Wind-solar power fluctuation stabilizing method based on hydraulic turbine set and energy storage

Abstract

The invention relates to a wind-solar power fluctuation stabilizing method based on a hydraulic turbine set and energy storage, and belongs to the technical field of power grids. The method comprises the following steps: firstly, constructing a wind-light-water storage integrated power generation system; performing adaptive moving average filtering on the wind and light output power to obtain power meeting grid connection requirements and stabilizing reference power; then, parameters K and # imgabs0 # of variational mode decomposition are optimized through a whale optimization algorithm, reference power is stabilized through parameter-optimized VMD decomposition, and a mode component is obtained; performing Hilbert marginal spectrum analysis on each modal component obtained by VMD decomposition, and determining high and low frequency components according to a modal aliasing phenomenon; low-frequency signals are distributed to a hydraulic turbine set, high-frequency signals are distributed to a lithium battery, and wind and light power is stabilized. The method can effectively stabilize the fluctuation of the wind-light output power and is easy to popularize and apply.

Description

Technical Field

[0001] The present invention belongs to the technical field of power grids, and in particular relates to a method for smoothing wind and solar power fluctuations based on a hydro turbine unit and energy storage. Background Art

[0002] In recent years, natural resources such as wind and light have been widely used to solve energy crises and environmental pollution problems, and wind power generation and photovoltaic power generation technologies have developed rapidly in countries around the world. Due to the randomness and volatility of wind and solar output power, direct grid connection will affect the frequency stability of the power grid. To solve this problem, energy storage systems are usually configured in wind and solar power generation systems to effectively smooth out the fluctuations in wind and solar output power and enable them to be stably connected to the grid.

[0003] Hydropower has the advantages of rapid start and stop and convenient regulation. Therefore, hydropower is often used in power grids to undertake important tasks such as peak load regulation and frequency regulation. Hydropower plays a very critical role in maintaining the safe and stable operation of the power grid. Hydropower accounts for about 20% of my country's total power capacity, and the remaining exploitable water resources are very limited. In modern smart grids, hydropower is extremely precious. How to make economical and efficient use of hydropower resources is a very important research topic facing my country's power science and technology workers.

[0004] Using energy storage systems to smooth wind power is a feasible solution to the current difficulty in connecting large-scale wind and solar power to the grid. Flywheels and conventional chemical energy storage equipment are usually used to smooth wind power, but this technical solution is often restricted by the capacity and cost of the energy storage system, and it is difficult to meet the needs of large-scale wind power grid connection; pumped storage power stations can also be used to smooth wind power, but the cost of building dedicated pumped storage power stations is often very high, and there may not be objective geographical resource conditions suitable for building pumped storage power stations near wind farms, so this solution is not ideal.

[0005] Therefore, how to overcome the shortcomings of the existing technology is an urgent problem to be solved in the field of power grid technology. Summary of the invention

[0006] The purpose of the present invention is to solve the deficiencies of the prior art and to provide a method for smoothing wind and solar power fluctuations based on a hydro turbine unit and energy storage.

[0007] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0008] A method for smoothing wind and solar power fluctuations based on a hydro turbine unit and energy storage comprises the following steps:

[0009] Step 1: Construct a wind-solar-water-storage integrated power generation system; the wind-solar-water-storage integrated power generation system includes a wind power generation system, a photovoltaic power generation system and a suppression system, and the suppression system includes a turbine unit and a lithium battery energy storage device;

[0010] Step 2: Perform adaptive moving average filtering on the output power of the wind power generation system and the photovoltaic power generation system to obtain the power P that meets the grid connection requirements. g (t) Smoothing reference power P h (t);

[0011] Step 3: Use Whale WOA to optimize the VMD parameter modal decomposition number K and quadratic penalty factor α, and use the VMD decomposition after parameter optimization to smooth the reference power and obtain the modal components;

[0012] Step 4: Perform Hilbert marginal spectrum analysis on each modal component obtained by VMD decomposition, determine the high and low frequency components according to the degree of modal aliasing, allocate the low frequency signal to the turbine unit, and the high frequency signal to the lithium battery energy storage device to smooth the wind and solar power.

[0013] Further, preferably, in Step 1, the wind power generation system and the photovoltaic power generation system are connected to the AC bus, the lithium battery energy storage device is connected to the AC bus through a DC / AC converter, the hydro-turbine generator set is connected to the AC bus, and the AC bus is connected to the power grid through a step-up transformer;

[0014] Power balance is required in the system, and the power balance relationship is as follows:

[0015] P h (t) = P W (t)+P V (t)-P g (t) (1)

[0016] P h (t) = P b (t)+P s (t) (2)

[0017] Where P W (t) is the output power of the wind power generation system; P V (t) is the output power of the photovoltaic power generation system; P g (t) is the power required for grid connection; P h (t) is the reference power for smoothing; P b (t) is the output power of the lithium battery; P s (t) is the output power of the turbine unit;

[0018] Further, preferably, the specific method of Step 2 is:

[0019] According to GB / T19963-2021 "Technical Regulations for Wind Farm Access to Power System" and GB / T 19964-2024 "Technical Regulations for Photovoltaic Power Station Access to Power System", the 1-minute active power and 10-minute active power change requirements are respectively set for the wind power generation system output power P W (t), photovoltaic power generation system output power P V (t) Perform moving average filtering, and directly connect the output power that meets the grid connection conditions to the grid. The output power that does not meet the grid connection conditions will be allocated to the turbine unit and lithium battery as a smoothing reference power.

[0020] Further, preferably, the specific method of Step 3 is:

[0021] Step 3.1: Input the smoothing reference power P h (t), set the parameter range of K and α in the VMD algorithm and initialize the parameters in the WOA model. The parameters in the WOA model include population size, maximum number of iterations, spatial dimension, and initial population individuals;

[0022] Step 3.2: Smoothing reference power P h (t) Perform VMD decomposition; smooth the reference power P h The time corresponding to (t) is a discrete time series {x(i), i = 1 to N}, with a total of N moments. The phase space of this series is reconstructed to obtain the following reconstruction matrix:

[0023]

[0024] Where: r = N-(m-1)τ, r is the number of reconstructed row vectors; j is the j-th row component of the reconstructed matrix, j = 1 to r; m is the embedding dimension; τ is the delay time;

[0025] Arrange the r row vectors in the reconstruction matrix in ascending order according to the element values ​​to obtain their respective symbol sequences S(q)=(j1,j2,···,jm); where q=1~r, r≤m!; m! is the total number of symbol sequences mapped in the m-dimensional phase space; j1,j2,···,jm represent the index number of each element in the original reconstruction component;

[0026] Calculate the probability P1, P2, ···, P of r different symbol sequences S(q) appearing r , then the permutation entropy of the time series {x(i), i=1~N} is defined as:

[0027]

[0028] Where: H p is the permutation entropy; P i is the probability of occurrence of each sequence;

[0029] When the permutation entropy is the smallest, the corresponding parameters are optimal;

[0030] Step 3.3: When |A|≤1, select the whale position corresponding to the minimum permutation entropy as the target value of local development, and then select formula (2) to update the position of the individual whale according to the randomly distributed number p; when |A|>1, randomly select a whale position, and update the position of the individual whale according to formula (3), retaining the optimal fitness and the corresponding parameter combination; the details are as follows:

[0031]

[0032] X(t+1)=X rand (t)-A·D(3)

[0033] in,

[0034] Where: X is the current target position; X * is the whale position; A is the coefficient matrix; a is the convergence factor; r1 is a number randomly distributed on [0,1]; tmax is the maximum number of iterations; D = |X * (t)-X(t)| represents the distance from the whale to the prey, D' is the transpose of D; b is a spiral constant; l is a random number in (-1,1); X rand (t) is the randomly selected whale position; p is a number randomly distributed on [0,1]; t is the tth iteration;

[0035] Step 3.4: Keep the updated whale population position as the initial population for the new round;

[0036] Step 3.5: Output the optimal whale individual and its corresponding fitness;

[0037] Step 3.6: Determine whether the maximum number of iterations is exceeded; if not, return to Step 3.2; if exceeded, output the optimal parameter modal decomposition number K and quadratic penalty factor α combination;

[0038] Step 3.7: Input the optimal parameter modal decomposition number K and the quadratic penalty factor α obtained in Step 3.6 into VMD to obtain K modal components IMF.

[0039] Furthermore, preferably, b=1; K=8.

[0040] Further, preferably, the specific method of Step 4 is:

[0041] A Hilbert marginal spectrum analysis is performed on the IMF, and according to the aliasing degree of the Hilbert marginal spectrum, IMF1~IMF2 are allocated to the turbine unit as low-frequency components for smoothing; IMF3~IMF8 are allocated to the lithium battery as high-frequency components for smoothing.

[0042] In the present invention, the power required for grid connection is the power that can be directly connected to the grid; the leveling reference power is allocated to the leveling system.

[0043] The present invention uses double-layer WOA-VMD and Hilbert to perform high- and low-frequency decomposition of the smooth reference power. The WOA-VMD is used to adaptively decompose the smooth reference power, and the optimal K and α are input into VMD to obtain K modal component IMFs.

[0044] The present invention proposes a method for smoothing wind and solar power fluctuations based on a hydro-turbine unit and energy storage, which is used to smooth the frequency instability of the power grid caused by the grid connection of wind and solar power generation. First, a wind, solar, and water storage integrated power generation system is constructed. The wind and solar output power is subjected to an adaptive moving average filter to obtain the power that meets the grid connection requirements and the reference power for smoothing. Then, WOA is used to optimize the parameter modal decomposition number K and the quadratic penalty factor α of VMD, and the VMD after parameter optimization is used to decompose the reference power for smoothing to obtain the modal components; the modal components obtained by VMD decomposition are subjected to Hilbert marginal spectrum analysis, and the high and low frequency components are determined according to the modal aliasing phenomenon; the low frequency signal is allocated to the hydro-turbine unit, and the high frequency signal is allocated to the lithium battery to smooth the wind and solar power. This method can effectively smooth the fluctuation of wind and solar output power.

[0045] In the present invention, the fitness of each individual in the initial population is calculated using formula (1). As the fitness function of this method, permutation entropy is used to measure the decomposition effect of parameter combination. The smaller the entropy, the more regular the time series distribution is, indicating that the IMF obtained by VMD processing contains more effective information; conversely, the closer the time series is to random distribution, the more noise components there are in the IMF. Therefore, when the permutation entropy is the smallest, the corresponding parameters are optimal.

[0046] Compared with the prior art, the present invention has the following beneficial effects:

[0047] (1) Using WOA to optimize VMD parameters has a faster search speed and higher optimization accuracy, with an accuracy of 95%.

[0048] (2) WOA-VMD is used to decompose the wind and solar power signals, and the hydro-turbine unit is combined with the lithium battery energy storage device to complete the smooth reference power allocation, avoiding the serious aliasing phenomenon caused by EMD decomposition. In addition, the proposed improved power allocation strategy is more accurate than the traditional power allocation strategy of single-layer VMD decomposition, and the accuracy is improved to 98%.

[0049] (3) By combining with lithium batteries, the turbine units can more effectively access and utilize renewable energy, reduce the phenomenon of wind and solar power abandonment, and reduce the annual average wind and solar power abandonment by 15%, thereby improving the utilization rate of renewable energy. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a topological diagram of the wind, solar, water and storage integrated power generation system of the present invention;

[0051] Figure 2 is a flow chart of the adaptive moving average algorithm of the present invention;

[0052] Figure 3 is a flow chart of optimizing VMD parameters by the WOA algorithm of the present invention;

[0053] Figure 4 is the I MF decomposed by WOA-VMD in Application Example 1 of the present invention;

[0054] Figure 5 is the Hilbert marginal spectrum of each IMF in Application Example 1 of the present invention;

[0055] Figure 6 This is a power comparison diagram before and after wind and solar power are stabilized in Application Example 1 of the present invention;

[0056] Figure 7 It is a flow chart of the method for smoothing wind and solar power fluctuations based on a hydro turbine unit and energy storage according to the present invention. DETAILED DESCRIPTION

[0057] The present invention is further described in detail below in conjunction with embodiments.

[0058] Those skilled in the art will appreciate that the following examples are only used to illustrate the present invention and should not be considered to limit the scope of the present invention. If no specific techniques or conditions are specified in the examples, the techniques or conditions described in the literature in the art or the product specifications are used. If the manufacturer of the materials or equipment used is not specified, they are all conventional products that can be purchased.

[0059] Example 1

[0060] like Figure 7 As shown, a method for smoothing wind and solar power fluctuations based on a hydro turbine unit and energy storage includes the following steps:

[0061] Step 1: Construct a wind-solar-water-storage integrated power generation system; the wind-solar-water-storage integrated power generation system includes a wind power generation system, a photovoltaic power generation system and a suppression system, and the suppression system includes a turbine unit and a lithium battery energy storage device;

[0062] Step 2: Perform adaptive moving average filtering on the output power of the wind power generation system and the photovoltaic power generation system to obtain the power P that meets the grid connection requirements. g (t) Smoothing reference power P h (t);

[0063] Step 3: Use Whale WOA to optimize the VMD parameter modal decomposition number K and quadratic penalty factor α, and use the VMD decomposition after parameter optimization to smooth the reference power and obtain the modal components;

[0064] Step 4: Perform Hilbert marginal spectrum analysis on each modal component obtained by VMD decomposition, determine the high and low frequency components according to the degree of modal aliasing, allocate the low frequency signal to the turbine unit, and the high frequency signal to the lithium battery energy storage device to smooth the wind and solar power.

[0065] Example 2

[0066] like Figure 7 As shown, a method for smoothing wind and solar power fluctuations based on a hydro turbine unit and energy storage includes the following steps:

[0067] Step 1: Construct a wind-solar-water-storage integrated power generation system; the wind-solar-water-storage integrated power generation system includes a wind power generation system, a photovoltaic power generation system and a suppression system, and the suppression system includes a turbine unit and a lithium battery energy storage device;

[0068] Step 2: Perform adaptive moving average filtering on the output power of the wind power generation system and the photovoltaic power generation system to obtain the power P that meets the grid connection requirements. g (t) Smoothing reference power P h (t);

[0069] Step 3: Use Whale WOA to optimize the VMD parameter modal decomposition number K and quadratic penalty factor α, and use the VMD decomposition after parameter optimization to smooth the reference power and obtain the modal components;

[0070] Step 4: Perform Hilbert marginal spectrum analysis on each modal component obtained by VMD decomposition, determine the high and low frequency components according to the degree of modal aliasing, allocate the low frequency signal to the turbine unit, and the high frequency signal to the lithium battery energy storage device to smooth the wind and solar power.

[0071] Specifically, in Step 1, the wind power generation system and the photovoltaic power generation system are connected to the AC bus, the lithium battery energy storage device is connected to the AC bus through a DC / AC converter, the hydro-turbine generator set is connected to the AC bus, and the AC bus is connected to the grid through a step-up transformer;

[0072] Power balance is required in the system, and the power balance relationship is as follows:

[0073] P h (t) = P W (t)+P V (t)-P g (t) (1)

[0074] P h (t) = P b (t)+P s (t) (2)

[0075] Where P W (t) is the output power of the wind power generation system; P V (t) is the output power of the photovoltaic power generation system; P g (t) is the power required for grid connection; P h (t) is the reference power for smoothing; P b (t) is the output power of the lithium battery; P s (t) is the output power of the turbine unit;

[0076] The specific method of Step 2 is:

[0077] According to GB / T19963-2021 "Technical Regulations for Wind Farm Access to Power System" and GB / T 19964-2024 "Technical Regulations for Photovoltaic Power Station Access to Power System", the 1-minute active power and 10-minute active power change requirements are respectively set for the wind power generation system output power P W (t), photovoltaic power generation system output power P V (t) Perform moving average filtering, and directly connect the output power that meets the grid connection conditions to the grid. The output power that does not meet the grid connection conditions will be allocated to the turbine unit and lithium battery as a smoothing reference power.

[0078] The specific method of Step 3 is:

[0079] Step 3.1: Input the smoothing reference power P h (t), set the parameter range of K and α in the VMD algorithm and initialize the parameters in the WOA model. The parameters in the WOA model include population size, maximum number of iterations, spatial dimension, and initial population individuals;

[0080] Step 3.2: Smoothing reference power P h (t) Perform VMD decomposition; smooth the reference power P h The time corresponding to (t) is a discrete time series {x(i), i = 1 to N}, with a total of N moments. The phase space of this series is reconstructed to obtain the following reconstruction matrix:

[0081]

[0082] Where: r = N-(m-1)τ, r is the number of reconstructed row vectors; j is the j-th row component of the reconstructed matrix, j = 1 to r; m is the embedding dimension; τ is the delay time;

[0083] Arrange the r row vectors in the reconstruction matrix in ascending order according to the element values ​​to obtain their respective symbol sequences S(q)=(j1,j2,···,jm); where q=1~r, r≤m!; m! is the total number of symbol sequences mapped in the m-dimensional phase space; j1,j2,···,jm represent the index number of each element in the original reconstruction component;

[0084] Calculate the probability P1, P2, ···, P of r different symbol sequences S(q) appearing r , then the permutation entropy of the time series {x(i), i=1~N} is defined as:

[0085]

[0086] Where: H p is the permutation entropy; P i is the probability of occurrence of each sequence;

[0087] When the permutation entropy is the smallest, the corresponding parameters are optimal;

[0088] Step 3.3: When |A|≤1, select the whale position corresponding to the minimum permutation entropy as the target value of local development, and then select formula (2) to update the position of the individual whale according to the randomly distributed number p; when |A|>1, randomly select a whale position, and update the position of the individual whale according to formula (3), retaining the optimal fitness and the corresponding parameter combination; the details are as follows:

[0089]

[0090] X(t+1)=X rand (t)-A·D(3)

[0091] in,

[0092] Where: X is the current target position; X * is the whale position; A is the coefficient matrix; a is the convergence factor; r1 is a number randomly distributed on [0,1]; tmax is the maximum number of iterations; D = |X * (t)-X(t)| represents the distance from the whale to the prey, D' is the transpose of D; b is a spiral constant; l is a random number in (-1,1); X rand (t) is the randomly selected whale position; p is a number randomly distributed on [0,1]; t is the tth iteration;

[0093] Step 3.4: Keep the updated whale population position as the initial population for the new round;

[0094] Step 3.5: Output the optimal whale individual and its corresponding fitness;

[0095] Step 3.6: Determine whether the maximum number of iterations is exceeded; if not, return to Step 3.2; if exceeded, output the optimal parameter modal decomposition number K and quadratic penalty factor α combination;

[0096] Step 3.7: Input the optimal parameter modal decomposition number K and the quadratic penalty factor α obtained in Step 3.6 into VMD to obtain K modal component IMFs.

[0097] b=1;K=8.

[0098] The specific method of Step 4 is:

[0099] The Hilbert marginal spectrum analysis is performed on the IMF, and according to the aliasing degree of the Hilbert marginal spectrum, IMF1-IMF2 are allocated to the turbine unit as low-frequency components for smoothing; IMF3-IMF8 are allocated to the lithium battery as high-frequency components for smoothing.

[0100] Example 3

[0101] like Figure 7 As shown, a method for smoothing wind and solar power fluctuations based on a hydro turbine unit and energy storage includes the following steps:

[0102] First, a wind-solar-water-storage integrated power generation system is constructed. The wind-solar output power is filtered by adaptive moving average to obtain the power P that meets the grid connection requirements. g (t) Smoothing reference power P h (t), where the power that meets the grid connection requirements is the power that can be directly connected to the grid; the smoothing reference power is allocated to the smoothing system. Then, the modal decomposition number K and the quadratic penalty factor α of VMD are optimized by WOA, and the smoothing reference power is decomposed by the VMD after parameter optimization to obtain the modal components; the modal components obtained by VMD decomposition are subjected to Hilbert marginal spectrum analysis, and the high and low frequency components are determined according to the modal aliasing phenomenon; the low frequency signal is allocated to the turbine unit, and the high frequency signal is allocated to the lithium battery to smooth the wind and solar power. This method can effectively smooth the fluctuation of wind and solar output power.

[0103] The specific steps are:

[0104] Step 1: Build an integrated wind, solar, and water storage power generation system. The power generation system includes a wind power generation system and a photovoltaic power generation system, and the leveling system includes a hydro turbine set and a lithium battery energy storage device; the wind power generation system and the photovoltaic power generation system are connected to the AC bus, the lithium battery is connected to the AC bus through a DC / AC converter, the hydro turbine set is connected to the AC bus, and the AC bus is connected to the power grid through a step-up transformer;

[0105] Power balance is required in the system. The power balance relationship is as follows:

[0106] P h (t) = P W (t)+P V (t)-P g (t)

[0107] P h (t) = P b (t)+P s (t)

[0108] Where P W (t) is the output power of the wind power generation system; P V (t) is the output power of the photovoltaic power generation system; P g (t) is the power required for grid connection; P h (t) is the reference power for smoothing; P b (t) is the output power of the lithium battery; P s (t) is the output power of the turbine unit.

[0109] Step 2: Use the adaptive moving average algorithm to obtain the power P that meets the grid connection requirements. g (t) Smoothing reference power P h (t); First, according to GB / T19963-2021 "Technical Regulations for Wind Farm Access to Power System" and GB / T 19964-2024 "Technical Regulations for Photovoltaic Power Station Access to Power System", the 1-minute active power and 10-minute active power change requirements for the wind power generation system output power P W (t), photovoltaic power generation system output power P V (t) Perform moving average filtering, and directly connect the output power that meets the grid connection conditions to the grid. The output power that does not meet the grid connection requirements will be allocated to the hydropower units and lithium batteries as a reference power for smoothing. The adaptive moving average algorithm process is as follows: Figure 2 shown.

[0110] Step 3: Use double-layer WOA-VMD and Hilbert to decompose the high and low frequencies of the smooth reference power. Use WOA-VMD to adaptively decompose the smooth reference power, and input the optimal K and α into VMD to obtain K modal component IMFs. The WOA algorithm optimizes VMD parameters flow chart as shown below: Figure 3 shown.

[0111] The specific steps are as follows:

[0112] Step 3.1: Input wind leveling reference power P h (t), set the parameter range of K and a in the VMD algorithm, and initialize various parameters in the WOA model, including population size, maximum number of iterations, spatial dimension, and initial population individuals.

[0113] Step 3.2: Smoothing reference power P h (t) VMD decomposition is performed, and the fitness of each individual in the initial population is calculated using formula (1). As the fitness function of this method, permutation entropy is used to measure the decomposition effect of parameter combinations. The smaller the entropy, the more regular the time series distribution is, indicating that the IMF obtained by VMD processing contains more effective information; conversely, the closer the time series is to random distribution, the more noise components there are in the IMF. Therefore, when the permutation entropy is the smallest, the corresponding parameters are optimal.

[0114] Smoothing reference power P h The time corresponding to (t) is a discrete time series {x(i), i = 1 to N}, with a total of N moments. The phase space of this series is reconstructed to obtain the following reconstruction matrix:

[0115]

[0116] Where: r = N-(m-1)τ, r is the number of reconstructed row vectors; j is the j-th row component of the reconstructed matrix, j = 1 to r; m is the embedding dimension; τ is the delay time;

[0117] Arrange the r row vectors in the reconstruction matrix in ascending order according to the element values ​​to obtain their respective symbol sequences S(q)=(j1,j2,···,jm); where q=1~r, r≤m!; m! is the total number of symbol sequences mapped in the m-dimensional phase space; j1,j2,···,jm represent the index number of each element in the original reconstruction component;

[0118] Calculate the probability P1, P2, ···, P of r different symbol sequences S(q) appearing r , then the permutation entropy of the time series {x(i), i=1~N} is defined as:

[0119]

[0120] Where: H p is the permutation entropy; P i is the probability of occurrence of each sequence;

[0121] Step 3.3: When |A|≤1, select the whale position corresponding to the minimum permutation entropy as the target value of local development, and then select formula (2) to update the position of the individual whale according to p; when |A|>1, randomly select a whale position, and update the position of the individual whale according to formula (3), retaining the optimal fitness and the corresponding parameter combination.

[0122]

[0123] X(t+1)=X rand (t)-A·D(3)

[0124] in,

[0125] Where: X is the current target position; X * is the whale position; A is the coefficient matrix; a is the convergence factor; r1 is a number randomly distributed on [0,1]; tmax is the maximum number of iterations; D = |X * (t)-X(t)| represents the distance from the whale to the prey, D' is the transpose of D; b is a spiral constant, which is 1; l is a random number in (-1,1); X rand (t) is the randomly selected whale position; p is a number randomly distributed on [0,1]; t is the tth iteration;

[0126] Step 3.4: Keep the updated whale population position as the initial population for the new round and iterate until the maximum number of iterations is reached.

[0127] Step 3.5: Output the optimal whale individual and its corresponding fitness.

[0128] Step 3.6: Determine whether the maximum number of iterations is exceeded. If not, return to Step 3.2; if exceeded, output the best parameter combination.

[0129] Step 4: Perform Hilbert marginal spectrum analysis on IMF and determine the cut-off frequency based on the aliasing degree of Hilbert marginal spectrum. The corresponding frequency value is used as the basis for frequency division of IMF. The high and low frequency signals obtained by decomposition are reconstructed and allocated to hydropower units and lithium batteries for smoothing.

[0130] The mathematical model corresponding to the smooth reference power allocation strategy is as follows:

[0131]

[0132] Where M is the high-frequency and low-frequency dividing point of VMD; K is the number of modal components generated by VMD decomposition; N is the Nth IMF; P b (t) is the output power of the lithium battery; P s (t) is the power generation capacity of the turbine unit.

[0133] Application Examples

[0134] Taking the measured data of wind and solar power output as an example, the wind power generation system and photovoltaic power generation system have a total capacity of 350MW.

[0135] Specific steps are as follows Figure 1 As shown:

[0136] Step 1: Build a wind, solar, water and storage integrated power generation system. The power generation system includes wind power generation system, photovoltaic power generation system, hydro turbine and lithium battery. Figure 1 shown.

[0137] Step 2: According to the 1-minute active power and 10-minute active power change requirements of GB / T19963-2021 "Technical Regulations for Wind Farm Access to Power System" and GB / T19964-2024 "Technical Regulations for Photovoltaic Power Station Access to Power System", the adaptive moving average algorithm is used to obtain the power that meets the grid connection requirements and the reference power for smoothing. W (t) and photovoltaic power generation power P V (t) Perform moving average filtering to obtain the output power P that meets the grid connection conditions. g (t) Directly connected to the grid, if the grid connection conditions are not met, it will be used as the reference power P h (t) is allocated to the turbine and lithium battery. The adaptive moving average algorithm process is as follows: Figure 2 shown.

[0138] Step 3: Use the double-layer WOA-VMD to decompose the high and low frequencies of the smooth reference power. Use WOA-VMD to adaptively decompose the smooth reference power and obtain 8 modal component IMFs. The WOA algorithm optimizes the VMD parameter flow chart as shown below: Figure 3 shown.

[0139] The specific steps are as follows:

[0140] Step 3.1: Input the smoothing reference power P h (t), set the range of K and α in the VMD algorithm, K is the upper limit of the interval of VMD decomposition layers, and is [3, 15]. The parameter range of α is [1000, 3000], and initialize various parameters in the WOA model, including population size, maximum number of iterations, spatial dimension, and initial population individuals.

[0141] Step 3.2: Smooth the input reference power P h (t) VMD decomposition is performed, and the fitness of each individual in the initial population is calculated using formula (1). As the fitness function of this method, permutation entropy is used to measure the decomposition effect of parameter combinations. The smaller the entropy, the more regular the time series distribution is, indicating that the IMF obtained by VMD processing contains more effective information; conversely, the closer the time series is to random distribution, the more noise components there are in the IMF. Therefore, when the permutation entropy is the smallest, the corresponding parameters are optimal.

[0142] Smoothing reference power P h The time corresponding to (t) is a discrete time series {x(i), i = 1 to N}, with a total of N moments. The phase space of this series is reconstructed to obtain the following reconstruction matrix:

[0143]

[0144] Where: r = N-(m-1)τ, r is the number of reconstructed row vectors; j is the j-th row component of the reconstructed matrix, j = 1 to r; m is the embedding dimension; τ is the delay time;

[0145] Arrange the r row vectors in the reconstruction matrix in ascending order according to the element values ​​to obtain their respective symbol sequences S(q)=(j1,j2,···,jm); where q=1~r, r≤m!; m! is the total number of symbol sequences mapped in the m-dimensional phase space; j1,j2,···,jm represent the index number of each element in the original reconstruction component;

[0146] Calculate the probability P1, P2, ···, P of r different symbol sequences S(q) appearing r , then the permutation entropy of the time series {x(i), i=1~N} is defined as:

[0147]

[0148] Where: H p is the permutation entropy; P i is the probability of occurrence of each sequence;

[0149] When the permutation entropy is the smallest, the corresponding parameters are optimal;

[0150] Step 3.3: When |A|≤1, select the whale position corresponding to the minimum permutation entropy as the target value of local development, and then select formula (2) to update the position of the individual whale according to p; when |A|>1, randomly select a whale position, and update the position of the individual whale according to formula (3), retaining the optimal fitness and the corresponding parameter combination.

[0151]

[0152] X(t+1)=X rand (t)-A·D(3)

[0153] in,

[0154] Where: X is the current target position; X * is the whale position; A is the coefficient matrix; a is the convergence factor; r1 is a number randomly distributed on [0,1]; tmax is the maximum number of iterations; D = |X * (t)-X(t)| represents the distance from the whale to the prey, D' is the transpose of D; b is a spiral constant, which is 1; l is a random number in (-1,1); X rand (t) is the randomly selected whale position; p is a number randomly distributed on [0,1]; t is the tth iteration;

[0155] Step 3.4: Keep the updated whale population position as the initial population for the new round.

[0156] Step 3.5: Output the optimal whale individual and its corresponding fitness.

[0157] Step 3.6: Determine whether the maximum number of iterations has been exceeded. If not, return to Step 3.2; if exceeded, output the best parameter combination (K = 8, α = 2051). The IMF after WOA-VMD decomposition is as follows: Figure 4 shown.

[0158] Step 4: Perform Hilbert marginal spectrum analysis on IMF. According to the aliasing degree of Hilbert marginal spectrum, the modal aliasing between IMF2 and IMF3 is particularly slight. Therefore, IMF1 to IMF2 are assigned to the hydropower unit as low-frequency components for smoothing. IMF3 to IMF8 are assigned to the lithium battery as high-frequency components for smoothing. The Hilbert marginal spectrum of each IMF is as follows: Figure 5 The power comparison before and after the wind and solar power are leveled is shown in Figure 6 shown.

[0159] The above shows and describes the basic principles, main features and advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for smoothing wind and solar power fluctuations based on a hydro turbine unit and energy storage, characterized in that: The steps include: Step 1: Construct a wind-solar-water-storage integrated power generation system; the wind-solar-water-storage integrated power generation system includes a wind power generation system, a photovoltaic power generation system and a suppression system, and the suppression system includes a turbine unit and a lithium battery energy storage device; Step 2: Perform adaptive moving average filtering on the output power of the wind power generation system and the photovoltaic power generation system to obtain the power P that meets the grid connection requirements. g (t) Smoothing reference power P h (t); Step 3: Use Whale WOA to optimize the VMD parameter modal decomposition number K and quadratic penalty factor α, and use the VMD decomposition after parameter optimization to smooth the reference power and obtain the modal components; Step 4: Perform Hilbert marginal spectrum analysis on each modal component obtained by VMD decomposition, determine the high and low frequency components according to the degree of modal aliasing, allocate the low frequency signal to the turbine unit, and the high frequency signal to the lithium battery energy storage device to smooth the wind and solar power.

2. The method for smoothing wind and solar power fluctuations based on a hydro turbine and energy storage according to claim 1 is characterized in that: In Step 1, the wind power generation system and the photovoltaic power generation system are connected to the AC bus, the lithium battery energy storage device is connected to the AC bus through a DC / AC converter, the hydro-turbine generator set is connected to the AC bus, and the AC bus is connected to the grid through a step-up transformer; Power balance is required in the system, and the power balance relationship is as follows: P h (t)=P W (t)+P V (t)-P g (t) (1) P h (t)=P b (t)+P s (t) (2) Where P W (t) is the output power of the wind power generation system; P V (t) is the output power of the photovoltaic power generation system; P g (t) is the power required for grid connection; P h (t) is the reference power for smoothing; P b (t) is the output power of the lithium battery; P s (t) is the output power of the turbine unit.

3. The method for smoothing wind and solar power fluctuations based on a hydro turbine and energy storage according to claim 1 is characterized in that: The specific method of Step 2 is: According to GB / T19963-2021 "Technical Regulations for Wind Farm Access to Power System" and GB / T 19964-2024 "Technical Regulations for Photovoltaic Power Station Access to Power System", the 1-minute active power and 10-minute active power change requirements are respectively set for the wind power generation system output power P W (t), photovoltaic power generation system output power P V (t) Perform moving average filtering, and directly connect the output power that meets the grid connection conditions to the grid. The output power that does not meet the grid connection conditions will be allocated to the turbine unit and lithium battery as a smoothing reference power.

4. The method for smoothing wind and solar power fluctuations based on a hydro turbine and energy storage according to claim 1 is characterized in that: The specific method of Step 3 is: Step 3.1: Input the smoothing reference power P h (t), set the parameter range of K and α in the VMD algorithm and initialize the parameters in the WOA model. The parameters in the WOA model include population size, maximum number of iterations, spatial dimension, and initial population individuals; Step 3.2: Smoothing reference power P h (t) Perform VMD decomposition; smooth the reference power P h The time corresponding to (t) is a discrete time series {x(i), i = 1 to N}, with a total of N moments. The phase space of this series is reconstructed to obtain the following reconstruction matrix: Where: r = N-(m-1)τ, r is the number of reconstructed row vectors; j is the j-th row component of the reconstructed matrix, j = 1 to r; m is the embedding dimension; τ is the delay time; Arrange the r row vectors in the reconstruction matrix in ascending order according to the element values ​​to obtain their respective symbol sequences S(q)=(j1,j2,···,jm); where q=1~r, r≤m!; m! is the total number of symbol sequences mapped in the m-dimensional phase space; j1,j2,···,jm represent the index number of each element in the original reconstruction component; Calculate the probability P1, P2, ···, P of r different symbol sequences S(q) appearing r , then the permutation entropy of the time series {x(i), i=1~N} is defined as: Where: H p is the permutation entropy; P i is the probability of occurrence of each sequence; When the permutation entropy is the smallest, the corresponding parameters are optimal; Step 3.3: When |A|≤1, select the whale position corresponding to the minimum permutation entropy as the target value of local development, and then select formula (2) to update the position of the individual whale according to the randomly distributed number p; when |A|>1, randomly select a whale position, and update the position of the individual whale according to formula (3), retaining the optimal fitness and the corresponding parameter combination; the details are as follows: X(t+1)=X rand (t)-A·D(3) in, Where: X is the current target position; X * is the whale position; A is the coefficient matrix; a is the convergence factor; r1 is a number randomly distributed on [0,1]; tmax is the maximum number of iterations; D = |X * (t)-X(t)| represents the distance from the whale to the prey, D' is the transpose of D; b is a spiral constant; l is a random number in (-1,1); X rand (t) is the randomly selected whale position; p is a number randomly distributed on [0,1]; t is the tth iteration; Step 3.4: Keep the updated whale population position as the initial population for the new round; Step 3.5: Output the optimal whale individual and its corresponding fitness; Step 3.6: Determine whether the maximum number of iterations is exceeded; if not, return to Step 3.2; if exceeded, output the optimal parameter modal decomposition number K and quadratic penalty factor α combination; Step 3.7: Input the optimal parameter modal decomposition number K and the quadratic penalty factor α obtained in Step 3.6 into VMD to obtain K modal component IMFs.

5. The method for smoothing wind and solar power fluctuations based on a hydro turbine and energy storage according to claim 4 is characterized in that: b=1;K=8.

6. The method for smoothing wind and solar power fluctuations based on a hydro turbine and energy storage according to claim 5 is characterized in that: The specific method of Step 4 is: The Hilbert marginal spectrum analysis is performed on the IMF, and according to the aliasing degree of the Hilbert marginal spectrum, IMF1-IMF2 are allocated to the turbine unit as low-frequency components for smoothing; IMF3-IMF8 are allocated to the lithium battery as high-frequency components for smoothing.

Images