value; This is the minimum weight value.

11. A multi-objective optimization device for control parameters of a grid-side converter in a medium-frequency grid-connected wind turbine generator set, characterized in that, The apparatus for using the multi-objective optimization method for control parameters of the grid-side converter of a medium-frequency grid-connected wind turbine as described in any one of claims 1 to 10 includes: The range unit is used to determine the grid-connected model of the grid-side converter of the medium-frequency grid-type wind turbine and to calculate the range of virtual inertia and virtual damping. Target unit for overshoot of grid-side converter power response in medium-frequency grid-connected wind turbine units. Minimize the maximum frequency fluctuation of the frequency response Minimize the damping ratio of the dominant oscillation mode. To maximize the optimization objective, a multi-objective optimization function is established. The modeling unit is used to build a multi-objective optimization model using a combination of virtual inertia and virtual damping parameters as population parameters and a multi-objective optimization function as the applicability function. The solution unit is used to solve the multi-objective optimization model using the particle swarm optimization algorithm to obtain the optimal solution for the combination of virtual inertia and virtual damping parameters.

12. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the method as described in any one of claims 1-10.

13. A storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method as described in any one of claims 1-10.

Images