Frequency modulation reserve capacity analysis method based on discrete Fourier transform

Abstract

The invention discloses a frequency modulation reserve capacity analysis method based on discrete Fourier transform, solves the problem that wind power output uncertainty affects frequency modulation reserve capacity demands, and belongs to the field of power grid frequency modulation reserve demand quantification. The method comprises the steps that sampling, scene induction and scene reduction are carried out on a large amount of wind power output data to obtain a typical scene data set of wind power output, time domain signals of system prediction error signals are screened out, time periods are divided according to time intervals, sampling is carried out in all the time periods to obtain time domain sampling signals, and the time domain sampling signals are sent to the system prediction error signals; and converting the time domain sampling signal into a frequency domain signal by using discrete Fourier transform to obtain corresponding power values in different frequency band ranges, and determining the demand of the system prediction error signal in different frequency band ranges in each time period of each scene for the spare capacity.

Description

Technical Field

[0001] The invention relates to a frequency regulation reserve capacity analysis method based on discrete Fourier transform, and belongs to the field of power grid frequency regulation reserve demand quantification. Background Art

[0002] Wind power will be widely used in new power systems and gradually integrated into the grid to serve market demand to reduce carbon emissions in the power industry. However, the random characteristics of new energy will cause an imbalance between supply and demand, which will bring secondary challenges to the dynamics of grid frequency regulation and peak regulation, and also put forward higher requirements for conventional units to flexibly respond to grid dispatch instructions. In the process of fundamental reconstruction of the grid structure, the performance requirements for traditional generators will be more stringent. The integration of energy storage into new energy power generation not only helps power plants improve performance indicators, but also reduces carbon emissions.

[0003] Since electrochemical energy storage has the advantages of large capacity, fast response and high efficiency, it can better solve the problems brought about by the grid connection of new energy, and can also provide auxiliary services such as frequency regulation, peak regulation and voltage regulation to maintain the voltage and frequency stability of the power grid and reduce the waveform distortion rate of power quality.

[0004] In recent years, hydrogen energy storage, as a new type of energy storage, has been widely explored in research for its unique advantages such as high energy density, low self-loss rate, and green environmental protection. In the new power system, hydrogen energy storage can be coupled with new energy such as wind and solar to adjust the randomness and volatility of new energy stations. At the same time, hydrogen, as a generalized energy storage, can better form a comprehensive energy network with electricity, heat, gas, etc.

[0005] Researching the configuration of energy storage and frequency optimization at wind farms is a key technology that needs to be solved urgently. However, traditional energy storage planning methods are mainly aimed at deterministic scenarios and are difficult to adapt to multi-scenario environments under high-order uncertainty of wind power; at the same time, the volatility of wind power needs to be considered. Summary of the invention

[0006] In order to solve the problem that the uncertainty of wind power output affects the demand for frequency regulation reserve capacity, the present invention provides a frequency regulation reserve capacity analysis method based on discrete Fourier transform.

[0007] The frequency modulation reserve capacity analysis method based on discrete Fourier transform of the present invention comprises:

[0008] S1. Sampling the wind power prediction error to obtain sampling samples, performing reduction operations on the sampling samples, and obtaining multi-scenario data of wind power output;

[0009] S2. Filter out the time domain signal of the system prediction error signal from the multi-scenario data of wind power output, divide the filtered time domain signal into time periods according to time intervals, and sample in each time period to obtain the time domain sampling signal, and use discrete Fourier transform to convert the obtained time domain sampling signal into the frequency domain signal Y s (n);

[0010] S3, according to the frequency domain signal Y corresponding to each time period of each scene in the optimization stage s (n), and obtain the frequency domain signal Y in each scene and time period s (n) The energy distribution and power distribution in different frequency bands are then used to determine the demand for spare capacity within different frequency bands by the system prediction error signal in each scenario and each time period based on the energy distribution and power distribution in different frequency bands.

[0011] As a preferred method, the system prediction error signal y is selected from the multi-scenario data of wind power output. c The time domain signal of <0 is used to determine the capacity demand of the upper reserve type in different frequency bands for the system prediction error signal in each scenario and each time period;

[0012] Filter out the system prediction error signal y from the multi-scenario data of wind power output c The time domain signal with a value greater than 0 is used to determine the reserve capacity demand for hydrogen production under different frequency bands based on the system prediction error signal in each scenario and time period.

[0013] Preferably, in S1, a Latin hypercube sampling method is used to sample the wind power prediction error to obtain a sampling sample.

[0014] Preferably, in S1, a backward reduction method is used to perform reduction operations on the sampled samples to obtain multi-scenario data of wind power output.

[0015] Preferably, when the number of sampling points N is an odd number, the actual amplitude of the spectrum component of the nth sampling sample | W s (n)|:

[0016]

[0017] When N is an even number, the actual amplitude of the spectral component of the nth sample |W s (n)|:

[0018]

[0019] As a preference, the frequency domain signal Y s (n) The energy distribution E(f1,f2) in the frequency band [f1,f2] is:

[0020]

[0021] The corresponding power distribution P(f1,f2) is:

[0022]

[0023] [f1,f2] are the frequency bands corresponding to each type of standby.

[0024] Preferably, in S2, the sampling parameters when sampling in each time period include:

[0025] Number of sampling points N = 900, total sampling time T = 1h, sampling period τ = 4s, sampling frequency f s =1 / 4Hz.

[0026] The beneficial effects of the present invention are as follows: the present invention can determine the system's demand for frequency regulation reserve capacity within different frequency ranges, so that reserve capacity suitable for different response time requirements can be configured in advance. The operating characteristics of the alkaline electrolyzer can be taken into account, and the electric hydrogen hybrid energy storage can avoid frequent start-stop cycles and the operation of the electrolyzer at too low power levels. Regarding frequency regulation requirements, the coordinated cooperation between batteries and energy storage can complement each other's advantages and solve the reserve capacity requirements for different response times. The electrolysis-based hydrogen production system can assume part of the frequency regulation function, allowing the energy storage system to concentrate on providing reserve capacity with high response time requirements. The frequency response speed and frequency regulation evaluation indicators of the system can be enhanced. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is a schematic diagram of the process of the present invention;

[0028] Figure 2 This is a typical scenario obtained after scenario reduction;

[0029] Figure 3 This is the distribution diagram of the frequency regulation reserve capacity in each frequency band of different wind farms. DETAILED DESCRIPTION

[0030] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0031] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0032] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments, but they are not intended to limit the present invention.

[0033] This embodiment is a frequency regulation reserve capacity analysis method considering the uncertainty of wind power output in a wind-hydrogen combined power system, specifically a frequency regulation reserve capacity analysis method based on discrete Fourier transform, including:

[0034] Step 1: Sampling the wind power prediction error to obtain a sampling sample, performing a reduction operation on the sampling sample, and obtaining multi-scenario data of wind power output.

[0035] Step 2: Filter out the time domain signal of the system prediction error signal from the multi-scenario data of wind power output, divide the filtered time domain signal into time periods according to time intervals, and sample in each time period to obtain the time domain sampling signal, and use discrete Fourier transform to convert the obtained time domain sampling signal into the frequency domain signal Y s (n);

[0036] Step 3: According to the frequency domain signal Y corresponding to each time period of each scene in the optimization stage s (n), and obtain the frequency domain signal Y in each scene and time period s (n) energy distribution and power distribution in different frequency bands, and then determining the demand for spare capacity in different frequency bands for the system prediction error signal in each scenario and each time period according to the energy distribution and power distribution in different frequency bands;

[0037] The obtained hydrogen production reserve capacity demand for each scenario and each time period will serve as the minimum constraint condition for the system to provide reserve in that scenario and that time period. That is, the sum of the reserved reserve capacity of the units is not less than the demand value, which can guarantee the demand for frequency regulation reserve of wind power output.

[0038] Specifically, step 1 of this embodiment includes:

[0039] Related research shows that the wind power output prediction error Δp w Subject to the expectation μ w , the variance is The normal distribution of can be expressed as The specific implementation method of the Latin hypercube sampling method is as follows:

[0040] (1) The random variable Δp w Divided into M regions, such that Δp1<Δp<Δp2, Δp2<Δp<Δp3, …, Δp M <Δp<Δp M+1 , the interval satisfies

[0041] (2) Assume that U is uniformly distributed on the interval [0,1], and draw a sample u from it. Dp is drawn from each region. The sample drawn in the mth interval is

[0042]

[0043] In the whole interval, a total of M random variables are drawn.

[0044] (3) In order to ensure the independence of each random variable, the order of the extracted random variables is changed, and certain disturbances are added to combine them into a new random vector.

[0045] Since there are a large number of similar wind power output scenarios after sampling, the backward reduction method will be used to reduce the wind power output data:

[0046] Scene i With scene w j The distance is defined as:

[0047]

[0048] Where v i,t is the wind speed of the i-th scene at time t, w i =(v i,1 ,v i,2 ,…v i,t ,…,v i,T ).

[0049] The probability distance between scene sets is:

[0050]

[0051] Where p i For scene w i The probability of occurrence, J is the set of deleted scenes.

[0052] The objective function of scene reduction is to minimize the probability distance between the scene sets before and after scene reduction. Figure 2 These are four typical scenarios after reduction.

[0053] Specifically, step 2 of this embodiment includes:

[0054] In order to quantify the capacity demand of the upper reserve type for the system prediction error, we first select the system prediction error y c Time domain signal y < 0 c (t), and then divide the time period into time periods according to the time interval, and sample within the time period to obtain the time domain sampling signal y s ,y s =y c (kτ), where k is the sampling sample number, which is a numerical point on the discrete time axis, k = 0, 1, 2, ..., N-1; τ is the sampling period, and the total sampling period is Nτ.

[0055] Then, the time domain sampling signal y of the system prediction error is transformed into sConvert to frequency domain signal Y s (n), defined by DFT:

[0056]

[0057] Where n is Y s The number of the sample value in (n) is the numerical point on the discrete frequency axis, n = 0, 1, 2, ..., N-1. The relationship between n and the angular frequency ω is

[0058]

[0059] It can be deduced that the relationship between n and frequency f is

[0060]

[0061] Where f s =1 / τ is the sampling frequency, Hz.

[0062] According to Shannon sampling theorem, frequency domain aliasing can be avoided only when the sampling frequency is greater than twice the highest frequency of the signal, that is, f s >2f max According to the frequency bands corresponding to each type of spare in Table 1, it can be known that f s >1 / 60Hz. Based on the prediction error data obtained by the Latin hypercube sampling method and the scene reduction method, this embodiment selects the number of sampling points N = 900, the total sampling time T = 1h, the sampling period τ = 4s, and the sampling frequency f s =1 / 4Hz, meeting the sampling requirements.

[0063] Table 1 Division of backup resources and their response time and frequency band

[0064]

[0065]

[0066] In the wind-hydrogen combined power system, the downward reserve can be borne by the electric hydrogen production system, reducing the regulation pressure of electrochemical energy storage.

[0067] Through the sampling period and sampling frequency, it can be determined that the actual time corresponding to the occurrence of prediction error is t = kτ, and the actual frequency is f = (f s / N)n. The actual amplitude of the nth spectral component in the frequency domain signal can be obtained |W s (n)|.

[0068] When N is an odd number, we have:

[0069]

[0070] When N is an even number, we have:

[0071]

[0072] From the symmetry, we can know that the frequency spectrum graph corresponds to the frequency f with k = N / 2 s / 2 is the axis of symmetry, so the spectrum analysis of the system prediction error only needs to consider [0,f s / 2] frequency range, in this implementation mode, f s / 2=1 / 8Hz. According to Parseval's theorem, the energy contained in a signal is equal to the sum of the energies of its components. Therefore, the frequency domain signal Y s (n) The energy distribution E(f1,f2) in the frequency band [f1,f2] is

[0073]

[0074] The corresponding power distribution P(f1,f2) is:

[0075]

[0076] [f1,f2] are the frequency bands corresponding to each type of standby.

[0077] In practical terms, P(f1, f2) is the capacity demand of the system prediction error signal for the response frequency in the range [f1, f2]. Repeating the above steps for each scenario and each time period in the optimization phase can quantify the capacity demand of backup types of different time scales caused by the system prediction error in this scenario and this time period. According to the above transformation method, Figure 3 As shown, the distribution diagram of the reserve capacity required for different wind farms is obtained.

[0078] Filter out the system prediction error signal y from the multi-scenario data of wind power output c The time domain signal with a value greater than 0 is used to determine the reserve capacity demand for hydrogen production under different frequency bands based on the system prediction error signal in each scenario and time period.

[0079] In summary, the present invention proposes a method for quantifying the frequency regulation reserve demand of wind power, which can determine the system's demand for frequency regulation reserve capacity within different frequency ranges, so that the reserve capacity suitable for different response time requirements can be configured in advance. Taking into account the operating characteristics of the alkaline electrolyzer, electric hydrogen hybrid energy storage can avoid frequent start-stop cycles and the operation of the electrolyzer at too low power levels. Regarding frequency regulation requirements, the coordinated cooperation between batteries and energy storage can complement each other's advantages and solve the reserve capacity requirements with different response times. The electrolysis-based hydrogen production system can assume part of the frequency regulation function, allowing the energy storage system to concentrate on providing reserve capacity with high response time requirements. The frequency response speed and frequency regulation evaluation indicators of the system can be enhanced.

[0080] Although the present invention is described herein with reference to specific embodiments, it should be understood that these embodiments are merely examples of the principles and applications of the present invention. It should therefore be understood that many modifications may be made to the exemplary embodiments and that other arrangements may be devised without departing from the spirit and scope of the present invention as defined by the appended claims. It should be understood that the various dependent claims and features described herein may be combined in a manner different from that described in the original claims. It should also be understood that features described in conjunction with individual embodiments may be used in other described embodiments.

Claims

1. A frequency modulation reserve capacity analysis method based on discrete Fourier transform, characterized in that: The method comprises: S1. Sampling the wind power prediction error to obtain sampling samples, performing reduction operations on the sampling samples, and obtaining multi-scenario data of wind power output; S2. Filter out the time domain signal of the system prediction error signal from the multi-scenario data of wind power output, divide the filtered time domain signal into time periods according to time intervals, and sample in each time period to obtain the time domain sampling signal, and use discrete Fourier transform to convert the obtained time domain sampling signal into the frequency domain signal Y s (n); S3, according to the frequency domain signal Y corresponding to each time period of each scene in the optimization stage s (n), and obtain the frequency domain signal Y in each scene and time period s (n) The energy distribution and power distribution in different frequency bands are then used to determine the demand for spare capacity within different frequency bands by the system prediction error signal in each scenario and each time period based on the energy distribution and power distribution in different frequency bands.

2. The frequency modulation reserve capacity analysis method based on discrete Fourier transform according to claim 1 is characterized in that: Filter out the system prediction error signal y from the multi-scenario data of wind power output c The time domain signal of <0 is used to determine the capacity demand of the upper reserve type in different frequency bands for the system prediction error signal in each scenario and each time period; Filter out the system prediction error signal y from the multi-scenario data of wind power output c The time domain signal with a value greater than 0 is used to determine the reserve capacity demand for hydrogen production under different frequency bands based on the system prediction error signal in each scenario and time period.

3. The frequency modulation reserve capacity analysis method based on discrete Fourier transform according to claim 1 is characterized in that: In S1, the Latin hypercube sampling method is used to sample the wind power prediction error to obtain a sampling sample.

4. The frequency modulation reserve capacity analysis method based on discrete Fourier transform according to claim 1 is characterized in that: In S1, the backward reduction method is used to perform reduction operations on the sampled samples to obtain multi-scenario data of wind power output.

5. The frequency modulation reserve capacity analysis method based on discrete Fourier transform according to claim 1 is characterized in that: When the number of sampling points N is an odd number, the actual amplitude of the spectral component of the nth sample | W s (n)|: When N is an even number, the actual amplitude of the spectral component of the nth sample |W s (n)|:

6. The frequency modulation reserve capacity analysis method based on discrete Fourier transform according to claim 5 is characterized in that: Frequency domain signal Y s (n) The energy distribution E(f1,f2) in the frequency band [f1,f2] is: The corresponding power distribution P(f1,f2) is: [f1,f2] are the frequency bands corresponding to each type of standby.

7. The frequency modulation reserve capacity analysis method based on discrete Fourier transform according to claim 6 is characterized in that: In S2, the sampling parameters for sampling in each time period include: Number of sampling points N = 900, total sampling time T = 1h, sampling period τ = 4s, sampling frequency f s =1 / 4Hz.

8. A computer-readable storage device storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the frequency modulation reserve capacity analysis method based on discrete Fourier transform as claimed in any one of claims 1 to 7 are implemented.

9. A disk satellite mass moment attitude control device, comprising a storage device, a processor, and a computer program stored in the storage device and executable on the processor, characterized in that: The processor executes the computer program to implement the steps of the frequency modulation reserve capacity analysis method based on discrete Fourier transform according to any one of claims 1 to 7.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the frequency modulation reserve capacity analysis method based on discrete Fourier transform as claimed in any one of claims 1 to 7 are implemented.

Images