Hybrid energy storage capacity configuration method considering wind and light power supply time multi-scale decomposition
By improving the wavelet packet decomposition method and configuring a hybrid energy storage system, the grid instability caused by the fluctuation of wind and solar power output was solved, thereby improving the grid's safety, stability, and economic benefits.
Patent Information
- Application Number
- CN202410961558.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-18
- Publication Date
- 2024-11-05
AI Technical Summary
Existing energy storage systems fail to effectively balance the randomness and volatility of combined wind and solar power output during configuration, resulting in unstable grid operation and poor economic benefits. Existing signal decomposition methods are complex and energy storage capacity configuration is unreasonable.
An improved wavelet packet decomposition method is used to decompose the combined wind and solar power output at multiple scales, and a probabilistic model of the combined wind and solar power output is established. The thermal power unit undertakes the low-frequency power, and the hybrid energy storage system allocates the medium and high-frequency power to configure the optimal energy storage capacity.
It has improved the grid's ability to absorb new energy power, enhanced the grid's safety, stability, and economic benefits, and suppressed the volatility of wind and solar power through the rational allocation of energy storage capacity.
Smart Images

Figure CN118899871A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical fields of wind-solar combined output prediction, signal decomposition and energy storage capacity configuration, and in particular to a hybrid energy storage capacity configuration method considering multi-scale time decomposition of wind-solar power sources. Background Art
[0002] With the change of global climate, renewable energy power generation, mainly wind and solar, has gradually become dominant. However, wind and photovoltaic power generation are uncertain, and their output is highly random and volatile due to the influence of natural conditions, which has a certain impact on the safe and stable operation of the power system.
[0003] The output of photovoltaic power stations and wind farms has a certain degree of randomness and volatility. After being connected to the grid, the uncertainty of the operation of the power grid system will increase, thereby reducing the quality of power. In some cases, it may cause the anti-interference ability, balance ability and regulation ability of the power grid system to drop significantly. The wind-solar complementary system has more advantages than a single power supply system. In order to ensure the smooth and safe operation of the power grid system, the grid-connected power needs to be strictly selected.
[0004] Energy storage systems, with their characteristics of absorbing and releasing electric energy in a short time, fast response speed and relatively flexible control, have become an effective means to stabilize the output power of new energy and improve the grid's ability to absorb new energy power. However, the access to energy storage systems will increase the relative investment cost, so it is necessary to do some research on the capacity configuration of energy storage systems.
[0005] Existing energy storage only suppresses fluctuations of single wind or light, and does not consider the difficulties brought by the combination of the two; the decomposition method of adaptive variational mode decomposition can effectively solve the impact of frequency band aliasing, but in fact the required data must be of sufficient length and is often more complicated; the grid-connected point power usually processed by wavelet packet decomposition uses node 0 as the reference power, which often results in excessive energy storage capacity configuration, which is not conducive to economic benefits. Summary of the invention
[0006] In order to solve the problem that the existing energy storage capacity is too large and is not conducive to economic benefits, the purpose of the present invention is to provide a hybrid energy storage capacity configuration method that considers multi-scale time decomposition of wind and solar power sources by performing signal decomposition of the output curve, selecting a suitable grid connection point and configuring the required energy storage capacity, so as to improve the grid's ability to absorb new energy electricity, thereby improving the safety and stability of the grid when new energy is connected to the grid.
[0007] To achieve the above object, the present invention adopts the following technical solution: a hybrid energy storage capacity configuration method considering the multi-scale decomposition of wind and solar power sources, the method comprising the following steps in order:
[0008] (1) Considering the historical power data of wind and solar power, the historical power data of wind and solar power includes local wind speed, wind turbine rated power, local light intensity, temperature and photovoltaic rated power, a probability model of wind and solar power combined output is established;
[0009] (2) The probability model of wind-solar joint output is decomposed by wavelet packet using an improved wavelet packet decomposition method to obtain the decomposed power, that is, the power required for each frequency band and the number of power data sets corresponding to each frequency band;
[0010] (3) The decomposed power is distributed to the hybrid energy storage according to the frequency as its reference power, and the optimal energy storage power is configured according to the reference power.
[0011] The step (1) specifically comprises the following steps:
[0012] (1a) According to the local wind speed and wind turbine rated power, the wind farm output probability curve P at time t is calculated. wind,t :
[0013]
[0014] Where, υ is the local wind speed; υ in ,υ r ,υ out are the cut-in wind speed, rated wind speed and cut-out wind speed of the fan respectively; P r is the rated power of the fan; A, B, and C are the polynomial fitting coefficients of the nonlinear part of the fan output curve;
[0015] The calculation formula of υ is:
[0016]
[0017] Where f(v) is the probability density function of the local wind speed υ; a is the shape coefficient, which represents the Weibull distribution corresponding to the regional wind speed distribution characteristics; b is the scale coefficient, which represents the size of the regional wind speed;
[0018] The calculation formulas for A, B, and C are:
[0019]
[0020] (1b) Based on the local light intensity, temperature and photovoltaic rated power, the photovoltaic output probability curve P at time t is calculated PI,t :
[0021]
[0022] Where P PI,R is the rated output of photovoltaic power; T t is the temperature at time t; I tis the ground light intensity at the current moment; I R is the light intensity under standard conditions, taking 1000W / m 2 ; T R is the temperature under standard conditions, which is 25°C; ε is the power temperature coefficient, which is 0.0039°C -1 ;η represents the photoelectric conversion efficiency;
[0023] I t The calculation formula is:
[0024]
[0025] In the formula, f(I t ) is the light intensity I t The probability density function of ; λ and μ are the shape parameters of the Beta distribution; I m is the maximum light intensity;
[0026] (1c) According to the wind and solar power output curve composed of equations (1) and (4), the two-dimensional vector space (X, Y) is obtained, where X = {x t |t=1,2,…,T},Y={y t |t=1,2,…,T},x t represents the wind farm power data at time t, y t represents the photovoltaic power data at the tth moment, and T represents the number of days predicted;
[0027] (1d) Normalize the two-dimensional space (X, Y) to obtain the normalized sample space (U, V), where space U = {u t |t=1,2,…,T},space V={v t |t=1,2,…,T},u t represents the normalized wind farm power data at time t, v t represents the normalized photovoltaic power data at time t;
[0028] (1e) The wind farm output probability density curve and the photovoltaic output probability density curve are obtained according to the non-parametric kernel density estimation;
[0029] (1f) Estimate the parameters θ of the connection function:
[0030]
[0031] Where L is the likelihood function of the sample space (U, V):
[0032]
[0033] In the formula, F(u t) is the marginal distribution function of space U, and the probability density function f(u t ) integration; F(v t ) is the marginal distribution function of space V, and the probability density function f(v t ) is integrated to obtain; c represents the derivative of the Frank copula function:
[0034]
[0035] Where, C represents the Frank copula function; represents the partial derivative of function C with respect to variable u; Represents the partial derivative of function C with respect to variable v;
[0036]
[0037] In the formula, yes The inverse function of is the generator of the Archimedean connection function:
[0038]
[0039] (1g) The probability model P of wind-solar combined output is obtained by fitting w (t):
[0040]
[0041] The step (2) specifically comprises the following steps in order:
[0042] (2a) Considering the rising step time of the thermal power unit, based on this, the minimum decomposition layer number n of the wavelet packet decomposition is calculated using the improved wavelet packet decomposition method;
[0043] n=INT(log2T th +log2f s -log2(T th f th )) (39)
[0044] In the formula, f s is the sampling frequency; INT is the rounding function; T th is the rise time of the step response of the thermal power unit; f th is the bandwidth of thermal power unit;
[0045] The thermal power units bear the corresponding part of the power shortage at low frequency, and the low frequency band number l corresponding to the wind and solar combined output is calculated:
[0046] l=1 1 / 4 2n-1 (40)
[0047] (2b) Calculate the decomposed frequency band power P j,l (t) and P j,l+1 (t):
[0048]
[0049] Where j represents the current layer number, j = 0, 1, ..., n; is the wavelet packet decomposition coefficient:
[0050]
[0051] In the formula, h k-2l , g k-2l They are respectively a low-pass filter bank and a high-pass filter bank of wavelet packet decomposition;
[0052] The calculation formula is:
[0053]
[0054] In the formula, They are respectively a low-pass filter bank and a high-pass filter bank reconstructed by wavelet packets;
[0055] (2c) Select k as the dividing point between the middle frequency and the high frequency;
[0056] (2d) Calculate the grid-connected power of each device, that is, the power P required for each frequency band w (t):
[0057]
[0058] Where: P n,0 (t) is a reference power of thermal power grid-connected; P n,1:2l (t) is the 2l-1 reference power of the wind-solar combined system connected to the grid; P n,2l:k (t) is the k-21 reference power of energy storage in hybrid energy storage; Supercapacitor 2 for hybrid energy storage n -k reference powers.
[0059] The step (3) specifically comprises the following steps in order:
[0060] (3a) Calculate the state of charge of energy storage:
[0061]
[0062] In the formula, SOC e (t), SOC e (t-1) are the charge states of the energy storage at time t and t-1 respectively; P dis,e,tis the discharge power of energy storage at time t; P ch,e,t is the charging power of energy storage at time t; η dis,e is the discharge efficiency of energy storage; η ch,e is the charging efficiency of energy storage; Q e is the rated capacity of the energy storage;
[0063] Calculate the state of charge of the supercapacitor:
[0064]
[0065] In the formula, SOC p (t), SOC p (t-1) are the charge states of the supercapacitors at time t and t-1 respectively; P dis,p,t is the discharge power of the supercapacitor at time t; P ch,p,t is the charging power of the supercapacitor at time t; η dis,p is the discharge efficiency of the supercapacitor; η ch,p is the charging efficiency of the supercapacitor; Q p is the rated capacity of the supercapacitor; f s is the sampling frequency;
[0066] (3b) Configure the capacity of hybrid energy storage: Calculate the required rated capacity S of energy storage e :
[0067]
[0068] In the formula, SOC is the maximum value of the energy storage state of charge; e (0) is the initial value of the energy storage charge state; P n,2l:k (t) is the k-21 reference power of energy storage in hybrid energy storage; It is the minimum value of the state of charge of energy storage;
[0069] Calculate the required rated capacity S of the supercapacitor p :
[0070]
[0071] In the formula, SOC is the maximum value of the supercapacitor state of charge; p (0) is the initial value of the supercapacitor charge state; Supercapacitor 2 for hybrid energy storage n -k reference powers; is the minimum state of charge of the supercapacitor;
[0072] Calculate the hybrid energy storage capacity S that needs to be configured HESS :
[0073] S HESS =ρS e +(1-ρ)S p (49)
[0074] Where ρ represents the energy storage ratio, ρ∈(0,1).
[0075] In claim 2, the step (1e) specifically refers to:
[0076] The wind farm output probability density curve f(u t );
[0077]
[0078] In the formula, α represents the number of quantiles; D represents the bandwidth; K1 represents the wind speed kernel function; represents the estimated standard deviation;
[0079] The photovoltaic output probability density curve f(v t ):
[0080]
[0081] Where K2 represents the wind speed kernel function:
[0082]
[0083] It can be seen from the above technical scheme that the beneficial effects of the present invention are as follows: first, the present invention quantifies the influence of wind speed on wind power generation, considers the influence of related factors of photovoltaic power generation on output, retains the randomness and volatility of wind and solar power output itself, and establishes a more accurate probability model of wind-solar joint output, which can quantify the uncertainty factors of the wind-solar complementary system and obtain a more realistic wind-solar joint output; second, the present invention uses an improved wavelet packet decomposition method to decompose the traditional signal into low, medium and high frequency bands, while considering the minimum power that can be connected to the grid by thermal power units and wind-solar joint systems, and converts the minimum The power of the low frequency is borne by the thermal power units, and the most appropriate power between the low frequency and the medium frequency is selected as the reference power for wind and solar grid connection, and its grid connection characteristics are quantified. On the one hand, it can meet the power required by the system for grid connection, and on the other hand, it can improve the grid connection stability of the power grid system; thirdly, the present invention considers the selection of the reference power required for the hybrid energy storage capacity configuration. On the premise of meeting the power of wind and solar grid connection, the medium frequency and high frequency power components are respectively allocated to energy storage and supercapacitors as reference power, so as to configure the rated capacity required for the hybrid energy storage grid connection, thereby suppressing the fluctuation of wind and solar power sources and improving the safety and stability of the grid connection. BRIEF DESCRIPTION OF THE DRAWINGS
[0084] Figure 1 This is a schematic diagram of the improved wavelet packet decomposition principle of the present invention;
[0085] Figure 2 The present invention is a flow chart of the method. DETAILED DESCRIPTION
[0086] like Figure 2 As shown, a hybrid energy storage capacity configuration method considering multi-scale time decomposition of wind and solar power sources includes the following steps in sequence:
[0087] (1) Considering the historical power data of wind and solar power, the historical power data of wind and solar power includes local wind speed, wind turbine rated power, local light intensity, temperature and photovoltaic rated power, a probability model of wind and solar power combined output is established;
[0088] (2) The probability model of wind-solar joint output is decomposed by wavelet packet using an improved wavelet packet decomposition method to obtain the decomposed power, that is, the power required for each frequency band and the number of power data sets corresponding to each frequency band;
[0089] (3) The decomposed power is distributed to the hybrid energy storage according to the frequency as its reference power, and the optimal energy storage power is configured according to the reference power.
[0090] The step (1) specifically comprises the following steps:
[0091] (1a) According to the local wind speed and wind turbine rated power, the wind farm output probability curve P at time t is calculated. wind,t :
[0092]
[0093] Where, υ is the local wind speed; υ in ,υ r ,υ out are the cut-in wind speed, rated wind speed and cut-out wind speed of the fan respectively; P r is the rated power of the fan; A, B, and C are the polynomial fitting coefficients of the nonlinear part of the fan output curve;
[0094] The calculation formula of υ is:
[0095]
[0096] Where f(v) is the probability density function of the local wind speed υ; a is the shape coefficient, which represents the Weibull distribution corresponding to the regional wind speed distribution characteristics; b is the scale coefficient, which represents the size of the regional wind speed;
[0097] The calculation formulas for A, B, and C are:
[0098]
[0099] (1b) Based on the local light intensity, temperature and photovoltaic rated power, the photovoltaic output probability curve P at time t is calculated PIt :
[0100]
[0101] Where P PI,R is the rated output of photovoltaic power; T t is the temperature at time t; I t is the ground light intensity at the current moment; I R is the light intensity under standard conditions, taking 1000W / m 2 ; T R is the temperature under standard conditions, which is 25°C; ε is the power temperature coefficient, which is 0.0039°C -1 ;η represents the photoelectric conversion efficiency;
[0102] I t The calculation formula is:
[0103]
[0104] In the formula, f(I t ) is the light intensity I t The probability density function of ; λ and μ are the shape parameters of the Beta distribution; I m is the maximum light intensity;
[0105] (1c) According to the wind and solar power output curve composed of equations (1) and (4), the two-dimensional vector space (X, Y) is obtained, where X = {x t |t=1,2,…,T},Y={y t |t=1,2,…,T},x t represents the wind farm power data at time t, y t represents the photovoltaic power data at the tth moment, and T represents the number of days predicted;
[0106] (1d) Normalize the two-dimensional space (X, Y) to obtain the normalized sample space (U, V), where space U = {u t |t=1,2,…,T},space V={v t |t=1,2,…,T},u t represents the normalized wind farm power data at time t, v t represents the normalized photovoltaic power data at time t;
[0107] (1e) The wind farm output probability density curve and the photovoltaic output probability density curve are obtained according to the non-parametric kernel density estimation;
[0108] (1f) Estimate the parameters θ of the connection function:
[0109]
[0110] Where L is the likelihood function of the sample space (U, V):
[0111]
[0112] In the formula, F(u t ) is the marginal distribution function of space U, and the probability density function f(u t ) integration; F(v t ) is the marginal distribution function of space V, and the probability density function f(v t ) is integrated to obtain; c represents the derivative of the Frank copula function:
[0113]
[0114] Where, C represents the Frank copula function; represents the partial derivative of function C with respect to variable u; Represents the partial derivative of function C with respect to variable v;
[0115]
[0116] In the formula, yes The inverse function of is the generator of the Archimedean connection function:
[0117]
[0118] (1g) The probability model P of wind-solar combined output is obtained by fitting w (t):
[0119]
[0120] like Figure 1 As shown, the step (2) specifically includes the following steps in order:
[0121] (2a) Considering the rising step time of the thermal power unit, based on this, the minimum decomposition layer number n of the wavelet packet decomposition is calculated using the improved wavelet packet decomposition method;
[0122] n=INT(log2T th +log2f s -log2(T th fth )) (66)
[0123] In the formula, f s is the sampling frequency; INT is the rounding function; T th is the rise time of the step response of the thermal power unit; f th is the bandwidth of thermal power unit;
[0124] The thermal power units bear the corresponding part of the power shortage at low frequency, and the low frequency band number l corresponding to the wind and solar combined output is calculated:
[0125] l=1 1 / 4 2n-1 (67)
[0126] (2b) Calculate the decomposed frequency band power P j,l (t) and P j,l+1 (t):
[0127]
[0128] Where j represents the current layer number, j = 0, 1, ..., n; is the wavelet packet decomposition coefficient:
[0129]
[0130] In the formula, h k-2l , g k-2l They are respectively a low-pass filter bank and a high-pass filter bank of wavelet packet decomposition;
[0131] The calculation formula is:
[0132]
[0133] In the formula, They are respectively a low-pass filter bank and a high-pass filter bank reconstructed by wavelet packets;
[0134] (2c) Select k as the dividing point between the middle frequency and the high frequency;
[0135] (2d) Calculate the grid-connected power of each device, that is, the power P required for each frequency band w (t):
[0136]
[0137] Where: P n,0 (t) is a reference power of thermal power grid-connected; P n,1:2l (t) is the 2l-1 reference power of the wind-solar combined system connected to the grid; P n,2l:k (t) is the k-21 reference power of energy storage in hybrid energy storage; Supercapacitor 2 for hybrid energy storage n -k reference powers.
[0138] The step (3) specifically comprises the following steps in order:
[0139] (3a) Calculate the state of charge of energy storage:
[0140]
[0141] In the formula, SOC e (t), SOC e (t-1) are the charge states of the energy storage at time t and t-1 respectively; P dis,e,t is the discharge power of energy storage at time t; P ch,e,t is the charging power of energy storage at time t; η dis,e is the discharge efficiency of energy storage; η ch,e is the charging efficiency of energy storage; Q e is the rated capacity of the energy storage;
[0142] Calculate the state of charge of the supercapacitor:
[0143]
[0144] In the formula, SOC p (t), SOC p (t-1) are the charge states of the supercapacitors at time t and t-1 respectively; P dis,p,t is the discharge power of the supercapacitor at time t; P ch,p,t is the charging power of the supercapacitor at time t; η dis,p is the discharge efficiency of the supercapacitor; η ch,p is the charging efficiency of the supercapacitor; Q p is the rated capacity of the supercapacitor; f s is the sampling frequency;
[0145] (3b) Configure the capacity of hybrid energy storage: Calculate the required rated capacity S of energy storage e :
[0146]
[0147] In the formula, SOC is the maximum value of the energy storage state of charge; e (0) is the initial value of the energy storage charge state; P n,2l:k (t) is the k-21 reference power of energy storage in hybrid energy storage; It is the minimum value of the state of charge of energy storage;
[0148] Calculate the required rated capacity S of the supercapacitor p :
[0149]
[0150] In the formula, SOC is the maximum value of the supercapacitor state of charge; p (0) is the initial value of the supercapacitor charge state; Supercapacitor 2 for hybrid energy storage n -k reference powers; is the minimum state of charge of the supercapacitor;
[0151] Calculate the hybrid energy storage capacity S that needs to be configured HESS :
[0152] S HESS =ρS e +(1-ρ)S p (76)
[0153] Where ρ represents the energy storage ratio, ρ∈(0,1).
[0154] In claim 2, the step (1e) specifically refers to:
[0155] The wind farm output probability density curve f(u t );
[0156]
[0157]
[0158] In the formula, α represents the number of quantiles; D represents the bandwidth; K1 represents the wind speed kernel function; represents the estimated standard deviation;
[0159] The photovoltaic output probability density curve f(v t ):
[0160]
[0161] Where K2 represents the wind speed kernel function:
[0162]
[0163] In summary, the present invention quantifies the influence of wind speed on wind power generation, considers the influence of relevant factors of photovoltaic power generation on output, retains the randomness and volatility of wind and solar output itself, establishes a more accurate probability model of wind and solar joint output, can quantify the uncertainty factors of wind and solar complementary system, and obtain more realistic wind and solar joint output; through the improved wavelet packet decomposition method, while decomposing the traditional signal in low, medium and high frequency bands, considers the minimum power that can be connected to the grid of thermal power units and wind and solar joint system, and the power with the lowest frequency is handed over to the thermal power unit, and the most suitable power between low frequency and medium frequency is selected as the wind and solar grid reference power, and its grid connection characteristics are quantified, which can meet the power required by the system for grid connection on the one hand, and improve the grid connection stability of the power grid system on the other hand; considers the selection of reference power required for hybrid energy storage capacity configuration, and on the premise of meeting the power of wind and solar grid connection, allocates medium frequency and high frequency power components to energy storage and super capacitor as reference power respectively, so as to configure the rated capacity required for hybrid energy storage grid connection, thereby suppressing the fluctuation of wind and solar power supply and improving the safety and stability of grid connection.
Claims
1. A hybrid energy storage capacity configuration method considering the multi-scale time decomposition of wind and solar power sources, characterized by: The method comprises the following steps in order: (1) Considering the historical power data of wind and solar power, the historical power data of wind and solar power includes local wind speed, wind turbine rated power, local light intensity, temperature and photovoltaic rated power, a probability model of wind and solar power combined output is established; (2) The probability model of wind-solar joint output is decomposed by wavelet packet using an improved wavelet packet decomposition method to obtain the decomposed power, that is, the power required for each frequency band and the number of power data sets corresponding to each frequency band; (3) The decomposed power is distributed to the hybrid energy storage according to the frequency as its reference power, and the optimal energy storage power is configured according to the reference power.
2. The hybrid energy storage capacity configuration method considering the multi-scale time decomposition of wind and solar power sources according to claim 1 is characterized by: The step (1) specifically comprises the following steps: (1a) According to the local wind speed and wind turbine rated power, the wind farm output probability curve P at time t is calculated. wind,t : Where, υ is the local wind speed; υ in ,υ r ,υ out are the cut-in wind speed, rated wind speed and cut-out wind speed of the fan respectively; P r is the rated power of the fan; A, B, and C are the polynomial fitting coefficients of the nonlinear part of the fan output curve; The calculation formula of υ is: Where f(v) is the probability density function of the local wind speed υ; a is the shape coefficient, which represents the Weibull distribution corresponding to the regional wind speed distribution characteristics; b is the scale coefficient, which represents the size of the regional wind speed; The calculation formulas for A, B, and C are: (1b) Based on the local light intensity, temperature and photovoltaic rated power, the photovoltaic output probability curve P at time t is calculated PI,t : Where P PI,R is the rated output of photovoltaic power; T t is the temperature at time t; I t is the ground light intensity at the current moment; I R is the light intensity under standard conditions, taking 1000W / m 2 ; T R is the temperature under standard conditions, which is 25°C; ε is the power temperature coefficient, which is 0.0039°C -1 ;η represents the photoelectric conversion efficiency; I t The calculation formula is: In the formula, f(I t ) is the light intensity I t The probability density function of ; λ and μ are the shape parameters of the Beta distribution; I m is the maximum light intensity; (1c) According to the wind and solar power output curve composed of equations (1) and (4), the two-dimensional vector space (X, Y) is obtained, where X = {x t |t=1,2,…,T},Y={y t |t=1,2,…,T},x t represents the wind farm power data at time t, y t represents the photovoltaic power data at the tth moment, and T represents the number of days predicted; (1d) Normalize the two-dimensional space (X, Y) to obtain the normalized sample space (U, V), where space U = {u t |t=1,2,…,T},space V={v t |t=1,2,…,T},u t represents the normalized wind farm power data at time t, v t represents the normalized photovoltaic power data at time t; (1e) The wind farm output probability density curve and the photovoltaic output probability density curve are obtained according to the non-parametric kernel density estimation; (1f) Estimate the parameters θ of the connection function: Where L is the likelihood function of the sample space (U, V): In the formula, F(u t ) is the marginal distribution function of space U, and the probability density function f(u t ) integration; F(v t ) is the marginal distribution function of space V, and the probability density function f(v t ) is integrated to obtain; c represents the derivative of the Frank copula function: Where, C represents the Frank copula function; represents the partial derivative of function C with respect to variable u; Represents the partial derivative of function C with respect to variable v; In the formula, yes The inverse function of is the generator of the Archimedean connection function: (1g) The probability model P of wind-solar combined output is obtained by fitting w (t):
3. The hybrid energy storage capacity configuration method considering the multi-scale time decomposition of wind and solar power sources according to claim 1 is characterized by: The step (2) specifically comprises the following steps in order: (2a) Considering the rising step time of the thermal power unit, based on this, the minimum decomposition layer number n of the wavelet packet decomposition is calculated using the improved wavelet packet decomposition method; n=INT(log2T th +log2f s -log2(T th f th )) (17) In the formula, f s is the sampling frequency; INT is the rounding function; T th is the rise time of the step response of the thermal power unit; f th is the bandwidth of thermal power unit; The thermal power units bear the corresponding part of the power shortage at low frequency, and the low frequency band number l corresponding to the wind and solar combined output is calculated: l=1 1 / 4 2n-1 (18) (2b) Calculate the decomposed frequency band power P j,l (t) and P j,l+1 (t): Where j represents the current layer number, j = 0, 1, ..., n; is the wavelet packet decomposition coefficient: In the formula, h k-2l , g k-2l They are respectively a low-pass filter bank and a high-pass filter bank of wavelet packet decomposition; The calculation formula is: In the formula, They are respectively a low-pass filter bank and a high-pass filter bank reconstructed by wavelet packets; (2c) Select k as the dividing point between the middle frequency and the high frequency; (2d) Calculate the grid-connected power of each device, that is, the power P required for each frequency band w (t): Where: P n,0 (t) is a reference power of thermal power grid-connected; P n,1:2l (t) is the 2l-1 reference power of the wind-solar combined system connected to the grid; P n,2l:k (t) is the k-21 reference power of energy storage in hybrid energy storage; Supercapacitor 2 for hybrid energy storage n -k reference powers.
4. The hybrid energy storage capacity configuration method considering the multi-scale time decomposition of wind and solar power sources according to claim 1 is characterized by: The step (3) specifically comprises the following steps in order: (3a) Calculate the state of charge of energy storage: In the formula, SOC e (t), SOC e (t-1) are the charge states of the energy storage at time t and t-1 respectively; P dis,e,t is the discharge power of energy storage at time t; P ch,e,t is the charging power of energy storage at time t; η dis,e The discharge efficiency of energy storage; η ch,e is the charging efficiency of energy storage; Q e is the rated capacity of the energy storage; Calculate the state of charge of the supercapacitor: In the formula, SOC p (t), SOC p (t-1) are the charge states of the supercapacitors at time t and t-1 respectively; P dis,p,t is the discharge power of the supercapacitor at time t; P ch,p,t is the charging power of the supercapacitor at time t; η dis,p is the discharge efficiency of the supercapacitor; η ch,p is the charging efficiency of the supercapacitor; Q p is the rated capacity of the supercapacitor; f s is the sampling frequency; (3b) Configure the capacity of hybrid energy storage: Calculate the required rated capacity S of energy storage e : In the formula, SOC is the maximum value of the energy storage state of charge; e (0) is the initial value of the energy storage charge state; P n,2l:k (t) is the k-21 reference power of energy storage in hybrid energy storage; It is the minimum value of the state of charge of energy storage; Calculate the required rated capacity S of the supercapacitor p : In the formula, SOC is the maximum value of the supercapacitor state of charge; p (0) is the initial value of the supercapacitor charge state; Supercapacitor 2 for hybrid energy storage n -k reference powers; is the minimum state of charge of the supercapacitor; Calculate the hybrid energy storage capacity S that needs to be configured HESS : S HESS =ρS e +(1-p)S p (27) Where ρ represents the energy storage ratio, ρ∈(0,1).
5. The hybrid energy storage capacity configuration method considering the multi-scale time decomposition of wind and solar power sources according to claim 2 is characterized by: In claim 2, the step (1e) specifically refers to: The wind farm output probability density curve f(u t ); In the formula, α represents the number of quantiles; D represents the bandwidth; K1 represents the wind speed kernel function; represents the estimated standard deviation; The photovoltaic output probability density curve f(v t ): Where K2 represents the wind speed kernel function:
