A power leveling method and system for a port wind-storage combined system

Abstract

The present invention relates to the technical field of new energy grid-connected control, and discloses a power smoothing method and system for a port wind-storage combined system. The method comprises: dynamically adjusting the filter weight of the original wind power based on a first degree of fluctuation of the original wind power and a second degree of fluctuation of the predicted wind power; adjusting the filter ratio based on the filter weight, filtering the original wind power, and determining the power fluctuation component that needs to be smoothed by the wind-storage combined system; using an adaptive frequency-division empirical mode decomposition method containing a local mean and an optimization operator to decompose the power fluctuation component and determine multiple modal power components of different frequencies; controlling the capacitor based on the energy router of the wind-storage combined system to smooth the modal power component with a frequency greater than the frequency division frequency, and controlling the battery based on the energy router of the wind-storage combined system to smooth the modal power component with a frequency less than the frequency division frequency. The present invention can improve the overall performance of the energy storage system.

Description

Technical Field

[0001] The present disclosure relates to the technical field of new energy grid-connected control, and in particular to a power stabilization method and system for a port wind-storage combined system. Background Art

[0002] Offshore wind power generation in ports is characterized by high turbulence intensity and frequent sudden changes in wind speed. Wind turbine systems must ensure that their active power output meets certain requirements before they can be connected to the grid. However, during the grid connection process, wind power fluctuates significantly, posing a severe challenge to grid stability. Therefore, it is necessary to smooth the raw wind power output to ensure stable output.

[0003] In the related art, it is difficult to take into account both high-frequency and low-frequency fluctuations, so the system operating cost is high and the service life is short. Summary of the Invention

[0004] In order to solve the above technical problems, the present disclosure provides a power leveling method for a port wind-storage combined system, comprising the following steps:

[0005] Obtain a first degree of fluctuation of the original wind power and a second degree of fluctuation of the predicted wind power, and dynamically adjust the filtering weight of the original wind power based on the first degree of fluctuation and the second degree of fluctuation; adjust the filtering ratio based on the filtering weight, filter the original wind power, and determine the power fluctuation component that needs to be smoothed by the wind-storage combined system; use an adaptive frequency-division empirical mode decomposition method including a local mean and an optimization operator to decompose the power fluctuation component and determine multiple modal power components of different frequencies; determine the frequency division frequency, and control the capacitor based on the energy router of the wind-storage combined system to smooth the modal power component with a frequency greater than the frequency division frequency, and control the battery based on the energy router of the wind-storage combined system to smooth the modal power component with a frequency less than the frequency division frequency.

[0006] Furthermore, the dynamically adjusting the filtering weight of the wind power original power based on the first fluctuation degree and the second fluctuation degree includes:

[0007] Obtain a first grid-connected power obtained by filtering the original wind power based on a sliding average filter, and a second grid-connected power obtained by filtering the original wind power based on an anti-pulse interference average filter; calculate a standard deviation of the original wind power based on the first degree of fluctuation, and calculate a standard deviation of the predicted wind power based on the second degree of fluctuation; calculate a first weight of the first grid-connected power and a second weight of the second grid-connected power based on the standard deviation of the original wind power and the standard deviation of the predicted wind power.

[0008] Furthermore, after dynamically adjusting the filtering weight of the wind power original power based on the first fluctuation degree and the second fluctuation degree, the method further includes:

[0009] Based on the first grid-connected power and the first weight, the first grid-connected component to be connected to the wind-storage combined system is determined; based on the second grid-connected power and the second weight, the second grid-connected component to be connected to the wind-storage combined system is determined; the sum of the first grid-connected component and the second grid-connected component is determined as the grid-connected power of the wind-storage combined system.

[0010] Furthermore, the optimization operator includes a first optimization operator and a second optimization operator; the adaptive frequency-fractionation empirical mode decomposition method including a local mean and an optimization operator is used to decompose the power fluctuation component to determine multiple modal power components of different frequencies, including:

[0011] Determine the power fluctuation signal, signal-to-noise ratio and number of times white noise is added of the power fluctuation component; optimize the white noise added for the first time based on the first optimization operator and the signal-to-noise ratio to obtain a first white noise; use the local mean to estimate the sum of the first white noise and the power fluctuation signal to determine the local power mean; optimize the local power mean based on the second optimization operator to determine a first power residual, and decompose the power fluctuation component based on the local power mean and the first power residual to determine multiple modal power components of different frequencies.

[0012] Furthermore, the decomposing the power fluctuation component based on the local power mean and the first power residual to determine multiple modal power components of different frequencies includes:

[0013] The difference between the power fluctuation signal and the first power residual is determined as the first modal power component; the sum of the first power residual and the first white noise is optimized according to the second optimization operator to determine the second power residual, and the difference between the first power residual and the second power residual is determined as the second modal power component; the white noise added for the second time is optimized based on the first optimization operator and the signal-to-noise ratio to obtain the second white noise; the sum of the second power residual and the second white noise is optimized according to the second optimization operator to determine the third power residual, and the third modal power component is determined based on the difference between the second power residual and the third power residual; until the number of the modal power components is the same as the number of times the white noise is added, the decomposition of the power fluctuation component is stopped to determine multiple modal power components of different frequencies.

[0014] Furthermore, determining the frequency division frequency includes:

[0015] Extract the instantaneous frequency of each modal power component and determine the correlation between the instantaneous frequency and time; determine a frequency interval based on the maximum and minimum values ​​of the instantaneous frequency; determine a temporary crossover frequency within the frequency interval, and calculate the energy of the aliased modal power between adjacent modal power components based on the temporary crossover frequency and the correlation; determine the temporary crossover frequency corresponding to the minimum value of the energy as the allocation frequency.

[0016] Furthermore, determining a temporary crossover frequency within the frequency interval, and calculating the energy of aliased modal powers between adjacent modal power components based on the temporary crossover frequency and the association relationship, includes:

[0017] Based on a preset frequency interval, multiple temporary division frequencies are determined within the frequency interval, and the time corresponding to each of the temporary division frequencies is determined according to the association relationship; the modal power components corresponding to the adjacent instantaneous frequencies are determined according to the temporary division frequencies, and the aliased modal powers of the adjacent modal power components are determined; based on the time and the adjacent modal power components, the energy of the aliased modal power is calculated.

[0018] Furthermore, the wind-storage combined system energy router is composed of multiple DC-DC converters, multiple DC-AC converters and an intelligent energy storage controller; one end of the DC-DC converter is connected to the port wind power and the hybrid energy storage respectively to obtain the original wind power and the power output of the hybrid energy storage system, and the other end is connected in parallel to the DC bus;

[0019] One end of the DC-AC converter is connected in parallel to the DC bus, and the other end is connected in parallel to the power grid, for outputting the AC voltage after the port wind storage combined system smoothes the power; the intelligent energy storage controller is used to control the capacitor to smooth the modal power component with a frequency greater than the division frequency, and to control the battery to smooth the modal power component with a frequency less than the division frequency.

[0020] Furthermore, the DC-AC converter is a three-level DC-AC converter; the three-level DC-AC converter includes a DC part and an AC part; the DC part and the AC part are respectively composed of multiple transistor common-emitter diode units connected in series and parallel.

[0021] A power leveling system for a port wind-storage combined system, the system comprising:

[0022] An adjustment module is used to obtain a first degree of fluctuation of the original wind power and a second degree of fluctuation of the predicted wind power, and dynamically adjust the filtering weight of the original wind power based on the first degree of fluctuation and the second degree of fluctuation; a determination module is used to adjust the filtering ratio based on the filtering weight, filter the original wind power, and determine the power fluctuation component that needs to be smoothed by the wind-storage combined system; an adaptive frequency-division empirical mode decomposition method including a local mean and an optimization operator is used to decompose the power fluctuation component to determine multiple modal power components of different frequencies; a smoothing module is used to determine a frequency division frequency, and control the capacitor based on the energy router of the wind-storage combined system to smooth the modal power component with a frequency greater than the frequency division frequency, and control the battery based on the energy router of the wind-storage combined system to smooth the modal power component with a frequency less than the frequency division frequency.

[0023] The embodiments of the present disclosure have the following technical effects:

[0024] The power smoothing method of the port wind-storage combined system provided by the present disclosure first obtains the first fluctuation degree of the original wind power and the second fluctuation degree of the predicted wind power, and dynamically adjusts the filter weight of the original wind power based on the first fluctuation degree and the second fluctuation degree, adjusts the filter ratio based on the filter weight, filters the original wind power, and determines the power fluctuation component that needs to be smoothed by the wind-storage combined system. Through the dynamic adjustment of the filter weight, the advantages of different filters are combined, and the possible time delay problem and the limitations of the smoothing effect are effectively solved, so as to achieve a more accurate and smooth filtering effect and reduce the system risk caused by power fluctuation. Then, an adaptive frequency division empirical mode decomposition method including a local mean and an optimization operator is used to decompose the power fluctuation component and determine multiple modal power components of different frequencies. The present disclosure decomposes the power fluctuation component into multiple modal power components of multiple different frequencies, and better and reasonably distributes the power fluctuation component, thereby improving the operating efficiency and stability of the wind farm. Finally, the crossover frequency is determined, and the energy router of the wind-storage combined system controls the capacitor to smooth the modal power components with frequencies greater than the crossover frequency. The energy router of the wind-storage combined system controls the battery to smooth the modal power components with frequencies less than the crossover frequency. It can be understood that by allocating signals of different frequency bands to the most suitable energy storage device, the overall performance of the energy storage system can be improved, the service life of the equipment can be extended, the output stability of the wind-storage combined system can be improved, and the operating cost of the system can be reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0026] Figure 1 This is a flow chart of a power leveling method for a port wind-storage combined system provided by an embodiment of the present disclosure;

[0027] Figure 2 is a flow chart of a wind power prediction method provided by an embodiment of the present disclosure;

[0028] Figure 3 is a flow chart of a method for decomposing power fluctuation components provided by an embodiment of the present disclosure;

[0029] Figure 4 is a flow chart of a method for smoothing power fluctuation components provided by an embodiment of the present disclosure;

[0030] Figure 5 This is a topological diagram of an energy router provided by an embodiment of the present disclosure;

[0031] Figure 6 This is a topological diagram of a three-level converter of an energy router of a port wind-storage combined system provided by an embodiment of the present disclosure;

[0032] Figure 7 is a structural diagram of a wind-storage combined system provided by an embodiment of the present disclosure;

[0033] Figure 8 This is a schematic diagram of wind power prediction provided by an embodiment of the present disclosure;

[0034] Figure 9 This is a diagram showing the effect of using dynamic weight filtering to smooth out wind power fluctuations provided by an embodiment of the present disclosure;

[0035] Figure 10 This is a schematic diagram of the power that needs to be smoothed by the hybrid energy storage system obtained by dynamic weight filtering provided by an embodiment of the present disclosure;

[0036] Figures 11(a)-(d) are diagrams of modal power components and their spectra provided by embodiments of the present disclosure;

[0037] Figure 12 is an instantaneous frequency-time curve of a modal power component provided by an embodiment of the present disclosure;

[0038] Figure 13 This is a diagram showing the effect of smoothing wind power fluctuations provided by an embodiment of the present disclosure;

[0039] Figure 14 is a schematic diagram of the operating conditions of the energy storage system provided by an embodiment of the present disclosure;

[0040] Figure 15 It is a schematic diagram of the power leveling system of the port wind-storage combined system provided by an embodiment of the present disclosure. DETAILED DESCRIPTION

[0041] To make the objectives, technical solutions, and advantages of the present invention more clear, the technical solutions of the present invention are described clearly and completely below. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. All other embodiments derived by persons of ordinary skill in the art based on the embodiments of the present invention without inventive effort are considered to be within the scope of the present invention.

[0042] Offshore wind power generation in ports is characterized by high turbulence intensity and frequent sudden changes in wind speed. Wind turbine systems must ensure that their active power output meets certain requirements before they can be connected to the grid. During the grid connection process, wind power fluctuates significantly, posing a severe challenge to grid stability. Therefore, it is necessary to smooth the raw wind power output to ensure stable output.

[0043] This application aims to quantify the smoothing effect of wind power fluctuations, and proposes a power smoothing method and system for a port wind-storage combined system. First, the filter ratio is adjusted using a determined filter weight to filter and adjust the original wind power. Subsequently, the adaptive frequency-division empirical mode decomposition method is used to decompose the power fluctuation components that need to be smoothed by the wind-storage combined system, and the power fluctuation components are decomposed into modal functions of different frequencies. The instantaneous frequency of the modal function is further extracted through the Hilbert transform, and the relationship between the instantaneous frequency and time is obtained. Finally, a power distribution smoothing control strategy and energy control equipment for a hybrid energy storage system are proposed, so that supercapacitors can better bear high-frequency fluctuations, while batteries are responsible for low-frequency fluctuations, thereby ensuring that the energy storage system can effectively respond to various power fluctuations, obtain power distribution for different energy storage devices, and achieve effective smoothing of wind power fluctuations.

[0044] Figure 1 This is a flow chart of the power leveling method for the port wind and storage combined system provided by the embodiment of the present disclosure. Figure 1 , specifically including the following steps:

[0045] In step S11 , a first fluctuation degree of the wind power original power and a second fluctuation degree of the wind power predicted power are obtained, and a filtering weight of the wind power original power is dynamically adjusted based on the first fluctuation degree and the second fluctuation degree.

[0046] In the embodiment of the present disclosure, a wind-storage combined system can be determined to be established. Generally, the wind-storage combined system consists of a wind turbine, a supercapacitor, a battery pack, a DC-DC converter, a DC-AC inverter and a transformer.

[0047] This disclosure obtains raw wind power from a wind turbine and determines the degree of fluctuation of the raw wind power. It also obtains predicted wind power based on a pre-built wind power prediction model and determines the degree of fluctuation of the predicted wind power. For ease of distinction, this disclosure defines the degree of fluctuation of the raw wind power as the first degree of fluctuation, and the degree of fluctuation of the predicted wind power as the second degree of fluctuation.

[0048] The degree of fluctuation includes the frequency and amplitude of fluctuation, which are used to determine the standard deviation of the wind power grid-connected power and its predicted wind power. The standard deviation of the wind power grid-connected power and its predicted wind power is used to determine the filtering weight of the wind power original power.

[0049] The filtering weight is used to adjust the filtering ratio of the original wind power. The filtering includes sliding average filtering and anti-pulse interference filtering.

[0050] In step S12, the filtering ratio is adjusted based on the filtering weight, the original wind power is filtered, and the power fluctuation component that needs to be smoothed by the wind-storage combined system is determined.

[0051] In the present disclosure, the original wind power is filtered by using sliding average filtering and anti-pulse interference filtering according to the filtering ratio, and the original wind power can be decomposed into two components: the grid-connected wind volume that can be directly incorporated into the wind power energy storage system and the power fluctuation component that needs to be smoothed by the wind-storage combined system.

[0052] In step S13, an adaptive frequency-fractional empirical mode decomposition method including a local mean and an optimization operator is used to decompose the power fluctuation component to determine multiple modal power components of different frequencies.

[0053] In the embodiment of the present disclosure, it is proposed to improve the adaptive frequency division empirical mode decomposition method by calculating the local mean and optimizing the operator, so as to decompose the power fluctuation component based on the improved adaptive frequency division empirical mode decomposition method and determine multiple modal power components of different frequencies.

[0054] Among them, the optimization operator can be a mathematical tool or function used to improve the defects of traditional noise-assisted decomposition methods, reduce modal aliasing and noise residue, and thus improve the accuracy and stability of signal decomposition.

[0055] In step S14, the crossover frequency is determined, and the capacitors of the wind-storage combined system are used to smooth the modal power components with frequencies greater than the crossover frequency, and the batteries of the wind-storage combined system are used to smooth the modal power components with frequencies less than the crossover frequency.

[0056] In the embodiment of the present disclosure, a suitable crossover frequency can be determined based on the instantaneous frequency of the modal power component, so that the capacitor of the wind-storage combined system can be used to smooth the modal power component with a frequency greater than the crossover frequency, and the battery of the wind-storage combined system can be used to smooth the modal power component with a frequency less than the crossover frequency.

[0057] The present disclosure allows adjusting the filter ratio based on the determined filter weights, filtering the raw wind power, and determining the power fluctuation components that require smoothing by the wind-storage combined system, thereby reducing unnecessary power output. This helps reduce the configuration capacity of the energy storage system, saves investment costs, and improves the energy utilization efficiency of the energy storage system. Furthermore, an adaptive frequency-fractionation empirical mode decomposition method, which includes a local mean and optimization operator, is used to decompose the power fluctuation components and determine multiple modal power components of different frequencies. This allows for better and more rational allocation of the power fluctuation components, thereby improving the operating efficiency and stability of the wind farm. A crossover frequency is determined, and the capacitors of the wind-storage combined system are used to smooth the modal power components with frequencies greater than the crossover frequency, while the batteries of the wind-storage combined system are used to smooth the modal power components with frequencies less than the crossover frequency. This can be understood as follows: by allocating signals of different frequency bands to the most suitable energy storage devices, the overall performance of the energy storage system can be improved, the equipment life can be extended, the output stability of the wind-storage combined system can be improved, and the operating costs of the system can be reduced. Furthermore, the power fluctuation components can be better and more rationally allocated, thereby improving the operating efficiency and stability of the wind farm.

[0058] The embodiment of the present disclosure can construct a wind power prediction model based on the extreme learning machine model, use the wind power prediction model to predict wind power, and thereby determine the weight of the dynamic weight filtering according to the predicted wind power and the original wind power.

[0059] Figure 2 This is a flow chart of the wind power prediction method provided by the embodiment of the present disclosure. Figure 2 , wherein, a wind power prediction model is constructed based on the Extreme Learning Machine (ELM) model, and the implementation method for predicting wind power is as follows:

[0060] First, the ELM needs to be trained. This requires obtaining a historical wind power dataset. This dataset includes environmental factors such as wind speed, wind direction, humidity, and temperature, which are input to the ELM model, along with the corresponding wind power output. This dataset is then divided into a training dataset and a test dataset in a 9:1 ratio.

[0061] The wind power historical data set is normalized using the normalization formula shown below.

[0062]

[0063] Where, g To normalize the historical data set of wind power; For the i Historical data of wind power is the minimum value in the historical wind power data set; is the maximum value in the historical wind power data set.

[0064] The purpose of data normalization is to eliminate the unit differences between different types of data, thereby improving the efficiency of model learning.

[0065] After that, ELM parameters can be configured, and ELM training can begin. This disclosure selects 90% of the historical wind power data set as the training dataset. Through continuous iterative training, the extreme learning machine model is judged to see if it meets the pre-set performance criteria. When the model meets the pre-set criteria, the ELM test is performed. If not, model training continues.

[0066] Furthermore, the ELM can be tested based on the divided test data set to obtain the predicted value of wind power. The predicted value is used to verify the prediction performance of the ELM, and the accuracy of the model is evaluated by calculating the error between the predicted value and the actual value.

[0067] After the ELM training is completed, a wind power prediction model that can be used to predict wind power is finally obtained, and the wind power can be predicted based on the wind power prediction model.

[0068] The embodiment of the present disclosure can obtain the first grid-connected power by filtering the original wind power based on the sliding average filter. Furthermore, the smoothing average filter uses a specific sliding window to smooth the data within a certain period of time, and represents any data point therein by calculating the arithmetic mean of adjacent data in the window. As the output data changes, the sliding window will be updated accordingly, and the characteristic of the fixed window size makes it possible to recalculate the average value of a set of data each time it slides. Therefore, the sliding average filtering technology can effectively represent the value of the center point of the time window. The sliding average filter smoothes the time series data by calculating the average value of the data points in the sliding window. The window moves with time, one data point at a time, and the corresponding average value is updated at the same time. Among them, the calculation formula for determining the grid-connected power by the sliding average filter is as follows:

[0069]

[0070] Where, N is the time step; is the first grid-connected power after sliding average filtering; is the original power of wind power, t For time.

[0071] A second grid-connected power can also be obtained by filtering the original wind power using pulse-interference-resistant average filtering, also known as median average filtering. Furthermore, pulse-interference-resistant average filtering first obtains the original wind power, selects M points from the original wind power at a time, and removes the maximum and minimum values ​​to reduce the impact of occasional pulse interference. The filtered data is then arithmetic averaged to obtain a relatively smooth set of output data. This method effectively reduces the noise component in the wind power signal and improves the accuracy and reliability of wind power output data.

[0072] The calculation formula for determining the grid-connected power by anti-pulse interference average filtering is as follows:

[0073]

[0074] Where, M is the number of data points intercepted from the original wind power series, To trace back from the current time t M The original wind power value at the sampling point is To backtrack from the current time t M +1 wind power original power value at sampling point is the second grid-connected power after average filtering for anti-pulse interference; is the maximum value of the original wind power, It is the minimum value of the original wind power.

[0075] Therefore, the present disclosure can obtain the original wind power and the predicted wind power, as well as the first grid-connected power obtained by filtering the original wind power based on the sliding average filter and the second grid-connected power obtained by filtering the original wind power based on the anti-pulse interference average filter, and calculate the standard deviation of the original wind power and the standard deviation of the predicted wind power.

[0076] The calculation formula for the standard deviation of wind power original power is as follows:

[0077]

[0078] Where, is the standard deviation of the original wind power, n is the number of wind power data; is the average value of wind power grid-connected power. When calculating the standard deviation of wind power original power, is the wind power grid-connected power, and the standard deviation of wind power prediction power is calculated. Forecast power for wind power.

[0079] Furthermore, a first weight of the first grid-connected power and a second weight of the second grid-connected power may be calculated according to the calculated standard deviation of the original wind power and the calculated standard deviation of the predicted wind power.

[0080] In the present disclosure, the calculation of the first weight of the grid-connected power determined by the sliding average filter is taken as an example. The calculation method of the first weight is shown in the following formula.

[0081]

[0082] Where, is the standard deviation of wind power prediction, θ is the first weight, h 1 is the upper limit of the standard deviation of wind power, h 2 is the lower limit of the standard deviation of wind power, .

[0083] The sum of the first weight and the second weight is 1.

[0084] In the present disclosure, a first grid-connected component to be incorporated into the wind-storage combined system is determined based on the first grid-connected power and the first weight; a second grid-connected component to be incorporated into the wind-storage combined system is determined based on the second grid-connected power and the second weight; and the sum of the first grid-connected component and the second grid-connected component is determined as the grid-connected power of the wind-storage combined system. The grid-connected power of the wind power directly incorporated into the storage combined system is as follows:

[0085]

[0086] Where, is the grid-connected power. is the first grid-connected power, is the second grid-connected power.

[0087] By using the filtering method proposed in this application that dynamically adjusts the hybrid filtering ratio according to the filtering weight, the original wind power power is adjusted, and the grid-connected power of the wind-storage combined system is determined, the unnecessary output of the energy storage system can be reduced. The reduced unnecessary output not only helps to reduce the configuration capacity of the energy storage system, thereby saving investment costs, but also improves the energy utilization efficiency of the energy storage system. In addition, the application of dynamically adjusting the hybrid filtering ratio can also enhance the stability and reliability of the wind power system and reduce the system risk caused by power fluctuations. It is of great significance for optimizing the operation of the wind-storage combined system and improving the economic and technical benefits of the energy storage system.

[0088] In the present disclosure, after determining the power fluctuation component, the adaptive frequency division empirical mode decomposition method is further improved based on the local mean and optimization operator, so that the power fluctuation component is decomposed according to the improved adaptive frequency division empirical mode decomposition method, and its implementation method is as shown in the following embodiment.

[0089] Figure 3 This is a flow chart of the power fluctuation component decomposition method provided by the embodiment of the present disclosure. Figure 3 First, it is necessary to determine the power fluctuation signal, signal-to-noise ratio, and the number of times white noise is added of the power fluctuation component. The added white noise is further optimized according to the optimization operator and the signal-to-noise ratio. The empirical mode decomposition method is then used to calculate and estimate the sum of the first white noise and the power fluctuation signal to determine the local power mean.

[0090] The calculation formula of the local power mean is as follows:

[0091]

[0092] Where, For the i The local power mean of the power fluctuation signal, is the power fluctuation signal, is the signal-to-noise ratio, For the i White noise is added, is the first optimization operator, t For time.

[0093] Among them, the signal-to-noise ratio The calculation formula is as follows:

[0094]

[0095] Where, is the inverse of the signal-to-noise ratio.

[0096] Finally, the local power mean is optimized based on the second optimization operator to determine the first power residual, and the power fluctuation component is decomposed based on the local power mean and the first power residual to determine multiple modal power components of different frequencies.

[0097] The calculation formula of the first power residual is as follows:

[0098]

[0099] Where, r 1 is the first power residual, N is the second optimization operator.

[0100] For ease of description, the present disclosure determines the difference between the power fluctuation signal and the first power residual as the first modal power component, wherein the calculation formula of the first modal power component is as follows:

[0101]

[0102] Where IMF1 is the first modal power component.

[0103] Furthermore, the sum of the first power residual and the first white noise is optimized according to the second optimization operator to determine the second power residual, and the difference between the first power residual and the second power residual is determined as the second modal power component.

[0104] The calculation formula of the second power residual is as follows:

[0105]

[0106] Where, r 2 is the second power residual, The signal-to-noise ratio corresponding to the second addition of white noise is: This is the first optimization operator corresponding to the second addition of white noise.

[0107] Therefore, the second modal power component is calculated as follows:

[0108]

[0109] Where IMF2 is the second modal power component.

[0110] Further, the second white noise added is optimized based on the first optimization operator and the signal-to-noise ratio to obtain a second white noise;

[0111] The sum of the second power residual and the second white noise is optimized using a second optimization operator to determine a third power residual. The difference between the second and third power residuals is determined as a third modal power component. Decomposition of the power fluctuation component is terminated until the number of modal power components equals the number of times white noise is added, and multiple modal power components of different frequencies are determined.

[0112] In summary, the calculation formula for the kth residual is as follows:

[0113]

[0114] Where, For the k The power residual, For the k -1 power residual, For the k +1 times the signal-to-noise ratio of adding white noise, is the signal-to-noise ratio corresponding to the k-th addition of white noise, For the k The first optimization operator corresponding to the addition of white noise is: k =3,…, K .

[0115] Then, k The calculation formula for the modal modulus component is as follows:

[0116]

[0117] Calculating the local power mean through this disclosure can effectively reduce the residual noise level in the decomposed mode, thereby significantly improving the accuracy and clarity of signal processing. Adding an optimization operator can reduce the white noise overlap phenomenon in the power residual.

[0118] In the present disclosure, after decomposing the power fluctuation component into multiple modal power components, the instantaneous frequency of the modal power components is further extracted through Hilbert Transform (HT), and the correlation between the instantaneous frequency and time is obtained.

[0119] In the present disclosure, the power fluctuation signal is as follows:

[0120]

[0121] Where, is the power fluctuation signal, for The Hilbert transform of .

[0122] For modal component IMF i ( t), the present invention adopts Hilbert transform to obtain the corresponding H [IMFk( t )]:

[0123]

[0124] Construct the analytical signal of each modal component, that is:

[0125]

[0126] Where, is the kth analytical signal, is the instantaneous value of each IMF amplitude; is the instantaneous value of each IMF phase; is the instantaneous value of each IMF frequency, τ is the actual time, and the final Hilbert time spectrum is expressed in polar coordinates H[ω, t] , j Is an imaginary unit.

[0127] Through the curve on the time-frequency spectrum, we can observe the change pattern of the power fluctuation signal with time and frequency in the entire frequency range. The specific pattern is shown in the following formula:

[0128]

[0129] According to the determined law of change of the power fluctuation signal with time and frequency in the entire frequency range, the correlation relationship between the instantaneous frequency and time of the modal power component is determined.

[0130] Furthermore, a frequency interval is determined based on the maximum and minimum values ​​of the instantaneous frequency, a temporary division frequency is determined within the frequency interval, and based on the temporary division frequency and the correlation relationship, the energy of the aliased modal power between adjacent modal power components is calculated, and the temporary division frequency corresponding to the minimum value of the energy is determined as the allocation frequency.

[0131] Furthermore, based on a preset frequency interval, multiple temporary crossover frequencies can be determined within the frequency interval, and the time corresponding to each of the temporary crossover frequencies can be determined based on the association relationship. Modal power components corresponding to adjacent instantaneous frequencies can be determined based on the temporary crossover frequencies, and aliased modal powers of the adjacent modal power components can be determined. Based on the time and the adjacent modal power components, the energy of the aliased modal power can be calculated.

[0132] In the present disclosure, the frequency interval may be 0.000001 Hz.

[0133] The energy of the aliased modal power between adjacent modal power components is calculated as follows:

[0134]

[0135] Where, is the energy of the aliased modal power between adjacent modal power components, For the h indivual Greater than Power; For the l indivual Less than The power, is the temporary crossover frequency, For the time difference.

[0136] In the context, after determining the crossover frequency, the modal power components of different frequencies can be distributed and smoothed. Figure 4 This is a flow chart of the power fluctuation component smoothing method provided by the embodiment of the present disclosure. Figure 4 The original wind power is read and analyzed using the ELM model to predict and determine the predicted wind power. A dynamic weighted filtering method is then used to decompose the original wind power to determine the grid-connected power directly fed into the system and the power fluctuation components that require energy storage to smooth out. Adaptive frequency-splitting empirical mode decomposition is then performed on the power fluctuation components. Based on the determined crossover frequency, the wind-storage combined system energy router controls capacitors to smooth out modal power components with frequencies greater than the crossover frequency, while the wind-storage combined system energy router controls batteries to smooth out modal power components with frequencies less than the crossover frequency.

[0137] The following examples are specific implementations of this application.

[0138] For example, Figure 5 This is a topological diagram of the energy router provided by the embodiment of the present disclosure, such as Figure 5As shown, the energy router for the port wind-storage system consists of multiple DC-DC converters (DC / DCs), multiple DC-AC converters (DC / ACs), and an intelligent energy storage (VSG) controller. By controlling wind turbines, energy storage, and other devices, it implements the proposed smoothing strategy. Furthermore, the DC-DC converters are used to output the AC voltage after the smoothed power of the port wind-storage system is smoothed. The intelligent energy storage controller used in the energy router includes four interfaces: two for wind power generation, one for energy storage devices, and one for grid connection. One end of each of the DC / DCs is connected to the port wind power source interface and the hybrid energy storage power source interface, respectively. In this disclosure, they are used to obtain the original wind power and the smoothed power output by the hybrid energy storage system. Their second ends are connected in parallel to the DC bus to achieve grid connection of the wind-storage system voltage. The Energy Storage System (ESS) port controls the DC / DCs to achieve adaptive voltage regulation and bidirectional energy flow. The VSG controller is connected to the second terminal of each DC / DC converter. It controls the output of low voltage direct current (LVDC) from the DC / DC converter connected to the port wind power plant to control the capacitor to smooth out modal power components with frequencies greater than the crossover frequency, and controls the battery to smooth out modal power components with frequencies less than the crossover frequency. The DC / AC converter serves as an interface for AC loads, grid-connected devices, and other AC equipment. The DC bus voltage is converted through the DC / AC converter into high voltage alternating current (HVAC) for connection to the grid. Figure 6 This is a topological diagram of a three-level converter of an energy router of a port wind-storage combined system provided by an embodiment of the present disclosure, such as Figure 6 As shown, the converter is a three-level DC / AC converter. A three-level DC / AC converter consists of a DC section and an AC section, each composed of multiple transistor-emitter-diode (FET) units connected in series and parallel. Sa1…Sa6, Sb1…Sb6, and Sc1…Sc6 are all FET-emitter-diode (FET) units, suitable for port scenarios. They ensure stable operation of the energy router and control of the wind-storage combined system.

[0139] Figure 7 This is a structural diagram of the wind-storage combined system provided by the embodiment of the present disclosure, such as Figure 7 As shown in the figure, it mainly consists of wind turbines, supercapacitors, batteries, DC-DC converters, DC-AC inverters, and transformers. After the wind power and energy storage output power are connected to the grid, they enter the grid through the substation.

[0140] This paper has established an ELM model for predicting wind power in advance, and uses the ELM model to predict and analyze wind power. In terms of data, this paper selects the 23-year wind power historical data from September to November at the Pacific Terminal of Tianjin Port as the input data of the ELM. The wind power data for 24 hours is predicted, with an interval of 15 minutes. The prediction results are as follows: Figure 8 As shown, Figure 8 This is a schematic diagram of wind power forecasting provided by an embodiment of the present disclosure. The diagram shows the actual wind power value and the predicted value from the wind power forecasting model. The predicted value curve obtained by the wind power forecasting model closely matches the actual value curve during wind power forecasting, fully demonstrating the model's excellent performance and accuracy in the field of wind power forecasting.

[0141] Furthermore, the original wind power is adjusted by adopting a dynamic weight filtering method to reduce the noise component in the wind power signal and improve the accuracy and reliability of the wind power grid-connected power data.

[0142] Figure 9 This is a diagram showing the effect of the dynamic weighted filter provided by the embodiment of the present disclosure on smoothing the fluctuation of the original power of wind power. Figure 9 As shown in the figure, the processed wind power has a high degree of fit with the original wind power directly connected to the grid, and there is basically no time delay. At t=11:20, the original wind power is 75.29 MW, and the wind power connected to the grid after sliding average processing increases to 89.22 MW, which requires the hybrid energy storage system to release 13.93 MW of electricity to smooth the power fluctuation. Figure 8 In the example above, the grid-connected power after filtering is reduced to 81.08 MW. Accordingly, the hybrid energy storage system only needs to release 5.79 MW of power to smooth out power fluctuations.

[0143] Table 1 shows the effectiveness of different filtering methods for smoothing wind power. Compared to the traditional sliding average filtering method, the maximum fluctuation rate using dynamic weight filtering is reduced from 11.247% to 8.5914%. Furthermore, the smoothness of the power fluctuation component decreases from 3.6609% using the sliding average filtering method to 1.3998% using the dynamic weight filtering method.

[0144] Table 1

[0145]

[0146] In summary, the dynamic weighted filtering proposed in this disclosure performs well in smoothing the original power fluctuations of wind power, especially for the long-term fluctuation smoothing effect, which provides more stable and reliable power output for wind power grid-connected operation and helps to improve the safety and economy of the power grid.

[0147] Figure 10This is a schematic diagram of the power that needs to be smoothed by the hybrid energy storage system obtained by dynamic weight filtering according to the embodiment of the present disclosure. Figure 10 As can be seen in the figure, the reference output of the hybrid energy storage system fluctuates frequently near zero. If traditional energy storage systems such as batteries are relied upon to smooth out wind power fluctuations, this frequent charging and discharging process will have a significant impact on the service life of the energy storage system, significantly shortening its operating cycle. The introduction of power storage elements such as supercapacitors can effectively cope with the high-frequency fluctuations in wind power. This can reduce the burden on the battery and reduce the number of times the battery is charged and discharged, thereby significantly extending the battery's cycle life. This hybrid energy storage system design not only improves the overall system performance, but also effectively avoids damage to the battery caused by excessive charging and discharging processes, ensuring the reliable operation of the wind-storage combined system.

[0148] Based on the above implementation process, the adaptive frequency-fractional empirical mode decomposition method can be used to decompose the hybrid energy storage power, decomposing the energy storage power fluctuation component into modal functions of different frequencies, namely modal power components. The signal-to-noise ratio is set to 0.5. Taking the addition of 500 white noise cycles as an example, operators Q1(.) and N(.) are introduced, and the iterative process is repeated 5000 times. The decomposition results are shown in Figures 11(a)-11(d), which show the modal power components and their spectra provided by the embodiments of the present disclosure. The figures show eight intrinsic mode functions (IMFs) from high to low frequencies, and their spectra are obtained. The power fluctuation component is decomposed into eight components within the operating frequency band of 0-1 Hz, effectively decomposing the power fluctuation into different frequency bands, with each IMF corresponding to a specific frequency range. This approach can better allocate the power fluctuation components, thereby improving the operating efficiency and stability of the wind farm.

[0149] After the power fluctuation component of wind power is processed by Hilbert transform, eight IMFs are extracted from the original data, and the Hilbert spectrum of each IMF is obtained, which provides a deep understanding of the dynamic correlation between instantaneous frequency and time. Figure 12 is the instantaneous frequency-time curve of the modal power component provided by the embodiment of the present disclosure. Figure 12 As shown in Figure 2, the Hilbert spectrum shows the main operating frequency band of the hybrid energy storage system. The energy storage system is most active in this frequency band and undertakes most of the energy storage and release tasks. Figure 12, it can be found that the two curves of IMF3 and IMF4 maintain clear separation in almost the entire time period, and there is no aliasing phenomenon. This shows that the signals they represent are independent of each other in frequency and time, and there is no mutual interference. Therefore, the present invention selects k=4 as the hybrid energy storage power division point to reconstruct the power of the hybrid energy storage system. The selection of this division point not only helps to improve the efficiency of energy management, but also helps to ensure the stable operation of the energy storage system in different working frequency bands. Based on the above embodiments, the present disclosure adopts dynamic weight filtering and adaptive frequency division empirical mode decomposition, which not only improves the response speed of the energy storage system, but also helps to optimize the distribution and utilization of energy storage energy. By allocating the modal power component signals of different frequency bands to the most suitable energy storage equipment, the overall performance of the energy storage system can be improved, the service life of the equipment can be extended, and the operating cost of the system can be reduced.

[0150] This paper processes the decomposed IMF components according to the wind power grid-connected fluctuation limit, and the high-frequency components are smoothed by the energy storage system. The smoothing control strategy is controlled by the energy router of the four-port port wind-storage joint system, thereby achieving stable output of wind power and healthy and economical operation of energy storage equipment. Its operating conditions can be found in Figure 13 and Figure 14 . Figure 13 This is a diagram showing the effect of smoothing wind power fluctuations provided by an embodiment of the present disclosure. Figure 14 This is a schematic diagram of the operating conditions of the energy storage system provided by the embodiment of the present disclosure. Figure 12 It can be concluded that most of the fluctuation components are effectively absorbed and compensated by the energy storage system. Compared with before stabilization, the wind power fluctuation limit is reduced after stabilization. Figure 13 It can be seen that the use of adaptive frequency-splitting empirical mode decomposition (EMD) results in a more gradual change in the battery's depth of discharge (DOD) and a smaller change in the battery's state of charge. This is primarily due to the supercapacitor shouldering more of the power fluctuations, resulting in a smoother charging and discharging process for the battery and effectively reducing overcharge and overdischarge. This improves the battery's service life and better matches and utilizes the absorption capacity of the supercapacitor and battery. Because adaptive frequency-splitting EMD can more finely decompose the energy signal, the energy storage system can more accurately determine the energy distribution between the supercapacitor and battery, thereby strengthening the synergy between the two.

[0151] Based on Figure 1 Based on the same principle as the method shown in , the present disclosure also provides a power leveling system for a port wind-storage combined system. Figure 15 This is a schematic diagram of the power leveling system of the port wind and storage combined system provided by the embodiment of the present disclosure. Figure 15 The power leveling system 1500 of the port wind power storage combined system may include:

[0152] An adjustment module 1501 is used to obtain a first degree of fluctuation of the original wind power and a second degree of fluctuation of the predicted wind power, and dynamically adjust the filtering weight of the original wind power based on the first degree of fluctuation and the second degree of fluctuation; a determination module 1502 is used to adjust the filtering ratio based on the filtering weight, filter the original wind power, and determine the power fluctuation component that needs to be smoothed by the wind-storage combined system; an adaptive frequency-division empirical mode decomposition method including a local mean and an optimization operator is used to decompose the power fluctuation component to determine multiple modal power components of different frequencies; a smoothing module 1503 is used to use the capacitors of the wind-storage combined system to smooth the modal power components with a frequency greater than the frequency division frequency, and use the batteries of the wind-storage combined system to smooth the modal power components with a frequency less than the frequency division frequency.

[0153] In the present disclosure, the adjustment module 1501 is used to obtain a first grid-connected power obtained by filtering the original wind power based on a sliding average filter, and a second grid-connected power obtained by filtering the original wind power based on an anti-pulse interference average filter; calculate the standard deviation of the original wind power based on the first degree of fluctuation, and calculate the standard deviation of the predicted wind power based on the second degree of fluctuation; calculate the first weight of the first grid-connected power and the second weight of the second grid-connected power based on the standard deviation of the original wind power and the standard deviation of the predicted wind power.

[0154] In the present disclosure, the determination module 1502 is used to determine the first grid-connected component connected to the wind-storage combined system based on the first grid-connected power and the first weight; determine the second grid-connected component connected to the wind-storage combined system based on the second grid-connected power and the second weight; and determine the sum of the first grid-connected component and the second grid-connected component as the grid-connected power of the wind-storage combined system.

[0155] In the present disclosure, the determination module 1502 is used to determine the power fluctuation signal, signal-to-noise ratio and number of times white noise is added of the power fluctuation component; optimize the white noise added for the first time based on the first optimization operator and the signal-to-noise ratio to obtain a first white noise; use the local mean to estimate the sum of the first white noise and the power fluctuation signal to determine the local power mean; optimize the local power mean based on the second optimization operator to determine the first power residual, and decompose the power fluctuation component based on the local power mean and the first power residual to determine multiple modal power components of different frequencies.

[0156] In the present disclosure, the determination module 1502 is used to determine the difference between the power fluctuation signal and the first power residual as a first modal power component; optimize the sum of the first power residual and the first white noise according to the second optimization operator to determine the second power residual, and determine the difference between the first power residual and the second power residual as a second modal power component; optimize the white noise added for the second time based on the first optimization operator and the signal-to-noise ratio to obtain the second white noise; optimize the sum of the second power residual and the second white noise according to the second optimization operator to determine the third power residual, and determine it as a third modal power component based on the difference between the second power residual and the third power residual; until the number of modal power components is the same as the number of times white noise is added, stop decomposing the power fluctuation wind volume and determine multiple modal power components of different frequencies.

[0157] In the present disclosure, the determination module 1502 is used to extract the instantaneous frequency of each of the modal power components and determine the correlation between the instantaneous frequency and time; determine the frequency interval according to the maximum and minimum values ​​of the instantaneous frequency; determine the temporary division frequency within the frequency interval, and calculate the energy of the aliased modal power between adjacent modal power components based on the temporary division frequency and the correlation; determine the temporary division frequency corresponding to the minimum value of the energy as the division frequency.

[0158] In the present disclosure, the determination module 1502 is used to determine multiple temporary division frequencies within the frequency range based on a preset frequency interval, and determine the time corresponding to each of the temporary division frequencies according to the association relationship; determine the modal power components corresponding to the adjacent instantaneous frequencies according to the temporary division frequencies, and determine the aliased modal powers of the adjacent modal power components; and calculate the energy of the aliased modal power based on the time and the adjacent modal power components.

[0159] In the present disclosure, the wind-storage combined system energy router is composed of multiple DC-DC converters, multiple DC-AC converters, and an intelligent energy storage controller; one end of the DC-DC converter is connected to the port wind power and the hybrid energy storage, respectively, to obtain the original wind power and the power output of the hybrid energy storage system, and the other end is connected in parallel to the DC bus;

[0160] One end of the DC-AC converter is connected in parallel to the DC bus, and the other end is connected in parallel to the power grid, for outputting the AC voltage after the port wind storage combined system smoothes the power; the intelligent energy storage controller is used to control the capacitor to smooth the modal power component with a frequency greater than the division frequency, and to control the battery to smooth the modal power component with a frequency less than the division frequency.

[0161] In the present disclosure, the DC-AC converter is a three-level DC-AC converter; the three-level DC-AC converter includes a DC part and an AC part; the DC part and the AC part are respectively composed of multiple transistor common-emitter diode units connected in series and parallel.

[0162] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the technical solutions of the embodiments of the present invention.

Claims

1. A power leveling method for a port wind and storage combined system, characterized in that: The steps include: Obtaining a first fluctuation degree of the original wind power and a second fluctuation degree of the predicted wind power, and dynamically adjusting a filtering weight of the original wind power based on the first fluctuation degree and the second fluctuation degree; Adjusting the filtering ratio based on the filtering weight, filtering the original wind power, and determining the power fluctuation component that needs to be smoothed by the wind-storage combined system; Adopting an adaptive frequency-fractional empirical mode decomposition method including a local mean and an optimization operator to decompose the power fluctuation component and determine multiple modal power components of different frequencies; Determine a crossover frequency, control a capacitor based on the wind-storage combined system energy router to smooth the modal power component having a frequency greater than the crossover frequency, and control a battery based on the wind-storage combined system energy router to smooth the modal power component having a frequency less than the crossover frequency; The above-mentioned method of dynamically adjusting the filtering weight of the original wind power based on the first degree of fluctuation and the second degree of fluctuation includes: obtaining a first grid-connected power by filtering the original wind power based on a sliding average filter, and a second grid-connected power by filtering the original wind power based on an anti-pulse interference average filter; calculating a standard deviation of the original wind power based on the first degree of fluctuation, and calculating a standard deviation of the predicted wind power based on the second degree of fluctuation; and calculating a first weight of the first grid-connected power and a second weight of the second grid-connected power based on the standard deviation of the original wind power and the standard deviation of the predicted wind power; The optimization operator includes a first optimization operator and a second optimization operator; In the above, an adaptive frequency division empirical mode decomposition method including a local mean and an optimization operator is used to decompose the power fluctuation component and determine multiple modal power components of different frequencies, including: determining the power fluctuation signal, signal-to-noise ratio and the number of times white noise is added of the power fluctuation component; optimizing the white noise added for the first time based on the first optimization operator and the signal-to-noise ratio to obtain the first white noise; using the local mean to estimate the sum of the first white noise and the power fluctuation signal to determine the local power mean; optimizing the local power mean based on the second optimization operator to determine the first power residual, and decomposing the power fluctuation component based on the local power mean and the first power residual to determine multiple modal power components of different frequencies.

2. A power leveling method for a port wind and storage combined system according to claim 1, characterized in that: After dynamically adjusting the filtering weight of the wind power original power based on the first fluctuation degree and the second fluctuation degree, the method further includes: Determining a first grid-connected component to be connected to the wind-storage combined system based on the first grid-connected power and the first weight; Determining a second grid-connected component to be connected to the wind-storage combined system based on the second grid-connected power and the second weight; The sum of the first grid-connected component and the second grid-connected component is determined as the grid-connected power of the wind-storage combined system.

3. The power leveling method of a port wind and storage combined system according to claim 1 is characterized in that: The decomposing the power fluctuation component based on the local power mean and the first power residual to determine a plurality of modal power components of different frequencies includes: determining a difference between the power fluctuation signal and the first power residual as a first modal power component; optimizing the sum of the first power residual and the first white noise according to the second optimization operator to determine a second power residual, and determining a difference between the first power residual and the second power residual as a second modal power component; Optimizing the white noise added for the second time based on the first optimization operator and the signal-to-noise ratio to obtain a second white noise; Optimizing the sum of the second power residual and the second white noise according to the second optimization operator to determine a third power residual, and determining a third modal power component based on a difference between the second power residual and the third power residual; Until the number of the modal power components is the same as the number of times the white noise is added, the decomposition of the power fluctuation component is stopped, and a plurality of modal power components of different frequencies are determined.

4. The power leveling method of a port wind power storage system according to claim 1 is characterized in that: Determining the frequency division frequency includes: Extracting the instantaneous frequency of each of the modal power components and determining a correlation between the instantaneous frequency and time; Determining a frequency interval according to the maximum and minimum values ​​of the instantaneous frequency; Determining a temporary frequency division frequency within the frequency interval, and calculating the energy of aliased modal powers between adjacent modal power components based on the temporary frequency division frequency and the association relationship; The temporary frequency division frequency corresponding to the minimum value of the energy is determined as the frequency division frequency.

5. The power leveling method of a port wind power storage system according to claim 4 is characterized in that: The determining of a temporary crossover frequency within the frequency interval, and calculating the energy of aliased modal powers between adjacent modal power components based on the temporary crossover frequency and the association relationship, includes: Based on a preset frequency interval, determining a plurality of temporary frequency division frequencies within the frequency interval, and determining a time corresponding to each of the temporary frequency division frequencies according to the association relationship; Determining the modal power components corresponding to the adjacent instantaneous frequencies according to the temporary frequency division frequency, and determining the aliased modal powers of the adjacent modal power components; The energy of the aliased modal power is calculated based on the time and the adjacent modal power components.

6. The power leveling method of a port wind and storage combined system according to claim 1 is characterized in that: The wind-storage combined system energy router is composed of multiple DC-DC converters, multiple DC-AC converters and an intelligent energy storage controller; One end of the DC-DC converter is connected to the port wind power and the hybrid energy storage system respectively, for obtaining the original power of the wind power and the power output of the hybrid energy storage system, and the other end is connected in parallel to the DC bus; One end of the DC-AC converter is connected in parallel to the DC bus, and the other end is connected in parallel to the power grid, for outputting the AC voltage after the power of the port wind-storage combined system is smoothed; The intelligent energy storage controller is used to control the capacitor to smooth the modal power component with a frequency greater than the frequency division frequency, and to control the battery to smooth the modal power component with a frequency less than the frequency division frequency.

7. A power leveling method for a port wind and storage combined system according to claim 6, characterized in that: The DC-AC converter is a three-level DC-AC converter; The three-level DC-AC converter includes a DC part and an AC part; The DC part and the AC part are respectively composed of a plurality of transistor common-emitter diode units connected in series and in parallel.

8. A power stabilization system for a port wind power storage system, used to implement the power stabilization method for a port wind power storage system according to any one of claims 1 to 7, characterized in that: The system comprises: an adjustment module, configured to obtain a first degree of fluctuation of the original wind power and a second degree of fluctuation of the predicted wind power, and dynamically adjust a filtering weight of the original wind power based on the first degree of fluctuation and the second degree of fluctuation; a determination module, configured to adjust a filtering ratio based on the filtering weight, filter the original wind power, and determine a power fluctuation component that needs to be smoothed by the wind-storage combined system; decompose the power fluctuation component using an adaptive frequency-fractional empirical mode decomposition method including a local mean and an optimization operator to determine multiple modal power components of different frequencies; The smoothing module is used to determine the crossover frequency, control the capacitor based on the wind-storage combined system energy router to smooth the modal power component with a frequency greater than the crossover frequency, and control the battery based on the wind-storage combined system energy router to smooth the modal power component with a frequency less than the crossover frequency.

Images