A frequency modulation optimization method for wind farm and energy storage cooperation

Abstract

The application discloses a kind of wind farm and energy storage collaborative frequency modulation optimization method, belong to new energy grid-connected technology field, comprising: according to power system frequency safety requirement, frequency safety operation trajectory of safety and economy is planned simultaneously;When system disturbance occurs, in initial stage, priority is given to the rotor kinetic energy of wind turbine generator set to carry out fast virtual inertia response, to suppress frequency sudden drop;In middle and later period, then by multi-resolution smoothing control, fine adjustment is carried out to frequency fluctuation, the control strategy can reduce energy storage frequent action, prolong its service life;In the top layer of frequency modulation, the comprehensive cost function considering wind turbine opportunity cost, energy storage life loss and frequency out-of-limit penalty is established, and the standby frequency modulation power of wind turbine and energy storage is dynamically optimized.The application effectively improves the stability of power grid frequency, and realizes more economic frequency modulation process under the premise of guaranteeing frequency safety, and is suitable for high proportion renewable energy penetration power system.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of new energy grid-connected power generation, in particular to a power system frequency stability control method suitable for high proportion of renewable energy access, and more particularly to an optimal control strategy for realizing fast, accurate and economic support of power grid frequency by optimizing the coordinated cooperation of wind turbine generators and battery energy storage systems. BACKGROUND

[0002] The frequency of a power system is a key indicator of real-time balance between power generation and power consumption, and its stability is directly related to the safe and stable operation of the power grid. The traditional power system relies on the rotational inertia of synchronous generators and the primary frequency regulation function to respond to load fluctuations or generator tripping disturbances. When there is a power shortage, the rotational inertia of the generator will immediately release kinetic energy, reducing the frequency change rate, and then the governor will act to increase the output, stabilizing the frequency at a new lower level. However, with the advancement of the "double carbon" goal, the penetration rate of renewable energy such as wind power and photovoltaic power in the power grid continues to rise. These energies are mostly connected to the grid through power electronic converters, and their output is intermittent and random, and the converter cannot provide natural inertia response to the system like a synchronous machine when it is running at the maximum power point tracking mode. This results in a significant reduction in the equivalent inertia of the high proportion of new energy power system, and when the same power disturbance occurs, the system frequency change rate is greater and the frequency drop is more severe, posing a serious challenge to frequency stability.

[0003] In order to give wind power and other new energy stations the ability to regulate frequency, the main technical route at present is to reserve part of the active reserve for wind turbine generators (i.e. not running at the maximum power point), and to respond to system frequency changes through "virtual inertia control" and "droop control" that simulate the characteristics of synchronous machines. Although the wind turbine responds quickly and is suitable for providing initial inertia support, its frequency regulation capability is severely limited by the rotor speed range. Excessive release of rotor kinetic energy can cause the wind turbine speed to be too low and trigger protection shutdown, and when the wind turbine exits frequency regulation, its power output will return to the pre-disturbance level, which may cause "secondary frequency drop" of the system. In addition, reserving the reserve means sacrificing power generation income, and how to economically allocate reserve capacity is a major problem.

[0004] Battery energy storage system has the excellent characteristics of millisecond level fast response and power accurate controllable, and is an ideal frequency modulation resource. However, its cost is high, and the cycle life is limited, and frequent charging and discharging will accelerate the aging. If the energy storage independently undertakes all the frequency modulation tasks, the economy is poor, and it is not conducive to long-term operation. The ideal collaborative strategy should be that the fan plays the advantage of "fast" in the initial stage of frequency disturbance, and the energy storage plays the advantage of "stable" in the later stage. However, the existing strategy often fails to clearly distinguish the frequency modulation tasks of the two in different time scales, resulting in that the frequency modulation resources cannot be used most efficiently. The existing wind storage collaborative frequency modulation lacks the fine design of in-depth mining and complementary advantages of the characteristics of the two. In addition, many collaborative strategies fail to fully consider the power generation economy of the wind turbine and the life loss of the energy storage system when formulating the frequency modulation instruction, and lack a comprehensive cost-benefit evaluation model to guide the optimal allocation of frequency modulation power. This may lead to high frequency modulation cost, or sacrifice the frequency modulation effect in pursuit of economy. SUMMARY

[0005] The present application provides a wind farm and energy storage collaborative frequency modulation optimization method, which can allocate the frequency modulation responsibilities of the fan and the energy storage respectively after the system disturbance, and combine the online monitored system frequency, frequency change rate and other parameters to perform real-time update feedback control on the output of the wind turbine and the charging and discharging power of the energy storage, and consider the system frequency safety and economy.

[0006] In order to solve the above problems, the technical scheme adopted by the present application is as follows:

[0007] A wind farm and energy storage collaborative frequency modulation optimization method, comprising the following steps:

[0008] S1, according to the frequency deviation action threshold and the frequency change rate action threshold set by the power system, and combining the real-time operation parameters of the system, a frequency safety operation trajectory considering frequency safety and economy is planned;

[0009] S2, when the system is disturbed, the system frequency deviation and the frequency change rate are detected, and if any of them exceeds the preset starting threshold, the collaborative frequency modulation control of the wind farm and the energy storage system is started;

[0010] S3, in the initial stage of the collaborative frequency modulation control, the virtual inertia response of the wind turbine of the wind farm is preferred, the difference between the frequency safety operation trajectory and the actual system frequency is output through an adaptive virtual inertia control model to suppress the frequency drop, and the rotor speed safety constraint is applied to the wind turbine;

[0011] S4, after providing inertia support for the wind turbine generator set, providing mid-late frequency modulation power by the energy storage system, decomposing the frequency deviation signal into high-frequency component and low-frequency component by a multi-resolution smoothing control method, and implementing differentiated charging and discharging strategies based on the characteristics of the high-frequency component and the low-frequency component;

[0012] S5, during the entire frequency modulation process, dynamically optimizing and adjusting the standby frequency modulation power of the energy storage system and the wind turbine generator set based on the comprehensive cost function, the load prediction information and the wind power prediction information, so as to minimize the comprehensive frequency modulation cost under the premise of ensuring frequency safety.

[0013] Further, the method for planning the frequency safety operation trajectory in S1 comprises: based on the initial actual frequency and the initial frequency change rate when the system starts frequency modulation, combining the frequency deviation action threshold and the frequency change rate action threshold, and calculating the target frequency trajectory by an exponential function model, the target frequency trajectory starts from the initial actual frequency in an exponential form, approaches the frequency boundary defined by the frequency deviation action threshold, and at the same time ensures that the target frequency trajectory is always between the minimum and maximum frequency limit values allowed by the system.

[0014] Further, the preset starting threshold in S2 comprises a frequency deviation starting threshold and a frequency change rate starting threshold; the method for starting the coordinated frequency modulation control comprises: when the detected absolute value of the frequency deviation is greater than the frequency deviation starting threshold, or the absolute value of the frequency change rate is greater than the frequency change rate starting threshold, a frequency modulation starting flag bit is generated; a confirmation delay is set, and it is stipulated that only when the threshold exceeding state duration exceeds the confirmation delay, the coordinated frequency modulation control instruction is issued.

[0015] Further, the adaptive virtual inertia control model in S3 comprises: a virtual inertia item proportional to the change rate of the difference between the frequency safety operation trajectory and the actual system frequency, and a droop control item proportional to the difference; the adaptive virtual inertia control model calculates the additional power required to be output by the wind turbine generator set according to the weighted sum of the virtual inertia item and the droop control item, wherein the weight coefficient of the virtual inertia item is the equivalent virtual inertia damping coefficient, and the weight coefficient of the droop control item is the droop coefficient.

[0016] Further, the method for applying rotor speed safety constraints comprises: monitoring the rotor speed of the wind turbine generator set in real time, and correcting the additional power instruction when the additional power instruction causes the rotor speed to approach the preset minimum or maximum rotor speed limit value; when the rotor speed approaches the maximum value, the additional power is limited; when the rotor speed approaches the minimum value, the additional power is limited, and the size of the limitation is related to the square of the difference between the current rotor speed and the limit rotor speed, so as to prevent the wind turbine generator set from triggering protective shutdown due to rotor speed exceeding limit.

[0017] Further, the step of decomposing in S4 by the multi-resolution smoothing control method comprises: using a wavelet transform method to perform multi-layer decomposition on the frequency deviation signal to obtain a plurality of detail components representing high-frequency fluctuation details and an approximation component representing a low-frequency trend, so as to separate the change characteristics of different time scales in the frequency deviation signal.

[0018] Further, the method of implementing differentiated charging and discharging strategies based on the characteristics of the high-frequency component and the low-frequency component comprises: adaptively adjusting the weight coefficients of each component to calculate the frequency modulation power instruction of the energy storage system; the adjustment method of the weight coefficients of each component comprises: determining the corresponding weight according to the variance of each detail component and the approximation component, and giving a smaller weight to a component with greater fluctuation intensity and a larger weight to a component with smaller fluctuation intensity.

[0019] Further, the comprehensive cost function in S5 comprises: an opportunity cost generated by the wind turbine generator set deviating from the maximum power point tracking operation due to participation in frequency modulation, an operation cost generated by the energy storage system charging and discharging loss and auxiliary equipment operation, a life loss cost of the energy storage system due to cyclic charging and discharging, and a frequency limit penalty cost generated by the system frequency exceeding the safety limit.

[0020] Further, the determination method of the life loss cost of the energy storage system comprises: establishing a relationship model of the cycle life of the energy storage battery and the depth of discharge; using a rain flow counting method to analyze the energy storage power or state of charge change curve in a period of time, identifying complete charging and discharging cycle events and determining the depth of discharge corresponding to each cycle event; based on the relationship model and the Miner linear cumulative damage rule, the loss caused by each cycle event is accumulated, and the total life loss cost is quantified by combining the initial investment cost of the energy storage system.

[0021] Further, the frequency modulation optimization method further comprises an offline planning step for formulating an optimization strategy, which comprises: using a Monte Carlo simulation method to randomly generate a plurality of future operation scenarios covering different wind speed sequences, load changes and potential system faults; calculating the occurrence probability and the corresponding frequency modulation comprehensive cost of each scenario, and optimizing based on the expected cost of all scenarios to determine the energy storage capacity configuration, the wind turbine standby frequency modulation capacity and the initial state of charge range strategy, and taking the strategy as a constraint boundary of the dynamic optimization adjustment.

[0022] Compared with the prior art, the beneficial effects of the present application are:

[0023] (1) The application pre-plans the optimal frequency safe operation trajectory, changes the passive lag response to active target tracking control, so that the frequency modulation process is more smooth and orderly. Through the rapid virtual inertia response of the fan in the initial stage, the frequency drop is effectively suppressed, the frequency change rate is reduced, and the first line of defense of frequency safety is maintained. The fine adjustment of the energy storage system in the middle and late stages can effectively eliminate the steady-state frequency deviation and ensure the accurate recovery of the frequency to the rated value, thereby comprehensively improving the frequency stability and safety margin of the power grid when facing disturbances.

[0024] (2) The application clearly divides the frequency modulation responsibilities of the wind turbine and the energy storage system in different time scales. It takes full advantage of the fast response speed and low cost of the fan rotor to deal with the initial frequency mutation, and takes full advantage of the precise controllable and sustainable power output of the energy storage system to deal with subsequent frequency fluctuations. This phased and differentiated collaborative strategy avoids simple resource superposition and functional redundancy, and achieves synergistic efficiency.

[0025] (3) The application optimizes in both technical and economic aspects. On the one hand, through multi-resolution smooth control, the energy storage system intelligently ignores high-frequency and small frequency fluctuations, significantly reducing unnecessary charging and discharging times and depths, avoiding frequent deep cycling, thereby greatly extending the cycle life of the energy storage battery and reducing the life cycle cost. On the other hand, the comprehensive cost function introduced at the top level can dynamically balance the opportunity cost of fan curtailment, the operation and life loss cost of energy storage, and the frequency safety penalty of the power grid in real time, and find an economically optimal standby power distribution scheme, so that the entire frequency modulation process is the lowest cost under the premise of safety.

[0026] (24) The adaptive virtual inertia control model used in the application can dynamically adjust the output power according to the deviation and its rate of change between the planned trajectory and the actual frequency, and has good adaptability. In addition, the optimization strategy generated by offline Monte Carlo simulation provides a reference and constraint for online operation considering future uncertainties such as wind power prediction error and load fluctuation, so that the method can still maintain good control effect and economy in complex and variable power grid environment, and has strong robustness.

[0027] In order to make the above-mentioned purposes, features and advantages of the application more obvious and easy to understand, the following embodiments of the application are described in detail below, and the accompanying drawings are described as follows. BRIEF DESCRIPTION OF DRAWINGS

[0028] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed to be used in the embodiments. It should be understood that the following drawings only show some of the embodiments of the present application, and therefore should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can also be obtained without creative labor.

[0029] Figure 1 is a flowchart of the frequency modulation optimization method of the wind farm and energy storage coordination according to the present application. DETAILED DESCRIPTION

[0030] In order to make the purpose, technical solutions and advantages of the embodiments of the present application more clear, the following will combine the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are some of the embodiments of the present application, but not all the embodiments.

[0031] As shown in Figure 1 , the frequency modulation optimization method of the wind farm and energy storage coordination according to the present application, the core idea of which is to realize the unity of system frequency safety and frequency modulation economy through the closed-loop strategy of "active trajectory planning + phased coordinated response + full-cycle cost optimization". The specific steps are as follows:

[0032] (1) Frequency safety operation trajectory planning

[0033] The goal of this step is to change the passive frequency response to active trajectory planning, which provides an optimal target for subsequent frequency modulation control. According to the frequency deviation action threshold and the frequency change rate action threshold set by the power system, combined with the real-time operation parameters of the system, the frequency safety operation trajectory considering frequency safety and economy is planned.

[0034] Specifically, the method for planning the frequency safety operation trajectory includes: based on the initial actual frequency and the initial frequency change rate when the system starts frequency modulation, and combined with the frequency deviation action threshold and the frequency change rate action threshold, the target frequency trajectory is calculated by an exponential function model. The target frequency trajectory starts from the initial actual frequency in the form of an exponential function, approaches the frequency boundary defined by the frequency deviation action threshold, and at the same time ensures that the target frequency trajectory is always between the minimum and maximum frequency limit values allowed by the system. The frequency safety operation trajectory can be calculated by the following formula:

[0035]

[0036] Wherein: f FTP is the planned frequency safety operation trajectory; is the initial actual frequency when the system starts frequency modulation; is the initial frequency change rate when the system starts frequency modulation; is the frequency deviation action threshold set for the power system, defining the boundary of the frequency safety zone; is the frequency change rate action threshold set for the power system; t is time; sign(·) is the sign function, taking 1 when >0, -1 when and are the minimum and maximum frequency limit values allowed by the system, respectively.

[0037] The trajectory planned by the method can minimize the overall power output cost of the wind storage system under the premise of ensuring system frequency safety.

[0038] (2) Start of the coordinated frequency modulation control

[0039] To ensure that the frequency modulation resources can be reliably put into operation at critical moments, the application sets a high-sensitivity composite start criterion. When the system is disturbed, the system frequency deviation and the frequency change rate are detected, and if either of them exceeds the preset start threshold, the coordinated frequency modulation control of the wind farm and the energy storage system is started.

[0040] The preset start threshold includes a frequency deviation start threshold and a frequency change rate start threshold; the method for starting the coordinated frequency modulation control includes: when the absolute value of the detected frequency deviation is greater than the frequency deviation start threshold, or the absolute value of the frequency change rate is greater than the frequency change rate start threshold, a frequency modulation start flag is generated. The composite start criterion can be expressed as:

[0041]

[0042] Wherein: T trig is the frequency modulation start flag, and is 1 when it indicates start; Δ f is the real-time detected system frequency deviation; R is the real-time detected system frequency change rate; F act is the frequency deviation start threshold; R act is the frequency change rate start threshold.

[0043] In addition, to prevent false start caused by transient disturbance, the application sets a confirmation delay (ΔT delay ), and provides that only when the over-threshold state (Δ T trig=1) The coordinated frequency modulation control command is only issued after the duration exceeds the confirmation delay. Similarly, to prevent frequent starts and stops, a return criterion can be set, such as exiting the frequency modulation mode only when the frequency deviation and frequency change rate are both below 80% of their start threshold for a period of time.

[0044] (3) Initial stage of wind turbine virtual inertia response

[0045] In the initial stage of disturbance, using the rotational kinetic energy of the wind turbine rotor for rapid support is the most efficient method. In the initial stage of coordinated frequency regulation control, the wind turbine generators of the wind farm are given priority to perform virtual inertia response. Based on the difference between the safe operating frequency trajectory and the actual system frequency, the adaptive virtual inertia control model outputs additional power to suppress the sudden frequency drop and applies a safe rotor speed constraint to the wind turbine generators.

[0046] The adaptive virtual inertia control model includes: a virtual inertia term proportional to the rate of change of the difference between the safe operating trajectory and the actual system frequency, and a droop control term proportional to the difference; the adaptive virtual inertia control model calculates the additional power required to be output by the wind turbine generator based on the weighted sum of the virtual inertia term and the droop control term. This model is expressed as:

[0047]

[0048] Where: Δ f FTP The difference between the frequency safety operating trajectory and the actual system frequency; f This refers to the actual system frequency. P wind,FR Additional power command for the wind turbine output; J eff This is the equivalent virtual inertia damping coefficient, i.e., the weighting coefficient of the virtual inertia term; K opt This is the droop coefficient, which is the weighting coefficient of the droop control term.

[0049] To ensure wind turbine safety, the method for applying rotor speed safety constraints includes: real-time monitoring of the wind turbine generator's rotor speed; correcting the additional power command when it causes the speed to approach a preset minimum or maximum speed limit; limiting increased power generation when the speed approaches the maximum value; and limiting decreased power generation when the speed approaches the minimum value. The magnitude of these limits is related to the square of the difference between the current speed and the limiting speed, to prevent the wind turbine generator from triggering a protective shutdown due to speed exceeding the limit. The corrected power command... P wind This can be achieved using the following piecewise functions:

[0050]

[0051] wherein P1 and P2 are defined as:

[0052]

[0053] wherein: ω is the real-time rotor speed of the fan; ω max , ω min are the maximum and minimum limits of the rotor speed, respectively; P max , P min are the maximum and minimum power outputs allowed by the fan converter, respectively; P MPPT is the theoretical output power of the fan in the maximum power point tracking (MPPT) mode.

[0054] Fine-tuning regulation of the energy storage in the middle and late stages

[0055] After the inertia support provided by the wind turbine, the frequency fluctuation enters the middle and late stages, at which the energy storage system provides the power for frequency regulation in the middle and late stages. To prolong the service life of the energy storage and avoid its frequent response to high-frequency and small fluctuations, the present application decomposes the frequency deviation signal into high-frequency components and low-frequency components through a multi-resolution smoothing control method, and implements differentiated charging and discharging strategies based on the characteristics of the high-frequency components and the low-frequency components.

[0056] The step of decomposition through the multi-resolution smoothing control method includes: using a wavelet transform method to perform multi-layer decomposition on the frequency deviation signal to obtain a plurality of detail components representing high-frequency fluctuation details and an approximation component representing a low-frequency trend, so as to separate the change characteristics of different time scales in the frequency deviation signal. The decomposition process is represented as:

[0057]

[0058] wherein: d j (t) is the high-frequency fluctuation detail component of scale j; a J (t) is the low-frequency trend approximation component on scale J, which represents the low-frequency trend component (slow change, large deviation) in the frequency deviation.

[0059] The method for implementing differentiated charging and discharging strategies based on the characteristics of high-frequency components and low-frequency components comprises: adaptively adjusting the weight coefficients of each component to calculate the frequency modulation power instruction of the energy storage system; wherein the adjustment method of the weight coefficients of each component comprises: determining the corresponding weight of each detail component and the approximate component according to the variance of each detail component and the approximate component, and giving a smaller weight to a component with a larger fluctuation intensity, and giving a larger weight to a component with a smaller fluctuation intensity. Energy storage frequency modulation power instruction P ess,ref and the weight coefficient thereof K j , K a The calculation is as follows:

[0060]

[0061]

[0062] Wherein: VAR[·] represents the variance operation, reflecting the fluctuation intensity of the signal. The strategy reduces unnecessary charging and discharging actions and prolongs the service life of the energy storage by mainly responding to low-frequency and large-amplitude frequency deviations in a manner similar to low-pass filtering. At the same time, the calculated P ess,ref Power limiting and SOC (state of charge) management are performed to ensure that the energy storage system provides safe and sustainable frequency modulation services.

[0063] (5) Dynamic optimization of comprehensive frequency modulation cost in the whole cycle

[0064] At the top level of the whole frequency modulation process, the invention introduces economic considerations. Specifically, based on the comprehensive cost function, the standby frequency modulation power of the energy storage system and the wind turbine generator set is dynamically optimized and adjusted according to the load prediction information and the wind power prediction information in the whole frequency modulation process, so as to minimize the comprehensive frequency modulation cost under the premise of ensuring frequency safety.

[0065] Wherein, the comprehensive cost C total includes: the opportunity cost C wind of the wind turbine generator set deviating from the maximum power point tracking operation due to participation in frequency modulation, the operation cost C ess of the charging and discharging loss of the energy storage system and the operation of auxiliary equipment, the life loss cost C degradation of the energy storage system due to cyclic charging and discharging, and the frequency limit penalty cost C penalty due to the system frequency exceeding the safety limit.

[0066]

[0067] Opportunity cost of wind turbine C wind is the lost power generation revenue of wind turbine due to deviating from the maximum power point tracking (MPPT) for participating in frequency regulation, which can be expressed as:

[0068]

[0069] wherein is the theoretical maximum output power of wind turbine in MPPT mode, which is determined by wind speed. is the actual output power of wind turbine. is the on-grid price of wind power, and T is the length of optimization time domain.

[0070] Operating cost of energy storage C ess mainly includes power-related losses and auxiliary system energy consumption, which can be expressed as:

[0071]

[0072] wherein k loss is the power-related operating loss coefficient, k fixed is the fixed operating cost coefficient.

[0073] Lifetime loss cost of energy storage C degradation is the main cost of energy storage participating in frequency regulation, and its determination method includes:

[0074] 1) Establish the relationship model between the cycle life of energy storage battery and the depth of discharge (DOD)

[0075]

[0076] wherein a and b are constants related to the battery model, represents the cycle life of the battery at a certain DOD.

[0077] 2) Use rainflow counting method to analyze the energy storage power or state of charge change curve in a period of time, identify complete charge and discharge cycle events and determine the corresponding discharge depth of each cycle event .

[0078] 3) Based on the relationship model and Miner linear cumulative damage rule, the loss caused by each cycle event is accumulated, and combined with the initial investment cost of energy storage system, the total lifetime loss cost is quantified:

[0079]

[0080] wherein C cell is the initial investment cost of energy storage unit.

[0081] Frequency over-limit penalty cost C penalty For converting the frequency safety constraint into part of the optimization objective, to ensure that the optimization result prioritizes the frequency safety.

[0082]

[0083] Where λ is the penalty coefficient, to ensure that once the frequency over-limit, the cost function increases sharply, forcing the optimizer to avoid this situation. f limit For the frequency deviation safety limit.

[0084] Finally, to provide strategic guidance for online dynamic optimization, the present application also includes an offline planning step for formulating optimization strategies, which includes: using Monte Carlo simulation method, randomly generating multiple future operation scenarios Ss covering different wind speed sequences, load changes and potential system failures (type, size, time); for each scenario, calculate its occurrence probability p s And the corresponding frequency regulation comprehensive cost, and based on the expected cost of all scenarios, optimize to determine the energy storage capacity configuration, wind turbine standby frequency regulation capacity and initial state of charge (SOC) range, etc. Strategy, and take this strategy as the reference or constraint boundary of dynamic optimization adjustment.

[0085] The present application effectively solves the balance problem of frequency stability and frequency regulation economy in high proportion of new energy power system through the coordinated optimization of wind farm and energy storage system. The method takes frequency safety operation as the core constraint, and adaptively allocates the fast inertia support of wind turbine and the fine regulation capacity of energy storage in stages, and combines with the whole life cycle cost model for dynamic optimization, realizes the efficient and economic utilization of frequency regulation resources. The present application is suitable for the power grid scene with high proportion of renewable energy access, significantly improves the system operation economy and the service life of energy storage under the premise of guaranteeing frequency safety, and has strong engineering application value.

[0086] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. For those skilled in the art, the present application can have various modifications and changes. Any modification, equivalent replacement, improvement, etc. within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for frequency regulation optimization of a wind farm in cooperation with energy storage, characterized in that, The method comprises the following steps: S1, according to the frequency deviation action threshold and the frequency change rate action threshold set by the power system, and in combination with the real-time operation parameters of the system, a frequency safety operation trajectory is planned, which takes into account the frequency safety and economy; S2, when the system is disturbed, the system frequency deviation and the frequency change rate are detected, and if either of them exceeds the preset starting threshold, the collaborative frequency modulation control of the wind farm and the energy storage system is started; S3, in the initial stage of the collaborative frequency modulation control, the wind turbine generator set of the wind farm is preferentially used for virtual inertia response, the difference between the frequency safety operation trajectory and the actual system frequency is used to output additional power through an adaptive virtual inertia control model to suppress frequency drop, and a rotor speed safety constraint is applied to the wind turbine generator set; S4, after the wind turbine generator set provides inertia support, the energy storage system provides mid-late frequency modulation power, a wavelet transform method is used to decompose the frequency deviation signal into multiple detail components representing high-frequency fluctuation details and an approximation component representing low-frequency trend, so as to separate the change characteristics of different time scales in the frequency deviation signal; The weight coefficients of each component are adaptively adjusted to calculate the frequency modulation power instruction of the energy storage system; The adjustment method of the weight coefficients of each component comprises: determining the corresponding weight according to the variance of each detail component and the approximation component, and giving smaller weight to the component with larger fluctuation intensity and giving larger weight to the component with smaller fluctuation intensity; S5, in the whole frequency modulation process, based on the comprehensive cost function, the standby frequency modulation power of the energy storage system and the wind turbine generator set is dynamically optimized and adjusted according to the load prediction information and the wind power prediction information, so as to minimize the comprehensive frequency modulation cost under the premise of ensuring frequency safety.

2. The method of claim 1, wherein, The method for planning the frequency safety operation trajectory in S1 comprises: based on the initial actual frequency and the initial frequency change rate when the system starts frequency modulation, and in combination with the frequency deviation action threshold and the frequency change rate action threshold, a target frequency trajectory is calculated through an exponential function model, which starts from the initial actual frequency in the form of an exponential function and approaches the frequency boundary defined by the frequency deviation action threshold, while ensuring that the target frequency trajectory is always between the minimum and maximum frequency limit values allowed by the system.

3. The method of claim 1, wherein, The preset starting threshold in S2 comprises a frequency deviation starting threshold and a frequency change rate starting threshold; The method for starting the collaborative frequency modulation control comprises: when the absolute value of the detected frequency deviation is greater than the frequency deviation starting threshold, or the absolute value of the frequency change rate is greater than the frequency change rate starting threshold, a frequency modulation starting flag bit is generated; a confirmation delay is set, and it is stipulated that only when the duration of the over-threshold state exceeds the confirmation delay, the collaborative frequency modulation control instruction is issued.

4. The method of claim 3, wherein, The adaptive virtual inertia control model in S3 comprises a virtual inertia item proportional to the rate of change of the difference between the frequency safe operation trajectory and the actual system frequency, and a droop control item proportional to the difference; the adaptive virtual inertia control model calculates the additional power required to be output by the wind turbine generator set according to the weighted sum of the virtual inertia item and the droop control item, wherein the weight coefficient of the virtual inertia item is the equivalent virtual inertia damping coefficient, and the weight coefficient of the droop control item is the droop coefficient.

5. The method of claim 4, wherein, The method for applying the rotor speed safety constraint comprises: monitoring the rotor speed of the wind turbine generator set in real time, and correcting the additional power instruction when the rotor speed tends to approach the preset minimum or maximum rotor speed limit; when the rotor speed tends to approach the maximum value, limiting the additional power; when the rotor speed tends to approach the minimum value, limiting the additional power, and the size of the limitation is related to the square of the difference between the current rotor speed and the limit rotor speed, so as to prevent the wind turbine generator set from triggering protective shutdown due to the rotor speed exceeding the limit.

6. The method of claim 1, wherein, The comprehensive cost function in S5 comprises: the opportunity cost of the wind turbine generator set deviating from the maximum power point tracking operation due to participating in frequency modulation, the operation cost of the energy storage system and auxiliary equipment due to charging and discharging loss and auxiliary equipment operation, the life loss cost of the energy storage system due to cyclic charging and discharging, and the frequency limit penalty cost due to the system frequency exceeding the safety limit.

7. The method of claim 6, wherein, The method for determining the life loss cost of the energy storage system comprises: establishing a relationship model of the cycle life of the energy storage battery and the depth of discharge; using the rain flow counting method to analyze the energy storage power or state of charge change curve in a period of time, identifying complete charging and discharging cycle events and determining the corresponding depth of discharge of each cycle event; based on the relationship model and the Miner linear cumulative damage rule, the loss caused by each cycle event is accumulated, and the total life loss cost is quantified by combining the initial investment cost of the energy storage system.

8. The method of claim 6, wherein, It also comprises an offline planning step for formulating an optimization strategy, which comprises: using the Monte Carlo simulation method to randomly generate a plurality of future operation scenarios covering different wind speed sequences, load changes and potential system faults; calculating the occurrence probability and corresponding frequency modulation comprehensive cost of each scenario, and optimizing based on the expected cost of all scenarios to determine the energy storage capacity configuration, wind turbine standby frequency modulation capacity and initial state of charge range strategy, and taking the strategy as the constraint boundary of the dynamic optimization adjustment. The adaptive virtual inertia control model in S3 comprises a virtual inertia item proportional to the rate of change of the difference between the frequency safe operation trajectory and the actual system frequency, and a droop control item proportional to the difference; the adaptive virtual inertia control model calculates the additional power required to be output by the wind turbine generator set according to the weighted sum of the virtual inertia item and the droop control item, wherein the weight coefficient of the virtual inertia item is the equivalent virtual inertia damping coefficient, and the weight coefficient of the droop control item is the droop coefficient.

Images