A method for wind power capacity optimization considering dynamic frequency constraints and multiple uncertainties

Abstract

The wind power installed capacity optimization method considering dynamic frequency constraint and multiple uncertainties comprises the following steps: step 1, a dynamic frequency response model containing load side inertia is established, and the expression of the lowest frequency point and the frequency change rate is obtained; step 2, a whole optimization model based on the wind turbine type and the number of wind turbines is established, and secondary integer variable linearization processing is carried out; step 3, an installed capacity optimization model considering multiple uncertainty fuzzy chance constraints is established; and step 4, the maximum wind farm income is taken as the objective function, and the optimal wind power installed capacity is solved. The method can enhance the ability of the system to resist sudden power disturbance, greatly improve the frequency safety of the system, improve the inertia response and frequency modulation capacity of the system, enhance the ability of wind energy consumption, and effectively improve the wind power installed capacity and the net income of the wind farm.

Description

Technical Field

[0001] This invention relates to the field of wind power installed capacity optimization technology, specifically to a wind power installed capacity optimization method that considers dynamic frequency constraints and multiple uncertainties. Background Technology

[0002] As one of the most promising renewable energy sources, wind power is increasingly accounting for a larger share of the power grid. By the end of 2022, my country's grid-connected wind power capacity had increased by 37.63 million kilowatts, bringing the cumulative grid-connected capacity to 365 million kilowatts, a year-on-year increase of 21%. However, the continuous increase in wind power capacity has exacerbated the problems of insufficient system inertia support and frequency response capabilities, posing new challenges to grid frequency security.

[0003] The determination of wind farm installed capacity is influenced not only by the size of wind energy resources in the proposed wind farm area and the wind farm operator's operating methods, but also by the grid's capacity to accommodate wind power in the region. In recent years, numerous large-scale blackouts caused by frequency instability have occurred worldwide, demonstrating that frequency stability issues have become increasingly important as the proportion of wind power connected to the grid increases. Therefore, frequency security constraints should be considered when studying wind power acceptance capacity and determining wind power installed capacity.

[0004] The dynamic frequency response capability of a system is closely related to its inertia level. Exploring system inertia can effectively improve the system's ability to withstand frequency safety incidents, thereby increasing the installed capacity of new energy sources. Previous dynamic frequency response models only considered the power source inertia, neglecting the impact of load-side inertia. Research shows that load-side inertia accounts for approximately 20% of the system's inertia. Taking load-side inertia into account can improve the system's inertial support capability, thus effectively increasing the installed capacity of wind farms and the revenue of wind farm operators. However, due to the randomness of users' energy consumption habits, the load is uncertain, and therefore, the load-side inertia is also uncertain. Therefore, the uncertainties of wind power, load power, and load-side inertia need to be considered in the optimization of wind power installed capacity. Summary of the Invention

[0005] Based on the above background, this invention comprehensively considers wind turbine type, dynamic frequency constraints, and load-side inertia to establish a wind power installed capacity optimization method with multiple uncertainties and fuzzy opportunity constraints. This method improves the grid's ability to resist sudden disturbances, enhances the grid's inertial response and frequency regulation capabilities, and effectively increases wind power installed capacity and wind farm net revenue while ensuring grid frequency safety and stability.

[0006] The technical solution adopted in this invention is as follows: The wind power installed capacity optimization method considering dynamic frequency constraints and multiple uncertainties includes the following steps: Step 1: Establish a dynamic frequency response model including the load-side inertia, and derive the expressions for the minimum frequency and the rate of frequency change; Step 2: Based on the overall optimization model of wind turbine type and number of wind turbines, and perform quadratic integer variable linearization processing; Step 3: Establish an installed capacity optimization model that considers multiple uncertainties and fuzzy opportunity constraints; Step 4: Using the maximization of wind farm revenue as the objective function, solve for the optimal wind power installed capacity.

[0007] In step 1, based on the equivalent frequency response model, the dynamic frequency response model is derived, and the load-side inertia is introduced into the dynamic frequency response model to derive the dynamic frequency response model containing the load-side inertia, thus obtaining the expressions for the minimum frequency point and the rate of frequency change.

[0008] The dynamic frequency response model including the load-side inertia is shown in equations (1) to (5) below: The rotor motion equations for this model are expressed as follows: (1); In the formula: The time constant of the synchronization inertia; The load-side inertia time constant; This refers to the system's frequency deviation. This refers to the active power adjustment amount for primary frequency regulation of the generating unit. The active power imbalance caused by the disturbance; is the damping coefficient of the system.

[0009] The results of the frequency minimum point derived from the dynamic frequency response model including load-side inertia are as follows: (2); (3); In the formula: Indicates the point of lowest frequency; This indicates the time corresponding to the point of lowest frequency. The initial frequency of the system; This is the system's reference frequency; This indicates the total number of thermal power units. The number indicates the number of thermal power units.

[0010] Each conventional unit The expression is as follows: (4); In the formula: For the first Start-up and shutdown status of the generator set; For the first The unit regulating power coefficient of a generator; For the first The response time constant of the generator. ; Indicates the start / stop status of the first generating unit; This represents the unit regulating power coefficient of the first generator; This represents the response time constant of the first generator.

[0011] The expression for the rate of change of frequency is as follows: (5); Indicates the rate of change of frequency; Indicates the frequency deviation of the system; Indicates the time of change in system frequency deviation; This represents the value of the frequency deviation when differentiated with respect to time.

[0012] In step 2, the historical wind speed of the region over a year is obtained, measured every 5 minutes. The wind speed is processed to obtain the typical daily wind speed value. An overall optimization model for wind turbine type and number of wind turbines is established. Based on the linearization of quadratic integer variables, the overall optimization model for wind turbine type and number of wind turbines is transformed into a mixed integer programming model.

[0013] This invention builds upon previous studies that only optimized the number of wind turbines to determine wind power capacity. It also incorporates wind turbine type as an optimization variable, proposing that optimizing the grid-connected capacity of a wind farm requires optimizing both the wind turbine type and the number of wind turbines connected to the grid. This improves the engineering applicability of the proposed scheme under different conditions and balances the benefits of the wind farm with grid security.

[0014] The overall optimization model for wind turbine type and number of wind turbines is as follows: (6); (7); (8); In the formula: This represents the total installed capacity of all wind turbine units connected to the grid. This represents the maximum number of installed units connected to the grid, and is the variable to be optimized in this invention. This represents a 0-1 type decision variable, used for screening wind turbine types; This represents the set of rated power for all fan types to be selected; Indicates the number of fan types; This indicates the rated power of the selected single fan; The number indicates the number of fan types.

[0015] In summary, equation (6) is a nonlinear expression. Therefore, the invention employs a quadratic integer variable linearization method, introducing decision variables... The expression is as follows: (9); (10); Where: decision variables Characterization and The product; It is a relatively large constant; This indicates the result based on the values ​​of the decision variables; A set representing the number of fan types; This indicates the maximum number of wind turbines that can be connected to the grid.

[0016] The two inequality constraints in equation (9) are to enable equation (6) to be successfully transformed into equation (10) and achieve the purpose of linearization.

[0017] In step 3, since wind power, load and load-side inertia all have uncertainties, the uncertainty of load-side inertia is analyzed, and fuzzy opportunity constraints are introduced. Based on the dynamic frequency constraints, a fuzzy opportunity constraint model with multiple uncertainties is constructed, and then the uncertainty constraints in the fuzzy opportunity constraint model are transformed into clear equivalence classes. The specific implementation steps of the fuzzy chance constraint model with multiple uncertainties are shown in step 3 below. The model is shown in formulas (14) to (28).

[0018] When the confidence level of the fuzzy chance constraint At that time, it was shaped like , Represents the constraint function; The trapezoidal fuzzy parameters can be represented as follows: The clear equivalence class form can be expressed as: (11); In equation (11): , This indicates the definition of a function; This represents the trapezoidal membership parameter. ; Represent real numbers; Indicates the number of possible values ​​in the set; Representation function Part of it.

[0019] In step 4, considering the high nonlinearity of the dynamic frequency constraint model, the optimization problem is divided into a main problem and sub-problems. The optimal wind power installed capacity is solved with the objective function of maximizing wind farm revenue. The established expression for separating the main problem and subproblems is as follows: (12); In the formula: The unit adjustable power corresponding to the failure to meet the dynamic frequency constraint; Representation system The unit adjustment power at any given time.

[0020] The expression for the net income of a wind farm is: (13); In the formula: For the revenue from selling electricity from wind farms; The investment cost of a wind farm; This refers to the operation and maintenance costs of wind farms.

[0021] This invention provides a method for optimizing wind power installed capacity that considers dynamic frequency constraints and multiple uncertainties, with the following technical advantages: 1) In step 1 of the present invention, dynamic frequency constraints are introduced on the basis of traditional wind power installed capacity optimization, which can ensure the safety of grid frequency; further, load-side inertia is introduced, which can improve the system inertial response and frequency regulation capability while taking into account the safety of grid frequency, thereby increasing wind power installed capacity and wind farm net revenue.

[0022] 2) In step 2 of this invention, based on the traditional scheme that only considers the number of wind turbines as the optimization variable, the wind turbine type is also proposed as the optimization variable, and a quadratic integer variable linearization method is used for linearization. Compared with separately optimizing the wind turbine type, the overall optimization can obtain the best wind turbine type faster, improve the solution efficiency, and enable wind farm operators to obtain greater benefits.

[0023] 3) In step 3 of this invention, fuzzy chance constraints are introduced, which can characterize uncertain variables through trapezoidal fuzzy parameters and perform clear equivalence class transformation based on the credibility theory in uncertain programming, so that the model can be solved quickly by the solver.

[0024] 4) Step 4 of the present invention takes the maximization of the net benefit of the wind farm as the objective function. In view of the high nonlinearity of the dynamic frequency constraint model, the optimization model is decomposed into the main problem and sub-problems, which can yield the optimal wind power installed capacity.

[0025] 5) The method of the present invention can enhance the system's ability to resist sudden power disturbances, greatly improve the system's frequency security, improve the system's inertial response and frequency regulation capability, and enhance the ability to absorb wind energy, thereby effectively increasing the installed capacity of wind power and the net income of wind farms. Attached Figure Description

[0026] Figure 1 This is a diagram of the dynamic frequency response model including the load-side inertia.

[0027] Figure 2 Flowchart for overall model optimization solution.

[0028] Figure 3 This is a distribution diagram of the lowest frequency points for two different schemes and a traditional invention.

[0029] Figure 4 The frequency change rate distribution diagrams for two different schemes and a traditional invention are shown. Detailed Implementation

[0030] The technical method of the present invention will be further described in detail below through embodiments and data analysis, including the following steps: Step 1: Based on the equivalent frequency response model, derive the dynamic frequency response model, and introduce the load-side inertia into the model to derive the dynamic frequency response model containing the load-side inertia, and obtain the expressions for the minimum frequency point and the rate of frequency change.

[0031] The proposed wind farm needs to address the frequency stability issues caused by large-scale wind power grid connection in the region, and measures should be taken in advance. Therefore, dynamic frequency constraints are introduced to limit wind power grid connection. The frequency response process typically uses the frequency minimum point and the rate of frequency change to measure the transient frequency stability of the system. Therefore, this invention uses an improved system equivalent frequency response model to establish a dynamic frequency response model including load-side inertia, as shown in the model below. Figure 1 As shown.

[0032] The rotor motion equations for this model are expressed as follows: (1); In the formula: The time constant of the synchronization inertia; The load-side inertia time constant; This refers to the system's frequency deviation. This refers to the active power adjustment amount for primary frequency regulation of the generating unit. The active power imbalance caused by the disturbance; is the damping coefficient of the system.

[0033] The results of the minimum frequency point derived from the dynamic frequency response model are as follows: (2); (3); In the formula: The initial frequency of the system; The reference frequency for the system; each conventional unit The expression is as follows: (4); In the formula: For the first Start-up and shutdown status of the generator set; For the first The unit regulating power coefficient of a generator; For the first The response time constant of the generator.

[0034] The expression for the rate of change of frequency is as follows: (5); Step 2: Obtain the historical wind speed of the region for one year, measuring every 5 minutes. Process the wind speed to obtain typical daily wind speed values. Establish an overall optimization model for wind turbine type and number of wind turbines. Use a quadratic integer variable linearization method to transform the nonlinear model composed of wind turbine type and number of wind turbines into a mixed integer programming model.

[0035] To study the impact of different wind turbine types on wind power installed capacity, this invention, building upon previous studies that only optimized the number of wind turbines to determine wind power installed capacity, also considers wind turbine type as an optimization variable, proposing an overall optimization model for wind turbine type and number of wind turbines. The model is as follows: (6); (7); (8); In the formula: This represents the total installed capacity of all wind turbine units connected to the grid. This represents the maximum number of installed units connected to the grid, and is the variable to be optimized in this invention. This represents a 0-1 type decision variable, used for screening wind turbine types; This represents the set of rated power for all fan types to be selected; Indicates the number of fan types; This indicates the rated power of the selected single fan.

[0036] In summary, equation (6) is a nonlinear expression. Therefore, this invention employs a quadratic integer variable linearization method, introducing decision variables... The expression is as follows: (9); (10); Where: decision variables Characterization and The product; The constant is relatively large; the two inequality constraints in equation (9) are to enable equation (6) to be successfully transformed into equation (10) and achieve the purpose of linearization.

[0037] Step 3: Given the uncertainties in wind power, load, and load-side inertia, analyze the uncertainty of load-side inertia and introduce fuzzy chance constraints. Based on dynamic frequency constraints, construct a fuzzy chance constraint model with multiple uncertainties, and then transform the uncertainty constraints in the model into clear equivalence classes. Step 4: Given the high nonlinearity of the dynamic frequency constraint model, the optimization problem is divided into a main problem and sub-problems. The optimal wind power installed capacity is solved with the objective function of maximizing wind farm revenue.

[0038] 1) Uncertainty analysis of load-side inertia: The total system inertia should include both the power source inertia and the load-side inertia. In traditional power systems, the system inertia is sufficient, and the load-side inertia accounts for a relatively small proportion compared to the power source inertia, often being ignored when quantifying the overall system inertia. Existing inventions have demonstrated that the load-side inertia accounts for approximately 20% of the system inertia. Therefore, considering the impact of load-side inertia during wind power capacity optimization can improve the frequency regulation capability of the power grid, thereby effectively increasing the installed capacity of wind farms and the revenue of wind farm operators.

[0039] Load-side inertia is provided by synchronous motors, asynchronous motors within the load, and wind power, solar power, energy storage, and electric vehicles connected to the distribution network. The number of synchronous and asynchronous motors connected to the load is related to people's production and daily life; the amount of inertia provided by wind power and solar power in the distribution network is affected by meteorological factors and new energy control strategies; the amount of inertia provided by energy storage is affected by its control strategies and electricity price incentives; and the amount of inertia provided by electric vehicles is related to residents' living habits and electricity price incentives. In summary, load-side inertia has time-varying characteristics and exhibits significant random fluctuations.

[0040] Existing studies have used Gaussian distribution to describe load-side inertia, but information collection on load-side inertia is difficult. Fuzzy parameters, on the other hand, can obtain the membership function of uncertain parameters with the help of an expert system when there is insufficient information or no existing information. Therefore, they can better characterize the uncertainty of load-side inertia and load power response. Thus, this invention uses trapezoidal fuzzy parameters to describe the uncertainty of load-side inertia.

[0041] 2) Power output characteristics of wind turbine units Numerous studies have shown that the distribution characteristics of wind speed can be described by the Weibull distribution, and its probability density function can be expressed as: (14); In equation (14): Indicates shape parameters; It represents a scale parameter that reflects the average wind speed in a region.

[0042] The relationship between the output power of a wind turbine and the real-time wind speed can be expressed by a piecewise function, as follows: (15); In the formula: Indicates the cut-in wind speed; Indicates the cut-out wind speed; Indicates the rated wind speed; Indicates the rated power.

[0043] 3) Wind power installed capacity optimization model with fuzzy chance constraints: To study the optimal installed wind power capacity of a wind farm, a mathematical model was constructed with the objective function of maximizing the net benefit of the wind farm. However, due to multiple uncertainties arising from wind power output, load, and load-side inertia, the optimization results for wind power capacity using deterministic equality constraints tend to be conservative. Therefore, this invention proposes a fuzzy chance constraint model that considers multiple uncertainties, allowing decisions to be made to some extent without meeting the constraints, as long as the probability meets a given confidence level.

[0044] Objective function: (16); The wind farm generates electricity and sells it based on the local electricity price. The revenue expression for wind farm power generation is as follows: (17); (18); (19); (20); In the formula: Indicates the electricity price; This represents the estimated annual power generation of all grid-connected wind turbine units; This represents the maximum number of installed units connected to the grid, and is the variable to be optimized in this invention. This represents the average annual expected value of the wind turbine's output power; This represents the wind power loss factor, taken as 0.08; This represents the present value factor of an annuity. , These represent the discount rate and the lifespan, respectively, and are set to 5% and 20 years.

[0045] The investment cost of a wind farm is determined by the type of wind turbine and the number of turbines connected to the grid, thus the total investment cost of the wind farm can be calculated as follows: (twenty one); In the formula: This indicates the unit capacity investment cost of the selected fan type.

[0046] Wind farm operation and maintenance costs include the initial costs of hiring professionals to survey the environment and leasing land, as well as the subsequent costs of operating, maintaining, and insuring all wind turbines. The total operation and maintenance cost is: (twenty two); In the formula: This indicates the unit capacity maintenance cost of the selected fan type.

[0047] The model's general constraints are as follows: (1) Power balance constraint: (twenty three); In the formula: For the first Taiwan thermal power units Always put in the effort; For all wind turbine units Predict output in real time; For load The predicted value at any given time.

[0048] (2) Output constraints of thermal power units: (twenty four); In the formula: For the first Taiwan thermal power units The start / stop status at any given moment; , These represent the upper and lower limits of the output of thermal power units.

[0049] (3) Investment constraints for wind farms: (25); In the formula: Investment limits for wind farms.

[0050] (4) Rotational spare constraint: (26); In the formula: , Positive and negative rotational spares are required to account for load forecasting errors; , Positive and negative rotating reserves are provided for wind power prediction errors; both load and wind power prediction errors are represented by trapezoidal fuzzy parameters. This indicates the confidence level at which the fuzzy chance constraint is satisfied. The confidence level.

[0051] (5) Minimum frequency constraint: (27); In the formula: for The point of lowest frequency at any given moment; This is a frequency limit.

[0052] (6) Frequency change rate constraint: (28); In the formula: for Rate of change of frequency at any given moment; This is the limit for the rate of change of frequency.

[0053] 5) Clear equivalence class transformation of fuzzy opportunity constraints The fuzzy chance-constrained programming model composed of equations (26) to (28) above can be transformed into clear equivalence classes based on the credibility theory in uncertain programming.

[0054] When the confidence level of the fuzzy chance constraint At that time, it was shaped like The clear equivalence class form can be represented as: (11); In the formula: , This indicates the definition of a function; Indicates the trapezoidal membership parameter; Representation function Part of it.

[0055] Trapezoidal fuzzy parameters can be represented by quadruples: (29); In the formula: For the membership parameters of wind power, load, and load-side inertia; These are the predicted values ​​for uncertain variables; This is the proportionality coefficient, which is generally determined using historical data of fuzzy parameters.

[0056] The clear equivalence class transformation of the spin spare constraint (26) is shown below: (30); (31); In the formula: , , , This refers to the membership parameter for wind power. , , , This is the membership parameter for the load.

[0057] The clear equivalence class transformations of the minimum frequency equation (27) and the rate of change of frequency equation (28) are shown below: (32); (33); In the formula: , This is the membership parameter for the load-side inertia.

[0058] 6) Solution process Because dynamic frequency constraints are nonlinear, mixed-integer programming solvers cannot solve them directly. Therefore, this invention decomposes the optimization problem into a main problem and sub-problems for solution. The main problem is to determine the optimal wind power installed capacity by optimizing the wind turbine type and the number of grid-connected units. The sub-problem is the dynamic frequency constraint verification problem based on the main problem's unit start-up and shutdown scheme.

[0059] The overall solution process is as follows: Figure 2As shown. The unit start-up and shutdown states obtained from solving the main problem are used as known values ​​to check the minimum frequency point and the frequency change rate exceeding the limit problem of the subproblem. If equations (42)-(43) are not satisfied, that is, there is an over-limit action in the dynamic frequency constraint, it indicates that the system frequency regulation capability is insufficient, and the system Unit adjustment power at any time It can characterize the system's frequency regulation capability and is related to the unit's start-up and shutdown status. Its expression is as follows: (34); To ensure that dynamic frequency constraints meet the requirements at all times and to prevent excessive wind power grid connection capacity from causing the dynamic frequency to deteriorate further and leading to grid frequency safety incidents, the following segmentation is implemented: (12); In the formula: For the unit adjustment power that does not meet the dynamic frequency constraints of equations (32)-(33); when the dynamic frequency constraints have over-limit problems and When the optimal cut equation (12) is generated, the main problem is returned. By adjusting the number of units starting and stopping, the system is made to increase the unit regulation power to improve the system frequency regulation capability, thereby continuously optimizing the wind power installed capacity until all constraints are met and the optimal solution is obtained.

[0060] Example: To verify the feasibility and effectiveness of the method proposed in this invention, a simulation was conducted using an improved 10-unit system, and the planning model was solved using the Gurobi solver. The example uses measured wind speed data from a specific location for analysis. The wind turbine types and cost parameters are shown in Table 1, and the wind power grid connection price is taken as 98. The system parameters of the thermal power unit are shown in Table 2. The rated frequency of the system is set to 50Hz, the minimum frequency limit is set to 49.5Hz, the frequency change rate limit is set to 0.25Hz / s, and the disturbance power of the N-1 extreme contingency is set to 10% of the total load for the corresponding period. The confidence level is... Set it to 0.95.

[0061]

[0062]

[0063] 1) The impact of fan type on optimization results: Three different wind turbine types were used to calculate the installed capacity and net revenue of wind farms, and the results are shown in Table 3. Table 3 shows that the optimal installed capacity and net revenue of wind farms differ depending on the wind turbine type. Comparing wind turbine types b and c also reveals that a larger installed capacity does not necessarily lead to better net revenue for the wind farm. This is because different types of wind turbines have different output power and wind energy utilization rates at the same wind speed. Therefore, this invention uses wind turbine type as an optimization variable for overall optimization, allowing for the selection of appropriate wind turbine types based on local wind speed characteristics, reducing investment costs for wind farms, and maximizing investment returns.

[0064]

[0065] 2) Consider the impact of frequency safety constraints and load-side inertia on the optimization results: With the continuous increase in wind power installed capacity, the frequency characteristics of the power grid are significantly affected, and frequency security issues are becoming increasingly prominent. Therefore, considering dynamic frequency constraints and load-side inertia is particularly important in traditional wind power installed capacity optimization studies. To investigate the impact of dynamic frequency constraints and load-side inertia on wind power installed capacity, the following three schemes are compared and analyzed: Option 1: Wind power installed capacity optimization scheme without considering dynamic frequency constraints; Option 2: Wind power installed capacity optimization scheme taking into account dynamic frequency constraints; Option 3: Wind power installed capacity optimization scheme that takes into account dynamic frequency constraints and load-side inertia.

[0066] Depend on Figure 3 and Figure 4 It can be seen that although Scheme 1 has the largest installed wind power capacity, its minimum frequency and rate of change in periods 1-3 and 22-24 are both below the frequency limit, which may easily lead to frequency safety accidents when power disturbances occur. Schemes 2 and 3 both consider dynamic frequency constraints, and the optimization results can meet the frequency limits. Compared with Scheme 2, Scheme 3 further considers the load-side inertia, which increases the system inertia it considers, improves the system frequency regulation capability, and enhances the ability to resist sudden disturbances. The minimum frequency curve and rate of change curve of Scheme 3 are generally improved compared with Scheme 2. In particular, the minimum frequency in periods 1-2 and 24 of Scheme 3 are lower than those of Scheme 2. This is because after considering the influence of load-side inertia, the inertia of the synchronous generator in Scheme 3 is smaller in these periods.

[0067]

[0068] Table 4 compares the wind power installed capacity and net wind farm revenue for three different schemes. Scheme 1 offers the highest wind power installed capacity and net wind farm revenue, but it fails to meet frequency safety constraints during certain periods. Therefore, if Scheme 1 is used for wind farm construction, the grid dispatching department will require wind farms to be curtailed to ensure grid stability after the wind farms are operational. Consequently, the actual net revenue for wind farm operators will not reach the optimized results of Scheme 1, leading to economic losses for the wind farm operators. Schemes 3 and 2 consider frequency safety constraints in their planning, ensuring sufficient frequency regulation capacity for the system while maximizing wind farm installed capacity. After the wind farms are operational, there is no risk of forced wind curtailment due to grid failure to meet frequency safety constraints. Compared with Scheme 2, Scheme 3 takes into account the load-side inertia that has been neglected. The system takes into account more inertia, has a stronger frequency response capability, a stronger ability to absorb wind energy, and a larger installed capacity, resulting in greater wind farm revenue. This proves the effectiveness and necessity of load-side inertia participating in frequency regulation from the perspective of increasing wind farm installed capacity and net revenue, and also provides the possibility and preliminary theoretical support for increasing the scale of wind power access.

[0069] 3) The impact of confidence level on optimization results: To investigate the impact of different confidence levels on wind power installed capacity and wind farm net revenue in the multiple uncertainty model of this invention, this invention uses Scheme 3 as an example for simulation analysis, and the comparison results are shown in Table 5.

[0070]

[0071] As the confidence level decreases, the installed wind power capacity and net wind farm revenue gradually increase. At a confidence level of α=0.65 compared to α=0.95, the installed wind power capacity increased by 31 MW, and the net wind farm revenue increased by $67 million. This is because the confidence level reflects the degree to which system safety is met. While a lower confidence level can increase installed wind power capacity and net wind farm revenue, it also increases the risk of not meeting system spinning reserve constraints and frequency safety constraints. In actual operation, an appropriate confidence level must be selected based on the specific circumstances.

[0072] This invention uses wind turbine type as an optimization variable and optimizes it in conjunction with the number of turbines. Based on dynamic frequency constraints, it further introduces the influence of load-side inertia and establishes a fuzzy-chance constraint-based wind power capacity optimization scheme with multiple uncertainties. This scheme can quickly select suitable wind turbine types, reducing investment costs for wind farm operators and enabling them to obtain greater investment returns. Furthermore, considering dynamic frequency constraints increases the system's ability to resist sudden disturbances, ensuring sufficient frequency regulation capability when the maximum wind power capacity is connected, guaranteeing that the system always operates within a safe frequency range, and ensuring stable operation after the wind farm is put into operation. When load-side inertia is considered in the dynamic frequency constraints, the system takes into account more inertia, resulting in stronger frequency response and a greater ability to absorb wind energy, thereby effectively improving wind power capacity and net wind farm revenue, and providing better guidance for the construction of large-scale wind farms.

Claims

1. A wind power installed capacity optimization method considering dynamic frequency constraints and multiple uncertainties, characterized in that... Includes the following steps: Step 1: Establish a dynamic frequency response model including the load-side inertia, and derive the expressions for the minimum frequency and the rate of frequency change; Step 2: Establish an overall optimization model based on the wind turbine type and number of wind turbines, and perform quadratic integer variable linearization processing; Step 3: Establish an installed capacity optimization model that considers multiple uncertainties and fuzzy opportunity constraints; Step 4: Using the maximization of wind farm revenue as the objective function, solve for the optimal wind power installed capacity; In step 2, the historical wind speed of the region for one year is obtained, measured every 5 minutes. The wind speed is processed to obtain the typical daily wind speed value. An overall optimization model for wind turbine type and number of wind turbines is established. Based on the linearization of quadratic integer variables, the overall optimization model for wind turbine type and number of wind turbines is transformed into a mixed integer programming model. The overall optimization model for wind turbine type and number of wind turbines is as follows: （6）； （7）； （8）； In the formula: This represents the total installed capacity of all wind turbine units connected to the grid. This represents the maximum number of installed units connected to the grid, and is a variable to be optimized. This represents a 0-1 type decision variable, used for screening wind turbine types; This represents the set of rated power for all fan types to be selected; Indicates the number of fan types; This indicates the rated power of the selected single fan; A number indicating the number of fan types; In summary, equation (6) is a nonlinear expression; a quadratic integer variable linearization method is adopted, and decision variables are introduced. The expression is as follows: （9）； （10）； Where: decision variables Characterization and The product; It is a relatively large constant; This indicates the result based on the values ​​of the decision variables; A set representing the number of fan types; This indicates the maximum number of wind turbines that can be connected to the grid.

2. The wind power installed capacity optimization method considering dynamic frequency constraints and multiple uncertainties according to claim 1, characterized in that: In step 1, the dynamic frequency response model containing the load-side inertia is shown in equations (1) to (5) below: The rotor motion equations for this model are expressed as follows: （1）； In the formula: The time constant of the synchronization inertia; The load-side inertia time constant; This refers to the system's frequency deviation. This refers to the active power adjustment amount for primary frequency regulation of the generating unit. The active power imbalance caused by the disturbance; The damping coefficient of the system; The results of the frequency minimum point derived from the dynamic frequency response model including load-side inertia are as follows: （2）； （3）； In the formula: Indicates the point of lowest frequency; This indicates the time corresponding to the point of lowest frequency. The initial frequency of the system; This is the system's reference frequency; This indicates the total number of thermal power units. The number indicating the number of thermal power units; Each conventional unit The expression is as follows: （4）； In the formula: For the first Start-up and shutdown status of the generator set; For the first The unit regulating power coefficient of a generator; For the first The response time constant of the generator. ; Indicates the start / stop status of the first generating unit; This represents the unit regulating power coefficient of the first generator; This represents the response time constant of the first generator; The expression for the rate of change of frequency is as follows: （5）； Indicates the rate of change of frequency; Indicates the frequency deviation of the system; Indicates the time of change in system frequency deviation; This represents the value of the frequency deviation when differentiated with respect to time.

3. The wind power installed capacity optimization method considering dynamic frequency constraints and multiple uncertainties according to claim 1, characterized in that: In step 3, since wind power, load, and load-side inertia all have uncertainties, the uncertainty of load-side inertia is analyzed, and fuzzy opportunity constraints are introduced. Based on the dynamic frequency constraints, a fuzzy opportunity constraint model with multiple uncertainties is constructed, and then the uncertainty constraints in the fuzzy opportunity constraint model are transformed into clear equivalence classes.

4. The wind power installed capacity optimization method considering dynamic frequency constraints and multiple uncertainties according to claim 3, characterized in that: When the confidence level of the fuzzy chance constraint At that time, it was shaped like , Represents the constraint function; The trapezoidal fuzzy parameters can be represented as follows: The clear equivalence class form can be expressed as: （11）； In equation (11): , This indicates the definition of a function; This represents the trapezoidal membership parameter. ; Represent real numbers; Indicates the number of possible values ​​in the set; Representation function Part of it.

5. The wind power installed capacity optimization method considering dynamic frequency constraints and multiple uncertainties according to claim 1, characterized in that: In step 4, considering the high nonlinearity of the dynamic frequency constraint model, the optimization problem is divided into a main problem and sub-problems. The optimal wind power installed capacity is solved with the objective function of maximizing wind farm revenue. The established expression for separating the main problem and subproblems is as follows: （12）； In the formula: The unit adjustable power corresponding to the failure to meet the dynamic frequency constraint; Representation system Unit regulating power at any given time; The expression for the net income of a wind farm is: （13）； In the formula: For the revenue from selling electricity from wind farms; The investment cost of a wind farm; This refers to the operation and maintenance costs of wind farms.

Images