Improved step-inertia control frequency regulation strategy for wind turbine based on optimized power curve

Abstract

The application discloses a wind turbine improved step inertia control frequency modulation strategy based on an optimized power curve, first constructs a mathematical model of power curve optimization, sets a target function of the model as a frequency deviation change rate of 0, ensures that the frequency does not drop again after frequency modulation is finished, then designs a neural network controller by using the frequency, speed and power of the model, establishes a relationship between system operation state information and output power, so that the application of the improved strategy in different operation scenes of the wind turbine is realized. The application avoids the power sudden drop phenomenon of the traditional step inertia control strategy by the method of improving the power curve, and solves the problem of secondary frequency drop.

Description

Technical Field

[0001] This invention belongs to the field of power system frequency control technology, specifically relating to an improved step inertial control frequency regulation strategy for wind turbines based on optimized power curves. Background Technology

[0002] Clean energy, primarily wind power, is highly favored. However, a high proportion and large scale of wind power will reduce the equivalent inertia of the power system when connected to the grid, threatening the system's frequency stability. Furthermore, the uncertain fluctuations in wind power generation will inevitably affect the system frequency. Therefore, wind power needs a certain frequency regulation capability when connected to the grid. Simultaneously, the rapid development of power electronics technology has led to the application of numerous power electronic devices in wind power generation equipment, decoupling the system frequency from the wind turbine, thus differentiating it from traditional synchronous generators. Therefore, understanding the impact of wind power generation on the power system's inertia response and improving the frequency regulation capability in wind power grid-connected scenarios can not only prevent frequency exceedances but also provide a fundamental guarantee for the safety of system operation and prevent serious cascading accidents.

[0003] When Stepwise Inertia Control (SIC) is applied to wind power frequency regulation, it can significantly reduce the frequency sag when the system is subjected to active power disturbances by releasing rotor kinetic energy for short-term power support. However, this strategy suffers from a large power drop during rotor speed recovery, leading to a severe secondary frequency sag and causing another impact on grid stability. Therefore, finding a method to mitigate the secondary frequency sag of this strategy is of great significance for enabling wind power to participate in frequency regulation. Summary of the Invention

[0004] The purpose of this invention is to provide an improved step inertial control frequency regulation strategy for wind power based on optimized power curves, which can be used to solve the frequency double drop problem existing in traditional SIC strategies.

[0005] The technical solution for achieving the objective of this invention is as follows: Firstly, this invention provides a wind power improved step inertial control frequency regulation strategy based on optimized power curves, comprising:

[0006] Step 1: Obtain the wind turbine speed ratio, system frequency deviation, and frequency deviation change rate. When the system frequency deviation is greater than the maximum value of the wind turbine frequency regulation dead zone, the wind turbine participates in system frequency regulation.

[0007] Step 2: Based on the predetermined frequency regulation strategy and the wind turbine output power obtained by the neural network controller, guide the wind turbine to participate in frequency regulation.

[0008] In a second aspect, the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the method described in the first aspect.

[0009] Thirdly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described in the first aspect.

[0010] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0011] This invention employs a neural network controller to correlate the wind turbine's own speed changes and system frequency changes with the wind turbine's output power during frequency regulation, enabling the wind turbine to improve the secondary frequency drop while avoiding speed overruns during frequency regulation.

[0012] The present invention will be further described below with reference to specific embodiments and accompanying drawings. Attached Figure Description

[0013] Figure 1 This is a control block diagram of an improved step inertial control frequency modulation strategy for wind turbines based on optimized power curves, as described in an embodiment of the present invention.

[0014] Figure 2 This is a diagram illustrating the neural network training effect in an embodiment of the present invention;

[0015] Figure 3 This is a schematic diagram of the frequency regulation principle of a wind turbine generator set in an embodiment of the present invention;

[0016] Figure 4 This is a block diagram of a low-order system frequency response model in an embodiment of the present invention;

[0017] Figure 5 To verify the adaptability of the improved strategy in the embodiments of the present invention, a comparison chart of the frequency regulation effect of the wind turbine using the strategy of the present invention under different initial conditions is shown.

[0018] Figure 6 To verify the effectiveness of the improved strategy in the embodiments of the present invention, a comparison chart of the effects of wind turbines using the strategy of the present invention, the traditional SIC strategy, and wind power not participating in frequency regulation is provided. Detailed Implementation

[0019] like Figure 1 As shown, an improved step inertial control frequency regulation strategy for wind turbines based on optimized power curves includes the following steps:

[0020] Step 1: Obtain the wind turbine speed ratio, system frequency deviation, and frequency deviation change rate. When the system frequency deviation is greater than the maximum value of the wind turbine frequency regulation dead zone, the wind turbine participates in system frequency regulation.

[0021] Specifically, the domain of the wind turbine speed ratio Δω is [1, 1.25], and the frequency deviation change rate... The domain of discourse is [-0.5, 0.5], and the preset frequency dead zone threshold of the fan is (±0.03Hz).

[0022] Step 2: Based on the predetermined frequency regulation strategy and the wind turbine output power obtained from the neural network controller, the wind turbine is guided to participate in frequency regulation. Specifically:

[0023] Step 2-1: Construct an improved mathematical model for the power curve

[0024] Power curve optimization is performed by selecting the maximum frequency deviation as the optimization objective.

[0025]

[0026] Wherein, ΔP W is the rate of change of the wind turbine power reference value, i.e., the decision variable of the optimization model; t is the total optimization time; Δf(t) is the frequency deviation at time t; μ P This is the penalty coefficient, similar to the penalty factor in optimal power flow calculation. Its value is 1 when the optimization result satisfies the constraints, and tends to infinity otherwise.

[0027] Step 2-2: Design the neural network controller:

[0028] The ratio of the frequency change rate of the optimization model to the wind turbine speed is used as the input of the neural network controller, and the output power of the wind turbine is used as the output of the neural network controller.

[0029] Steps 2-3: Use the Radial Basis Function (RBF) algorithm as the neural network learning algorithm:

[0030]

[0031] y = w1h1 + w2h2 + ... + w m h m

[0032] Where x is the input layer of the neural network, h j b j c j Let be the basis vector, Gaussian function, and center point vector value of the j-th neuron, respectively; ||·|| is the Euclidean norm; y is the output value of the neural network; and w is the weight vector from the hidden layer to the output layer.

[0033] The wind turbine's participation in frequency regulation is guided by a predetermined frequency regulation strategy and the wind turbine's output power obtained from the neural network controller, and is divided into the following three stages:

[0034] The power generation phase consists of three stages: short-term power increase, power transition, and power recovery.

[0035] The short-term power boost phase is the stage where the rotor releases kinetic energy to provide power support. The start time of this phase is when the wind turbine participates in frequency regulation, and the end time is when the wind turbine exits frequency regulation. The power transition phase is the stage where the system frequency is maintained at a constant value while ensuring that the speed does not exceed the limit. The start time of this phase is when the wind turbine exits frequency regulation, and the end time is when the wind turbine resumes MPPT operation mode. The power recovery phase is the stage where the wind turbine recovers to the pre-disturbance state along the MPPT operation mode. The start time of this phase is when the wind turbine resumes MPPT operation mode, and it will continue to operate in MPPT operation mode thereafter.

[0036] The short-term boost phase employs the following methods to participate in frequency modulation:

[0037]

[0038] Where P is the output power of the wind turbine, P0 is the power of the wind turbine when it is operating at ω0, and ΔP up For short-term power generation, t is the current operating time of the wind turbine, t0 is the moment when the system is subjected to load disturbance, and t up This is to support the power output for a short period of time.

[0039] When the system is subjected to load disturbances, the wind turbine can quickly release rotor kinetic energy to support frequency. The difference between the improved SIC strategy and the traditional SIC strategy is that the improved strategy requires the wind turbine to exit the short-term power generation phase earlier, with the optimal exit time being the moment the system frequency reaches its lowest point. This helps reduce the consumption of rotor kinetic energy. In the low-order system frequency response (SFR) model, when subjected to a step load disturbance, the time t when the system frequency reaches its lowest point... out The parameters related to the SFR model can be calculated, specifically:

[0040]

[0041] In the formula, H S D is the equivalent inertial time constant. S R is the equivalent damping coefficient, R is the governor adjustment coefficient, and K is the speed governor adjustment coefficient. m F is the mechanical power gain coefficient. H T represents the work ratio of an equivalent high-pressure cylinder. RThe time constants are equivalent reheat time constants, where K1, K2, and K3 are all constants, and are related to H. S D S R, K m F H T R It depends on the value.

[0042] The power transition phase employs the following method to participate in frequency modulation:

[0043] When t≥t0+t up At that time, the wind turbine unit is based on the system frequency deviation rate of change. The output power of the wind turbine is adjusted by the ratio of its own rotational speed to Δω, specifically as follows:

[0044]

[0045] Among them, P t Let t be the output power of the wind turbine. w1 and w2 are the weights obtained from training the neural network, where w1 represents the rate of change of frequency deviation.

[0046] The power recovery phase employs the following methods for frequency modulation:

[0047] When the output power of the wind turbine is close to the rotational speed ω t When the corresponding MPPT operating power is reached, the control mode switches back to MPPT operating mode, specifically as follows:

[0048] P = P MPPT (ω t )

[0049] Among them, P MPPT (ω t ) represents the power on the MPPT curve.

[0050] Figure 2 , Figure 3 The images shown are a neural network training effect diagram and a wind turbine frequency regulation principle diagram, respectively, from an embodiment of the present invention.

[0051] Figure 4 This is a block diagram of the frequency response model of a low-order system, where H... s =8.0s, D s =1, R=0.05, K m =0.95, T R =8s, F H =0.3s.

[0052] Other system parameters are set as follows: simulation duration is 200s, load disturbance occurs at t=30s, and initial load P. LWith a power consumption of 1.0 pu, the wind speed was set to a constant 9 m / s when verifying the adaptability of the improved SIC strategy, and the wind power penetration rate was also kept constant at 15%. The load disturbance gradually increased from 5% to 13%. The frequency regulation effect diagram is shown below. Figure 5 As shown, under gradually increasing load disturbance, the power change of this invention during the transition phase is almost consistent, and the frequency deviation change rate is approximately zero, effectively avoiding the secondary frequency drop phenomenon and achieving the expected frequency regulation effect. The design of the neural network controller can rely on its own learning ability to adapt the improved SIC strategy to different operating scenarios of wind turbines. In verifying the effectiveness of the improved SIC strategy, the wind speed was maintained at 9 m / s and the wind power penetration rate at 15%. The difference is that the parameters of the traditional SIC strategy were all set to optimal: t off =8.754s, ΔP up =0.0181 pu; The short-term issuance support time of the strategy in this article is t. min1 =3.8s, ΔP set =0.001pu,t MPPT =60s. Frequency modulation effect comparison chart as shown below. Figure 6 As shown, this invention can significantly improve the minimum frequency of the power grid, with a maximum frequency deviation of only -0.25Hz. This represents a significant improvement compared to wind power not participating in frequency regulation and the traditional SiC strategy, and the frequency remains stable during the transition phase. In contrast, the traditional SiC strategy can only mitigate the magnitude of the second frequency drop to some extent by improving parameters, but it cannot avoid the second frequency drop altogether. Furthermore, it can be seen that even with the optimal parameters, the magnitude of the second frequency drop caused by a sudden power drop is almost the same as the frequency drop during load disturbances, resulting in limited frequency regulation effectiveness. On the other hand, this invention leads to an increase in the frequency regulation time of wind turbines, with a longer speed recovery time than the traditional SiC strategy. This results in a loss of power generation efficiency for wind turbines, which is the cost of the improved strategy.

Claims

1. An improved stepper inertial control frequency regulation strategy for wind turbines based on optimized power curves, characterized in that, Includes the following steps: Step 1: Obtain the wind turbine speed ratio, system frequency deviation, and frequency deviation change rate. When the system frequency deviation is greater than the maximum value of the wind turbine frequency regulation dead zone, the wind turbine participates in system frequency regulation. Step 2: Based on the predetermined frequency regulation strategy and the wind turbine output power obtained from the neural network controller, the wind turbine is guided to participate in frequency regulation; the wind turbine output power is determined by the following method: Step 2-1: Construct an improved mathematical model for the power curve Power curve optimization is performed by selecting the maximum frequency deviation as the optimization objective. ； Wherein, ΔP W is the rate of change of the wind turbine power reference value, i.e., the decision variable of the optimization model; t is the total optimization time; Δf(t) is the frequency deviation value at time t; This is the penalty coefficient; Step 2-2: Design the neural network controller: The frequency deviation rate of the optimization model and the ratio of wind turbine speed are used as inputs to the neural network controller, and the output power of the wind turbine is used as the output of the neural network controller. Steps 2-3: Use the radial basis function algorithm as the neural network learning algorithm: ； ； Where x is the input layer of the neural network, , , The first The basis vectors, Gaussian function, and center point vector values ​​of each neuron. Let y be the Euclidean norm, and y be the output value of the neural network. ~ These represent the weight vectors from the 1st to the mth neurons in the hidden layer to the output layer.

2. The improved stepper inertial control frequency regulation strategy for wind turbines based on optimized power curves according to claim 1, characterized in that, The wind turbine speed ratio is determined by the following method: ； in, This refers to the rotational speed of the wind turbine at the current operating moment. This is the minimum speed limit for wind turbine generators.

3. The improved stepper inertial control frequency regulation strategy for wind turbines based on optimized power curves according to claim 2, characterized in that, The wind turbine's participation in frequency regulation is guided by a predetermined frequency regulation strategy and the output power obtained from the neural network controller. The process is divided into three phases: a short-term boost phase, a power transition phase, and a power recovery phase. The short-term boost phase involves the rotor releasing kinetic energy for power support, and this phase begins when the wind turbine participates in frequency regulation. The power transition phase ensures the system frequency remains constant while maintaining the rotational speed within limits, and this phase ends when the wind turbine resumes MPPT (Multi-Pulse Test) operation. The power recovery phase is when the wind turbine returns to its pre-disturbance state under MPPT operation, and this phase begins when the wind turbine resumes MPPT operation, after which it will continue to operate in MPPT mode.

4. The improved stepper inertial control frequency regulation strategy for wind turbines based on optimized power curves according to claim 3, characterized in that, The short-term boost phase employs the following methods to participate in frequency modulation: ； Where P is the output power of the wind turbine, and P0 is the operating temperature of the wind turbine. Power at that time This is the initial speed of the wind turbine. For short-term power generation, t represents the current operating time of the wind turbine, and t0 represents the moment when the system is subjected to load disturbance. This is to support the power output for a short period of time.

5. The improved stepper inertial control frequency regulation strategy for wind turbines based on optimized power curves according to claim 4, characterized in that, The power transition phase employs the following method to participate in frequency modulation: ； in, Let t be the output power of the wind turbine. w1 and w2 are the weights obtained from training the neural network, where w1 represents the rate of change of frequency deviation.

6. The improved stepper inertial control frequency regulation strategy for wind turbines based on optimized power curves according to claim 5, characterized in that, The power recovery phase employs the following methods for frequency modulation: ； in, This represents the power on the MPPT curve.

7. An electronic 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 program, it implements the steps of the strategy as described in any one of claims 1-6.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps of the strategy as described in any one of claims 1-6.

Images