Power grid and wind farm joint power dispatching method considering wind turbine governor time delay

Abstract

The application provides a grid-station combined power scheduling method considering the frequency modulation time delay of a wind turbine. The method comprises the following steps: a wind turbine frequency modulation response characteristic model considering the frequency modulation time delay is constructed; the frequency safety constraint is obtained and linearized by substituting into a system frequency evolution ordinary differential equation; a system level optimization scheduling model is established, taking the minimum operation cost as the target, containing the node power balance, thermal power operation, wind power output, DC power flow and frequency safety constraint, and the wind farm target output curve is obtained by solving; the station level scheduling takes the minimum deviation of the curve as the target, containing the wind turbine output, self-provisioning energy storage and frequency modulation standby constraint, and the station scheduling result is obtained by solving. The grid-station combined scheduling method provided by the application solves the frequency safety problem caused by the time delay of the wind turbine frequency modulation response, improves the safety of the system operation, is easy to implement, and has high engineering application value.

Description

Technical Field

[0001] This invention relates to the field of power grid-storage joint power dispatch technology, and in particular to a power grid-storage joint power dispatch method that takes into account the frequency regulation delay of wind turbine generators. Background Technology

[0002] With the large-scale integration of renewable energy sources, the installed capacity of wind power in the global power system continues to grow. Wind power, with its clean and low-carbon characteristics, plays a vital role in the energy transition. However, the intermittency and volatility of wind power pose unprecedented challenges to the safe and stable operation of the power system. Especially in regions with a high proportion of wind power connected to the grid, the frequency regulation capability of the power system faces severe tests. Currently, many countries have required wind power to have frequency response capabilities. Therefore, fully exploring the frequency regulation potential of wind turbines and improving the ability of wind power to participate in grid frequency regulation has become an important issue of common concern to the academic and engineering communities.

[0003] In recent years, wind turbines have initially acquired certain frequency regulation capabilities through improvements in main control strategies, energy storage assistance, and joint control with traditional units. However, compared with traditional synchronous generators, the frequency regulation characteristics of wind turbines differ significantly. On the one hand, wind turbines typically rely on power electronic interfaces to connect to the grid, limiting their inertial support and primary frequency regulation capabilities. On the other hand, in actual frequency regulation processes, wind turbine responses exhibit significant time delays. These delays mainly originate from measurement, communication, controller calculations, and active power regulation execution. Existing field studies show that frequency regulation delays vary considerably across different wind farms and turbine models, and these delays are influenced by multiple factors, including operating conditions, wind speed fluctuations, and control strategies. Field data from wind farms in multiple regions show that the power response delay after primary frequency regulation triggering can range from hundreds of milliseconds to several seconds. This delay directly affects the frequency regulation effect and the dynamic response of the system frequency. Test results from existing power system frequency response models that consider the frequency regulation delay of new energy power plants indicate that high delays reduce system inertia, worsen the system frequency change rate, and exacerbate system frequency deviations.

[0004] Existing optimization scheduling typically incorporates frequency security constraints to achieve frequency security control. However, current research has not yet considered the impact of wind turbine frequency regulation delay on frequency security constraints, and there is currently a lack of research solutions to reduce the impact caused by frequency regulation delay from the perspective of optimization scheduling. Summary of the Invention

[0005] The purpose of this invention is to overcome the aforementioned shortcomings of the prior art and provide a grid-plant joint power dispatching method and system that considers the frequency regulation delay of wind turbines. This invention derives a frequency safety constraint considering the frequency regulation delay of wind turbines, incorporates this constraint into system-level optimized dispatching, and achieves coordinated control of thermal and wind power through system-level optimized dispatching. While considering the frequency regulation delay of wind turbines, it ensures that the frequency meets the frequency safety constraint, thus improving system security. Then, the obtained wind farm output results are distributed to each wind farm, and the wind farm output deviation is minimized through plant-level dispatching.

[0006] The objective of this invention is achieved by at least one of the following technical solutions.

[0007] A grid-storage joint power dispatch method considering the frequency regulation delay of wind turbines includes the following steps: Step S1: Construct a frequency regulation response characteristic model of wind turbine considering frequency regulation delay; Step S2: Substitute the response characteristic model constructed in step S1 into the ordinary differential equation describing the frequency evolution of the power system to obtain the frequency security constraints considering the frequency regulation delay of wind turbine units. The frequency security constraints include the initial frequency change rate constraint, the maximum frequency difference constraint, and the steady-state frequency difference constraint. The maximum frequency difference constraint is linearized to obtain the linearized set of frequency security constraints. Step S3: Establish a system-station joint scheduling architecture, which includes a system-level scheduling layer and a station-level scheduling layer; wherein the system-level scheduling layer is used to realize the coordinated operation of thermal power units and wind farms, and the station-level scheduling layer is used to track the wind farm output curve obtained from the system-level scheduling results; Step S4: In the system-level scheduling layer, a system-level optimal scheduling model is established. The system-level optimal scheduling model aims to minimize the system operating cost. It uses the linearized set of frequency security constraints, node power balance constraints, thermal power unit operation constraints, wind power output constraints, and DC power flow constraints obtained in step S2 as constraints to solve the system-level optimal scheduling model and obtain the target output curves of each wind farm. Step S5: Send the target output curves of each wind farm obtained in step S4 to each wind farm dispatch substation. Step S6: In the station-level scheduling layer, establish and solve the station-level optimized scheduling model to track the target output curve of each wind farm issued in step S5 and minimize the actual output deviation of the wind farm. Solve to obtain the station-level scheduling result including wind farms and self-supplied energy storage. Step S7: Based on the site-level scheduling results obtained in Step S6, perform scheduling control on the wind farm and its self-contained energy storage.

[0008] Furthermore, the construction of the wind turbine frequency regulation response characteristic model considering the frequency regulation delay in step S1 specifically includes: adding a delay element to the traditional wind turbine primary frequency response characteristic model, and using a first-order Padé expansion to perform an equivalent transformation on the delay element.

[0009] Furthermore, in step S2, the maximum frequency difference constraint is linearized, specifically including: analyzing the functional monotonicity between the maximum frequency difference of the system frequency and the total inertia of the system, determining that the maximum frequency difference decreases as the total inertia of the system increases, thereby transforming the nonlinear constraint that the maximum frequency difference does not exceed the limit into a linear constraint that the total inertia of the system is not lower than the corresponding minimum inertia requirement.

[0010] Furthermore, the nonlinear constraint that the maximum frequency difference does not exceed the limit is transformed into a linear constraint that the total inertia of the system is not lower than the corresponding minimum inertia requirement. Specifically, this is achieved by introducing auxiliary variables and a preset constant M in the form of a mixed integer linear constraint. The constant M is a sufficiently large positive number whose value is greater than the upper limit of the absolute value of any relevant variable (such as the output of thermal power units, power flow of lines, energy storage power, etc.) under any actual physical constraint.

[0011] Furthermore, the system operating cost in the objective function of the system-level optimization scheduling model in step S4 includes: thermal power unit operating cost, wind power operating cost, energy curtailment cost, and frequency regulation cost; the constraints of the system-level optimization scheduling model also include node power balance constraints, thermal power unit operating constraints, wind power output constraints, and DC power flow constraints.

[0012] Furthermore, the operating constraints of the thermal power units include: upper and lower limits of thermal power unit output and ramp rate constraints; the DC power flow constraints include: constraints on the relationship between line transmission power and the phase angle of the starting and ending node voltages, upper limit constraints on line transmission power, and upper and lower limits constraints on the phase angle of the node voltages.

[0013] Furthermore, the constraints of the site-level optimized scheduling model in step S6 include: wind turbine output constraints, wind farm self-supplied energy storage power and energy constraints, and wind farm energy storage frequency regulation reserve constraints; the wind farm energy storage frequency regulation reserve constraints are used to limit the positive and negative reserve capacity of wind farm self-supplied energy storage at each time to meet the reserve requirements of wind farm participating in primary frequency regulation.

[0014] Furthermore, the self-contained energy storage power and energy constraints of the wind farm include: rated output constraints of energy storage, upper and lower limits constraints of energy storage state of charge, constraints on the relationship between energy storage charging and discharging power and state of charge change, and mutual exclusion constraints of energy storage charging and discharging states.

[0015] Furthermore, in both steps S4 and S6, a commercial solver is used to solve the model. The commercial solver is the Gurobi solver, and the modeling is performed using the YALMIP toolbox in the MATLAB environment.

[0016] The present invention also provides a grid-storage joint power dispatching system that considers the frequency regulation delay of wind turbine generators, comprising: The frequency regulation response characteristic modeling and frequency security constraint acquisition module is used to execute steps S1 and S2 in the method, construct a wind turbine frequency regulation response characteristic model considering frequency regulation time delay, and substitute it into the ordinary differential equation describing the frequency evolution of the system to obtain and output the linearized set of frequency security constraints. The system-level scheduling module is connected to the frequency regulation response characteristic modeling and frequency security constraint acquisition module. It is used to receive the linearized set of frequency security constraints, execute step S4 in the method, establish and solve the system-level optimized scheduling model, and output the target output curve of each wind farm. The site-level scheduling module is connected to the system-level scheduling module. It is used to receive the target output curves of each wind farm and execute step S6 in the method to establish and solve the site-level optimized scheduling model and output the site-level scheduling results including wind farms and self-contained energy storage. The execution control module is connected to the site-level scheduling module and is used to execute step S7 in the method, and to perform power scheduling control on the wind farm and self-contained energy storage according to the site-level scheduling result.

[0017] Furthermore, the frequency regulation response characteristic modeling and frequency safety constraint acquisition module is specifically used to: use a first-order Padé expansion to perform equivalent modeling of the delay element in the frequency regulation response of the wind turbine; and by determining the monotonically decreasing relationship between the maximum frequency difference of the system and the total inertia of the system, introduce auxiliary variables and a large M constant to transform the maximum frequency difference constraint into a mixed integer linear constraint on the total inertia of the system.

[0018] Compared with the prior art, the present invention has the following advantages and beneficial effects: (1) Improved the accuracy of frequency security assessment. This invention constructs a wind turbine frequency regulation response characteristic model that considers frequency regulation delay and substitutes it into the system frequency evolution differential equation. For the first time, it quantifies the impact of wind turbine frequency regulation delay on RoCoF, maximum frequency difference and steady-state frequency difference. This avoids the problem of overestimating wind power frequency regulation capability due to neglecting delay in traditional methods, and makes the calculation of frequency security constraints more consistent with the actual physical process.

[0019] (2) Efficient solution of nonlinear frequency safety constraints is achieved. This invention transforms the maximum frequency difference constraint into a linear constraint with respect to the total inertia of the system through monotonicity analysis, and introduces auxiliary variables to express it in the form of mixed integer linear constraints. This enables the nonlinear frequency safety constraints, which are originally difficult to embed directly into the optimization model, to be solved efficiently by commercial solvers, thus ensuring the computability and engineering applicability of the scheduling model.

[0020] (3) Improved system operation safety. Simulation results show that after adopting the method proposed in this invention, the maximum frequency difference of the system when facing load disturbances is always kept within the limit of 0.6Hz, the initial frequency change rate is significantly lower than that of the scheme without considering time delay, the total inertia level of the system is higher, and the anti-disturbance capability and frequency stability margin of the low inertia power grid are effectively enhanced.

[0021] (4) The system-site joint scheduling architecture provided by this invention achieves coordinated optimization of system-level economy and site-level tracking accuracy. At the system level, it ensures frequency security and economic operation, and at the site level, it achieves accurate tracking of the target output curve through flexible regulation of self-contained energy storage. It fully leverages the frequency regulation reserve value of energy storage resources and takes into account both global optimization and local controllability.

[0022] (5) Simple engineering implementation and strong applicability. The models built in this invention are all mixed integer linear programming models, which can be solved directly by calling mature commercial solvers. There is no need to develop special algorithms, and they are easy to deploy and apply in existing scheduling systems, which has high engineering application value. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating a grid-storage joint power dispatching method that considers the frequency regulation delay of wind turbine generators in one embodiment. Figure 2 This is a system topology diagram of one embodiment.

[0024] Figure 3 This is a diagram of the system-site level joint scheduling architecture in the embodiment.

[0025] Figure 4 This is a predicted power output curve of the wind farm in the embodiment.

[0026] Figure 5 This is a load forecast curve diagram from the embodiment.

[0027] Figure 6a The figure shows the simulation results of the peak-shaving output of the unit in Scheme 1 of the embodiment.

[0028] Figure 6b The figure shows the simulation results of the peak-shaving output of the unit in Scheme 2 of the embodiment.

[0029] Figure 6cThe figure shows the simulation results of the peak-shaving output of the unit in Scheme 3 of the embodiment.

[0030] Figure 7a The figure shows the simulation results of the frequency regulation output of each unit in Scheme 1 of the embodiment.

[0031] Figure 7b The figure shows the simulation results of the frequency regulation output of each unit in Scheme 2 of the embodiment.

[0032] Figure 8 This is a comparison chart of the maximum frequency difference of the system at each time point obtained from the scheduling results of each scheme in the embodiment.

[0033] Figure 9 The chart shows a comparison of the system RoCoF at different times obtained from the scheduling results of various schemes in the example.

[0034] Figure 10a The output curves of wind farm 1 and its self-contained energy storage are shown in the embodiment.

[0035] Figure 10b The output curves of wind farm 2 and its self-contained energy storage are shown in the embodiment.

[0036] Figure 10c The output curves of wind farm 3 and its self-contained energy storage are shown in the embodiment.

[0037] Figure 11 The diagram shows the reserve power curves for each wind farm in the example. Detailed Implementation

[0038] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0039] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and method steps have been omitted so as not to obscure the description of this application with unnecessary detail.

[0040] It should be understood that, when used in this specification, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements, and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components, and / or collections thereof. References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof all mean "including but not limited to," unless otherwise specifically emphasized.

[0041] like Figure 1 As shown, this embodiment of the invention provides a step-by-step flowchart of a grid-storage joint power dispatch method considering the frequency regulation delay of wind turbines, including: Step 1: Construct a frequency regulation response characteristic model for wind turbines that considers frequency regulation delay based on the traditional primary frequency response characteristics and response delay model of wind turbines; Step 2: Based on the response characteristic model established in Step 1, substitute it into the ordinary differential equation describing the frequency evolution of the system to obtain the frequency security constraints considering the frequency regulation delay of the wind turbine, including the initial frequency change rate ( Rate of Change of Frequency , RoCoF Constraints include maximum frequency deviation constraint and steady-state frequency deviation constraint, and nonlinear constraints are linearized. Step 3: Establish a system-plant joint scheduling architecture. System-level scheduling enables the coordinated operation of thermal power units and wind farms, while plant-level scheduling tracks the wind farm output curve obtained from the system scheduling results. Step 4: Establish a system-level optimization scheduling model. The system-level scheduling aims to minimize the system's operating cost. The constraints include node power balance constraints, thermal power unit operation constraints, wind power output constraints, DC power flow constraints, and frequency security constraints obtained in Step 2 that take into account the frequency regulation delay of wind turbine units. Step 5: Use the Gurobi commercial solver to solve the system-level scheduling model established in Step 4, obtain the scheduling results, and send the wind power output results to each wind farm scheduling substation. Step 6: Establish a wind farm-level optimization scheduling model. Based on the wind power output results obtained in Step 5, the wind farm-level scheduling aims to minimize the wind farm output deviation. The constraints include wind turbine output constraints, wind farm self-supplied energy storage power and energy constraints, and wind farm energy storage frequency regulation reserve constraints. Step 7: Use the Gurobi commercial solver to solve the site-level scheduling model established in Step 6 to obtain the scheduling results. Based on the final scheduling results, perform scheduling control on the wind farm and its self-contained energy storage.

[0042] As one embodiment, the specific method for constructing a wind turbine frequency regulation response characteristic model considering frequency regulation delay based on the traditional wind turbine primary frequency response characteristics and response delay model in step 1 is as follows: The traditional wind turbine primary frequency response process assumes that the wind power response power increases linearly and considers the influence of the frequency dead zone in the model, as shown in equation (1): (1); In the formula: t DB Dead time; T d This is the duration of a single frequency modulation response; R W This represents the maximum power of the wind turbine's primary frequency regulation response. For wind turbine units t The frequency modulation response power at a given moment; To accurately characterize the response delay of wind turbine units, a delay element is added to the above model. e -τs The first-order Padé expansion is used to approximate the delay element with higher accuracy. Therefore, the first-order Padé expansion is used to approximate the delay element, as shown in equation (2): (2); In the formula: s for t The frequency domain corresponding to that moment; G ( s This is the frequency domain form of the frequency regulation response characteristics of a wind turbine considering frequency regulation delay; F ( s This is the frequency domain form of the frequency regulation response characteristics of a wind turbine without considering frequency regulation delay. τ This is the frequency modulation delay.

[0043] Perform a Laplace transform on the original function: (3); In the formula: This is a Laplace transform.

[0044] Substituting equation (3) into equation (2), we get: (4); Performing an inverse Laplace transform on equation (4), we get: (5); In the formula: g ( t This is the time-domain representation of the frequency regulation response characteristics of a wind turbine considering frequency regulation delay.

[0045] Therefore, the frequency regulation response characteristic model of wind turbine considering frequency regulation delay is as follows: (6); As one embodiment, the frequency security constraint derivation in step 2 is specifically described as follows: The time evolution of frequency in a traditional system can be described by a first-order ordinary differential equation: (7); In the formula: M ( t )for t Total system inertia at any given moment; f 0 represents the system's rated frequency; for t Time-dependent system frequency deviation; D This is the load damping coefficient; P L ( t )for t Total system load at any time; for t Response power of conventional frequency regulation units at all times; for t Constant load disturbance.

[0046] The system frequency-time evolution equation considering the frequency regulation delay of wind turbines can be described as follows: (8); By solving equation (8), the system frequency safety index considering the frequency regulation delay of the wind turbine is obtained, including the initial frequency change rate ( Rate of Change of Frequency , RoCoF ), maximum frequency deviation and steady-state frequency deviation.

[0047] 1) RoCoF constraint Due to the existence of the frequency modulation dead zone, the initial response power of the frequency modulation unit is 0, and the system frequency deviation is also 0. RoCoF for: (9); In the formula: H ( t )for t Total inertia of the system at any given moment.

[0048] The total inertia of the system is calculated as follows: (10); In the formula:u g,t For thermal power units g exist t The start / stop status at any given moment; h g , h w , respectively, are the inertial time constants of thermal power units and wind power units; G and W are the total number of thermal power units and wind farms, respectively; S g , S w These represent the installed capacity of thermal power units and wind farms, respectively.

[0049] The initial frequency change rate of the system must not exceed the limit. The constraint on the initial frequency change rate of the system is as follows: (11); In the formula: RoCoF max This represents the maximum initial frequency change rate of the system.

[0050] 2) Maximum frequency deviation constraint When the rate of change of frequency is 0, the system frequency deviation reaches its extreme point, at which point the system frequency deviation is at its maximum. Substituting the specific function forms of wind turbine units and conventional thermal power frequency regulating units into equation (8), the specific function form of conventional thermal power frequency regulating units is consistent with equation (1), and will not be repeated here, we get: (12); in, ; ; ; ; t ’ This is the actual response time after removing the dead time. R G This represents the maximum power output of the primary frequency regulation response of the thermal power unit. α , β , γ All are response coefficients.

[0051] The expression for the absolute value of the system frequency deviation is obtained by integration: (13); in, ; ; K 1. K 2 is the calculation coefficient.

[0052] when t ≥ t DB season ,have to: (14); In the formula: t * This refers to the time when the system reaches its maximum frequency difference. .

[0053] Therefore, the maximum frequency difference is: (15); The maximum frequency difference of the system must not exceed the limit. The maximum frequency difference constraint is as follows: (16); In the formula: This is the maximum frequency difference limit for the system.

[0054] Since this constraint is a nonlinear constraint, it needs to be transformed. t * The value is determined by H ( t Therefore, equation (15) shows a certain... H ( t The function of ) is analyzed for monotonicity, when H ( t As the value increases, the maximum frequency difference of the system decreases, and the function is a monotonically decreasing function; therefore, a unique solution exists. H ( t * s ) makes =Δ f max If true, equation (16) can be transformed into: (17); In the formula: y g,t As an auxiliary variable; M For a sufficiently large number (also known as a large number) M (Constant, which can be 1000 in this embodiment). The constraint shown in Equation (17) is a mixed integer linear constraint, which can be solved efficiently by a commercial solver.

[0055] 3) Steady-state frequency deviation constraint When the system reaches a stable frequency, the rate of frequency change is 0. At this point, the system frequency deviation depends primarily on the response power of the frequency regulating unit, and the steady-state frequency difference is: (18); In the formula: t ss This represents the time when the system reaches its steady-state frequency difference.

[0056] The steady-state frequency deviation of the system must not exceed the limit. The steady-state frequency deviation constraint is as follows: (19); In the formula: This represents the maximum steady-state frequency difference.

[0057] As one embodiment, the specific method for establishing the system-site joint scheduling architecture in step 3 is as follows: To address the frequency stability challenges posed by a high proportion of wind power integration, constructing a system-level and site-level collaborative scheduling architecture has become a crucial direction for improving the frequency regulation efficiency of new energy sources and ensuring the safe operation of the power grid. In this embodiment, the architecture is divided into two scheduling levels: system-level scheduling and site-level scheduling. System-level scheduling is handled by the main grid dispatch center, aiming to achieve coordinated frequency control and economical operation of thermal and wind power while meeting frequency security constraints. Site-level scheduling is executed by the internal control system of the wind farm, primarily tasked with tracking the output curves set by the system based on frequency regulation requests issued by the system dispatch center, thereby improving the sensitivity and accuracy of frequency support response. The provided system-site joint scheduling architecture is as follows: Figure 3 As shown.

[0058] As one embodiment, the establishment of the system-level optimized scheduling model in step 4 is carried out at the system-level scheduling layer of the system-site joint scheduling architecture established in step 3, and the modeling method is as follows: The system-level optimized scheduling model proposed in this invention aims to achieve frequency-safe scheduling by coordinating thermal power units and wind power units. Frequency safety constraints considering the frequency regulation delay of wind power units are incorporated into the model, with the objective function being to minimize the system's operating cost, as shown below: (20); (twenty one); (twenty two); (twenty three); (twenty four); In the formula: F 1 represents the overall operating cost of the power grid; C 1 represents the operating cost of thermal power units; C 2 represents the operating cost of wind power; C 3 represents the cost of energy curtailment; C 4 represents the cost of frequency modulation; T The scheduling period; N This represents the total number of thermal power units. a i , b i , c ithermal power units i The consumption characteristic coefficients of the quadratic, linear, and constant terms; u gi,t For thermal power units i exist t The start / stop status at any given moment; S gi For thermal power units i Start-up and shutdown costs; N w The number of wind farms; σ w This represents the operation and maintenance cost coefficient for wind farms. For the system for wind farms w exist t Efforts should be made to achieve the desired goals at all times; c w For wind farm w Cost coefficient of wind curtailment; For wind farm w exist t Predicting output at any given moment; For scheduling step size; and thermal power units i and wind farm w Frequency modulation cost reserve factor; and thermal power units i and wind farm w exist t A frequency adjustment output at any given moment.

[0059] The system-level optimization scheduling model needs to satisfy the following constraints during the solution process: 1) Node power balance constraints (25); In the formula: and In the system respectively n Node No. t Peak-shaving output of thermal and wind power at all times; To and n The line connected to node number 1 is in t Transmission power at any given moment; and respectively with n The node is a set of lines consisting of a termination node and a start node. For the system n Node No. t The workload of the moment.

[0060] 2) Operating constraints of thermal power units (26); (27); In the formula: and The first i The lower limit of the thermal power unit; and The first i The maximum uphill and downhill speeds of the thermal power units.

[0061] 3) Wind power output constraints Wind power in t The output at any given time shall not exceed its predicted output: (28); 4) DC power flow constraint (29); In the formula: For the line L exist t Transmission power at any given moment; and The lines are respectively L exist t Voltage phase angle at the beginning and end of the time; X L For the line L The reactance; For the line L The upper limit of transmission power; For nodes n exist t Voltage phase angle at any given moment; and They are nodes n The upper limit of the voltage phase angle; the influence of voltage amplitude and reactive power is ignored in this embodiment.

[0062] 5) Frequency security constraints As shown in equations (9) to (19).

[0063] As one embodiment, the specific method of the system-level optimization scheduling solution process described in step 5 is as follows: In MATLAB, use the YALMIP toolbox or other software development platforms to build the system-level optimization scheduling model described in step 4, and use the Gurobi solver or other commercial optimization solving software to solve it, obtain the system-level optimization scheduling results, and extract the power output results of each wind farm.

[0064] The establishment of the station-level optimized scheduling model described in step 6 is carried out at the station-level scheduling layer of the system-station joint scheduling architecture established in step 3, and its modeling method is as follows: The wind farm tracks the target output issued by the system as accurately as possible, based on the target curve given by the system. (30); In the formula: For wind farm w exist t The actual output at any given moment.

[0065] The station-level optimal scheduling model needs to satisfy the following constraints during the solution process: 1) Fan output constraints (31); (32); In the formula: For wind farm w Self-supplied energy storage output; For wind farm w The charging and discharging states of self-contained energy storage; M Ew It is a sufficiently large positive number (e.g., 100).

[0066] 2) Wind farm self-contained energy storage capacity and energy constraints (33); (34); (35); (36); (37); In the formula: For wind farm w Self-contained energy storage rated output; , , Wind farm w Self-contained energy storage t SOC at time point, initial time point, and final time point; For wind farm w Self-contained energy storage rated capacity; η To improve the charging and discharging efficiency of energy storage; , Wind farm w Self-contained energy storage charging and discharging power; , Wind farm w Lower limit of self-contained energy storage SOC; , Wind farm w Self-contained energy storaget Positive and negative backup at any time.

[0067] 3) Constraints on frequency regulation backup for wind farm energy storage (38); In the formula: , Wind farm w Positive and negative primary frequency modulation limiting coefficients.

[0068] As one embodiment, the specific method for solving the station-level optimization scheduling in step 7 is as follows: In MATLAB, the YALMIP toolbox or other software development platforms are used to establish the site-level optimal scheduling model described in step 6. The model is then solved using the Gurobi solver or other commercial optimization software to obtain the site-level optimal scheduling results. The wind farm and its self-contained energy storage output are executed based on these results.

[0069] As one embodiment, a grid-storage joint power dispatch system considering the frequency regulation delay of wind turbines includes: The frequency regulation response characteristic modeling and frequency security constraint acquisition module is used to execute steps 1 and 2 of the method, construct a wind turbine frequency regulation response characteristic model considering frequency regulation time delay, and substitute it into the ordinary differential equation describing the frequency evolution of the system to obtain and output the linearized set of frequency security constraints. The system-level scheduling module is connected to the frequency regulation response characteristic modeling and frequency security constraint acquisition module. It is used to receive the linearized frequency security constraint set, execute step 4 of the method, establish and solve the system-level optimized scheduling model, and output the target output curve of each wind farm. The site-level scheduling module, connected to the system-level scheduling module, is used to receive the target output curves of each wind farm, execute step 6 of the method, establish and solve the site-level optimized scheduling model, and output the final scheduling instructions for each wind farm and its self-contained energy storage. The execution control module is connected to the site-level scheduling module and is used to execute step 7 of the method, performing power scheduling control on the wind farm and self-contained energy storage according to the final scheduling instruction.

[0070] The frequency regulation response characteristic modeling and frequency security constraint acquisition module uses a first-order Padé expansion to perform equivalent modeling of the delay element in the frequency regulation response of the wind turbine.

[0071] The frequency modulation response characteristic modeling and frequency safety constraint acquisition module determines the monotonically decreasing relationship between the system's maximum frequency deviation and total system inertia, and introduces an auxiliary variable and a sufficiently large number to transform the maximum frequency deviation constraint into a linear constraint on the system's total inertia. Both the system-level scheduling module and the site-level scheduling module solve their respective optimal scheduling models by calling the Gurobi solver.

[0072] The system-level scheduling module establishes and solves the system-level optimization scheduling model. The system operating cost in its objective function includes the operating cost of thermal power units, the operating cost of wind power, the cost of energy curtailment, and the cost of frequency regulation. Its constraints include node power balance constraints, thermal power unit operating constraints, wind power output constraints, DC power flow constraints, and a linearized set of frequency security constraints from the frequency regulation response characteristic modeling and frequency security constraint acquisition module. The site-level scheduling module establishes and solves a site-level optimal scheduling model, whose constraints include wind turbine output constraints, wind farm self-supplied energy storage power and energy constraints, and wind farm energy storage frequency regulation reserve constraints. The wind farm energy storage frequency regulation reserve constraints are used to limit the positive and negative reserve capacity of the wind farm's self-supplied energy storage at each time point. Both the system-level scheduling module and the site-level scheduling module solve their respective optimal scheduling models by calling the Gurobi solver.

[0073] The following is an application example of the method of the present invention, using an improved IEEE 39-node system for simulation analysis. The system topology diagram is shown below. Figure 2 As shown, it includes 6 thermal power units (G represents thermal power units) and 3 wind farms (W represents wind turbine units). Among the 39 nodes of the IEEE 39-node system, there are generator access points (such as nodes 30, 31, 36, 37, 38, and 39), wind farm access points (such as nodes 32, 33, and 34), and the remaining nodes are conventional transmission network intermediate nodes or load nodes. The system load damping coefficient is 1%, the maximum frequency change rate limit is 0.25 Hz / s, the maximum frequency difference limit is 0.6 Hz, the steady-state frequency difference limit is 0.1 Hz, the frequency regulation dead time and frequency regulation dead zone are 0.1 s and 0.033 Hz, respectively. The frequency regulation delay of wind turbines is set to 0.8 s, and the droop coefficient is 0.05. The droop coefficient of thermal power units is 0.042. The inertial time constants of thermal power units are 6.75 s, 6.4 s, 5.3 s, 5.3 s, 5.6 s, and 4.8 s, respectively. The inertial time constant of wind turbines is 8 s. The maximum load disturbance power is set to 5%. The predicted output curve of the wind farm is shown below. Figure 4 As shown, the load forecast curve is as follows: Figure 5 As shown; in this embodiment, the scheduling model is programmed using the YALMIP toolbox in MATLAB R2023a and solved using the Gurobi commercial solver.

[0074] To verify the effectiveness of the system-level scheduling model that considers the frequency regulation delay of wind turbines proposed in this invention, three schemes were set up for comparative analysis.

[0075] Option 1: The frequency safety constraint part of the model takes into account the influence of the frequency regulation delay of the wind turbine, which is the model proposed in this invention.

[0076] Option 2: The frequency safety constraint part of the model does not consider the impact of wind turbine frequency regulation delay.

[0077] Option 3: The model does not consider either the frequency regulation delay of the wind turbine or the frequency safety constraints.

[0078] The peak-shaving output of each unit obtained from the scheduling results is as follows: Figures 6a-6c As shown. By analyzing... Figures 6a-6c Simulation results analysis shows that, despite consistent predictions of total system load and wind power output across the three scheduling schemes, the different handling of frequency safety constraints leads to significant differences in the activation strategies of thermal power units. Scheme 1 indicates that, due to the response lag in wind turbine frequency regulation, the system tends to activate thermal power units with rapid response capabilities to ensure that frequency safety indicators such as RoCoF and maximum frequency difference do not exceed limits. Therefore, a larger number of thermal power units are activated, sacrificing some economic efficiency for frequency control stability and disturbance resistance. Scheme 2 results show that the scheduling model assumes immediate response to wind power frequency regulation, overestimating its frequency regulation capability. The system considers frequency regulation resources relatively abundant, thus reducing reliance on thermal power units, decreasing the number of activated units, and consequently lowering system peak-shaving costs. Scheme 3 ignores the system's frequency regulation safety requirements, focusing solely on minimizing costs. To further reduce operating costs, the system further reduces the number of activated thermal power units without affecting total output, resulting in economical operation, but with severely insufficient potential frequency regulation capability. When faced with wind power frequency regulation delays, the system will proactively introduce more conventional power sources with inertia and responsiveness to provide support in order to maintain safe frequency operation. Conversely, when the frequency regulation capability is "overestimated" or the constraints are removed, the system will excessively pursue economic efficiency, which may cause a potential shortage of frequency regulation resources and threaten system safety.

[0079] The frequency regulation output of each unit obtained from the scheduling results of Scheme 1 and Scheme 2 are as follows: Figure 7a and Figure 7b As shown. By analyzing... Figure 7a and Figure 7b Simulation results show that, compared to Scheme 2, Scheme 1 mobilizes more thermal power units to participate in frequency regulation support at certain times to make up for the frequency regulation capability gap caused by the frequency regulation delay of wind turbine units.

[0080] In terms of system frequency security comparison, Scheme 1 has the advantage in maximum system frequency difference compared to... RoCoFAll three schemes outperformed other schemes in performance. A comparison of the maximum system frequency difference at each time point obtained from the scheduling results of each scheme is shown in the figure below. Figure 8 As shown in Table 1, the system frequency safety index levels of each scheme are illustrated using average values. Scheme 1 sets frequency safety constraints considering the frequency regulation delay of wind turbines. The maximum deviation calculated by considering the maximum frequency difference is below 0.2Hz, and the average maximum deviation is also below 0.2Hz, which meets the maximum frequency difference limit constraint. RoCoF Scheme 2 has the lowest average level; Scheme 2 sets traditional frequency safety constraints, and the maximum deviation calculated by considering the maximum frequency difference exceeds 0.2Hz at some times, exceeding the maximum frequency difference limit. The average maximum frequency difference is close to 0.2Hz. From a physical point of view, although the control system detects the frequency drop, it lags in performing active power adjustment, which increases the maximum frequency difference; Scheme 3 does not set frequency safety constraints, and the system's maximum frequency difference exceeds 0.2Hz at most times. The average maximum frequency difference exceeds 0.2Hz, and the frequency exceeds the limit. Figure 9 Scheme 1 shows a significantly lower rate of frequency change after the initial disturbance than Schemes 2 and 3, and its fluctuation range is more controllable, demonstrating its stronger ability to suppress frequency abrupt changes. The system inertia calculated according to Equation (10) for the three schemes are 17720MWs, 16370MWs, and 14930MWs, respectively, which also proves that the system is the most stable among the scheduling schemes in Scheme 1.

[0081] Table 1 Comparison of Frequency Security Indicators

[0082] Scheme 1 enhances the defense against low-inertia disturbances by taking into account the frequency regulation delay characteristics of wind turbines during the dispatching process. Although the proposed strategy slightly increases operating costs, it significantly improves frequency security, prevents extreme control measures such as load shedding and turbine tripping caused by frequency instability, avoids greater potential economic losses, and enhances the resilience of the power system for long-term safe and stable operation. To quantitatively evaluate the overall performance of the scheme, Table 2 compares the key economic indicators of the three scenarios.

[0083] Under the premise of ensuring frequency security, the peak-shaving cost of Scheme 1 is 4% higher than that of Scheme 2 and 6% higher than that of Scheme 3. The frequency modulation cost of the proposed method is only 0.01% higher than that of Scheme 2, which is within an acceptable range for engineering practice. The resulting economic cost yields significant safety benefits: compared to Scheme 2, Scheme 1 increases the average maximum frequency deviation of the system by 0.092Hz and decreases the average RoCoF by 0.048Hz / s. Compared to Scheme 3, Scheme 1 increases the average maximum frequency deviation of the system by 0.107Hz and decreases the average RoCoF by 0.031Hz / s. Furthermore, Scheme 1 does not result in frequency exceeding limits.

[0084] Table 2 Economic Comparison

[0085] This embodiment includes three wind farms. Each wind farm performs its own power output target curve based on the scheduling results of Scheme 1, and the power output curves of each wind farm and its self-contained energy storage are obtained as follows: Figures 10a-10c As shown, the reserve power curves for each wind farm are as follows: Figure 11 As shown.

[0086] Depend on Figures 10a-10c The results show that the power output curves of each wind farm can effectively track the target power output issued by the system, and the flexible charging and discharging of energy storage can compensate for the impact of the fluctuation of wind power output. Figure 11 The results show that each wind farm has sufficient reserve power, providing strong support for frequency regulation needs.

[0087] Therefore, the method of the present invention can effectively solve the frequency security problem in power system dispatch and can realize the tracking of wind farm output to the system's instructions, which is feasible.

[0088] In summary, through simulation analysis of the improved IEEE 39-node model, the effectiveness of the grid-storage joint power dispatching method considering the frequency regulation delay of wind turbines provided by this invention has been verified.

[0089] The above embodiments are preferred embodiments of the present invention, but the embodiments of the present invention are not limited to the above embodiments. Any modifications, alterations, substitutions, combinations, or simplifications made without departing from the spirit and principle of the present invention shall be considered equivalent substitutions and shall be included within the protection scope of the present invention.

Claims

1. A grid-storage joint power dispatch method considering the frequency regulation delay of wind turbine generators, characterized in that, Includes the following steps: Step S1: Construct a frequency regulation response characteristic model of wind turbine considering frequency regulation delay; Step S2: Substitute the response characteristic model constructed in step S1 into the ordinary differential equation describing the frequency evolution of the power system to obtain the frequency security constraints considering the frequency regulation delay of wind turbine units. The frequency security constraints include the initial frequency change rate constraint, the maximum frequency difference constraint, and the steady-state frequency difference constraint. The maximum frequency difference constraint is linearized to obtain the linearized set of frequency security constraints. Step S3: Establish a system-station joint scheduling architecture, which includes a system-level scheduling layer and a station-level scheduling layer; wherein the system-level scheduling layer is used to realize the coordinated operation of thermal power units and wind farms, and the station-level scheduling layer is used to track the wind farm output curve obtained from the system-level scheduling results; Step S4: In the system-level scheduling layer, a system-level optimal scheduling model is established. The system-level optimal scheduling model aims to minimize the system operating cost. The linearized set of frequency security constraints, node power balance constraints, thermal power unit operation constraints, wind power output constraints, and DC power flow constraints obtained in step S2 are used as constraints. The system-level optimal scheduling model is solved to obtain the target output curves of each wind farm. Step S5: Send the target output curves of each wind farm obtained in step S4 to each wind farm dispatch substation. Step S6: In the station-level scheduling layer, establish and solve the station-level optimized scheduling model to track the target output curve of each wind farm issued in step S5 and minimize the actual output deviation of the wind farm. Solve to obtain the station-level scheduling result including wind farms and self-supplied energy storage. Step S7: Based on the site-level scheduling results obtained in Step S6, perform scheduling control on the wind farm and its self-contained energy storage.

2. The method according to claim 1, characterized in that, The construction of the wind turbine frequency regulation response characteristic model considering the frequency regulation delay in step S1 specifically includes: adding a delay element to the traditional wind turbine primary frequency response characteristic model, and using a first-order Padé expansion to perform an equivalent transformation on the delay element.

3. The method according to claim 1, characterized in that, In step S2, the maximum frequency difference constraint is linearized, specifically by analyzing the functional monotonicity between the maximum frequency difference of the system and the total inertia of the system, determining that the maximum frequency difference decreases as the total inertia of the system increases, thereby transforming the nonlinear constraint that the maximum frequency difference does not exceed the limit into a linear constraint that the total inertia of the system is not lower than the corresponding minimum inertia requirement.

4. The method according to claim 3, characterized in that, The nonlinear constraint that the maximum frequency difference does not exceed the limit is transformed into a linear constraint that the total inertia of the system is not lower than the corresponding minimum inertia requirement. Specifically, this is achieved by introducing auxiliary variables and a preset constant M in the form of a mixed integer linear constraint.

5. The method according to claim 1, characterized in that, The system operating cost in the objective function of the system-level optimization scheduling model in step S4 includes: thermal power unit operating cost, wind power operating cost, energy curtailment cost, and frequency regulation cost; the constraints of the system-level optimization scheduling model also include node power balance constraints, thermal power unit operating constraints, wind power output constraints, and DC power flow constraints.

6. The method according to claim 5, characterized in that, The operating constraints of the thermal power units include: upper and lower limits of thermal power unit output and ramp rate constraints; the DC power flow constraints include: constraints on the relationship between line transmission power and the phase angle of the starting and ending node voltages, upper limit constraints on line transmission power, and upper and lower limits constraints on the phase angle of the node voltages.

7. The method according to claim 1, characterized in that, The constraints of the site-level optimized scheduling model in step S6 include: wind turbine output constraints, wind farm self-supplied energy storage power and energy constraints, and wind farm energy storage frequency regulation reserve constraints; the wind farm energy storage frequency regulation reserve constraints are used to limit the positive and negative reserve capacity of wind farm self-supplied energy storage at each time to meet the reserve requirements of wind farm participating in primary frequency regulation.

8. The method according to claim 7, characterized in that, The self-contained energy storage power and energy constraints of the wind farm include: rated output constraints of energy storage, upper and lower limits constraints of energy storage state of charge, constraints on the relationship between energy storage charging and discharging power and state of charge change, and mutual exclusion constraints of energy storage charging and discharging states.

9. A grid-storage joint power dispatching system considering the frequency regulation delay of wind turbine generators for implementing the method of any one of claims 1 to 8, characterized in that, include: The frequency regulation response characteristic modeling and frequency security constraint acquisition module is used to execute steps S1 and S2 in the method, construct a wind turbine frequency regulation response characteristic model considering frequency regulation time delay, and substitute it into the ordinary differential equation describing the frequency evolution of the system to obtain and output the linearized set of frequency security constraints. The system-level scheduling module is connected to the frequency regulation response characteristic modeling and frequency security constraint acquisition module. It is used to receive the linearized frequency security constraint set and execute step S4 in the method as described in claim 1 to establish and solve the system-level optimized scheduling model and output the target output curve of each wind farm. The site-level scheduling module is connected to the system-level scheduling module and is used to receive the target output curves of each wind farm, and execute step S6 in the method as described in claim 1, to establish and solve the site-level optimized scheduling model, and output the site-level scheduling results including wind farms and self-contained energy storage. The execution control module is connected to the site-level scheduling module and is used to execute step S7 in the method, and to perform power scheduling control on the wind farm and self-contained energy storage according to the site-level scheduling result.

10. The system according to claim 9, characterized in that, The frequency regulation response characteristic modeling and frequency safety constraint acquisition module is specifically used to: use the first-order Padé expansion to perform equivalent modeling of the delay element in the frequency regulation response of the wind turbine; and by determining the monotonically decreasing relationship between the maximum frequency difference and the total inertia of the system, introduce auxiliary variables and a large M constant to transform the maximum frequency difference constraint into a mixed integer linear constraint on the total inertia of the system.

Images