Wind-photovoltaic-hydro storage coordinated generation modeling method based on multi-time scale feature decomposition

Abstract

The application relates to the technical field of power systems, and particularly discloses a wind-solar-water-storage coordinated output modeling method based on multi-time-scale feature decomposition, which comprises the following steps: S1, acquisition and preprocessing of multi-source output and meteorological data; S2, adaptive multi-time-scale decomposition based on meteorological driving, which comprises a mixed decomposition framework of EWT, a meteorological driving mechanism and VMD; the framework realizes adaptive multi-scale decomposition on the multi-energy power signal through three-level nesting of frequency domain identification, meteorological adjustment and time-frequency decomposition; S3, construction of an energy coordination mapping matrix; and S4, coordinated output modeling and reconstruction. The application has the advantages that joint modeling and complementary feature quantization of wind power, photovoltaic power, hydropower and energy storage systems are realized, and support is provided for optimal configuration and safe and efficient operation of a wind-solar-water-storage integrated base.

Description

Technical Field

[0001] This invention relates to the field of power system technology, and in particular to a method for modeling the coordinated power output of wind, solar, hydro, and storage systems based on multi-timescale feature decomposition. Background Technology

[0002] In power systems with a high proportion of renewable energy, wind and solar power generation exhibit typical intermittency and randomness. Their power output is significantly affected by meteorological conditions (wind speed, irradiance, temperature, etc.), displaying strong non-stationarity and high volatility. With the rapid growth of renewable energy installed capacity, these output characteristics place higher demands on the real-time balancing, frequency regulation, peak shaving, and reserve dispatch of the power grid. Hydropower and energy storage, as energy types with controllable and adjustable capabilities, can compensate for wind and solar fluctuations over time, thereby achieving dynamic coordination and energy complementarity among different energy sources. How to quantitatively characterize the dynamic coupling relationship between multiple energy sources and establish a collaborative output modeling method that reflects the overall smoothness of the system has become an important scientific problem and engineering requirement for constructing high-proportion clean energy systems.

[0003] Existing research on the synergistic power output modeling of multi-energy systems mainly focuses on two approaches: one is power simulation models based on physical mechanisms, which calculate the output of a single energy source using meteorological driving formulas, such as wind power models based on wind speed-power curves, photovoltaic models corrected for irradiance-temperature, and hydropower output models based on water level difference and flow rate. These models can accurately describe the power generation characteristics of a single energy source, but they often fail to reflect the temporal interaction characteristics between different energy sources in multi-energy coupling scenarios. The other approach is synergistic analysis models based on statistical correlation or probabilistic joint distribution, such as using the Copula function to describe the joint distribution characteristics of wind and solar power output, or using correlation coefficients such as Pearson and Spearman to assess the complementarity between wind, solar, and hydropower. However, these methods are usually limited to a single time scale or static statistical analysis, and cannot reveal the synergistic evolution of multi-energy output at different time scales.

[0004] Furthermore, some studies have attempted to use signal processing methods to decompose the power output sequences of new energy sources into time-scale components, such as wavelet packet decomposition (WPD) and empirical mode decomposition (EMD), to extract multi-scale fluctuation characteristics of power output. These methods can reflect the short-term fluctuations and long-term trends of wind and solar power to some extent, but their decomposition levels and parameters are usually fixed, lacking adaptive responses to meteorological non-stationarity and failing to integrate with the time-varying characteristics of regulating energy sources such as hydropower and energy storage. Moreover, existing decomposition results mostly remain at the level of single-energy analysis, lacking a systematic characterization of energy flow relationships between multi-energy systems.

[0005] In practical applications, collaborative output modeling is not only an intermediate step in power prediction but also a crucial foundation for complementarity analysis, capacity allocation, and scheduling optimization. By modeling the output characteristics of multi-energy systems in the time and frequency domain, the synergistic and offsetting patterns among wind, solar, hydro, and storage systems can be revealed, providing a quantitative basis for determining the capacity ratio of different energy sources, the scale of energy storage configuration, and output smoothing control strategies. However, a unified modeling method that can simultaneously consider meteorological driving characteristics, multi-timescale output decomposition, and multi-energy collaborative mapping is currently lacking. Existing technologies struggle to accurately reflect the dynamic complementary mechanisms of wind, solar, hydro, and storage systems at different timescales and cannot achieve collaborative reconfiguration across energy outputs or smoothing modeling at the system level.

[0006] Therefore, it is necessary to propose a collaborative output modeling method that is oriented towards multiple time scales and integrates meteorological feature-driven and energy coupling mapping, so as to realize the joint modeling and complementary feature quantification of wind power, photovoltaic, hydropower and energy storage systems, and provide support for the optimized configuration and safe and efficient operation of the "wind-solar-hydro-storage" integrated base. Summary of the Invention

[0007] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a wind-solar-hydro-storage synergistic output modeling method based on multi-timescale feature decomposition. This aims to solve three major problems in existing "wind-solar-hydro-storage" multi-energy system output modeling: First, existing output modeling methods are mostly limited to single-energy sources or static statistical analysis, failing to depict the dynamic synergistic patterns of different energy sources at multiple timescales; second, traditional signal decomposition methods have a fixed number of decomposition layers and are insensitive to meteorological changes, making it difficult to adapt to the non-stationary characteristics of external driving conditions such as wind speed, irradiance, and runoff; third, there is a lack of a unified modeling framework that can comprehensively reflect the energy flow relationships among multiple energy sources, making it difficult to use the synergistic modeling results for complementarity analysis and capacity optimization.

[0008] The objective of this invention is achieved through the following technical solution: a wind-solar-hydro-storage coordinated power output modeling method based on multi-timescale feature decomposition, which includes the following steps:

[0009] S1. Acquisition and preprocessing of multi-source power output and meteorological data; acquiring power output sequences of wind power, photovoltaic, hydropower and energy storage systems and their corresponding meteorological driving data, performing normalization and outlier removal to obtain standardized multi-energy output sequences.

[0010] S2. Meteorological-driven adaptive multi-timescale decomposition, which includes constructing a hybrid decomposition framework of EWT, meteorological driving mechanism and VMD; the framework performs adaptive multi-scale decomposition of multi-energy power signals through a three-level nesting of frequency domain identification, meteorological regulation and time-frequency decomposition.

[0011] S3. Construct an energy synergy mapping matrix; For each time scale, integrate the correlation coefficient index, mutual information index and dynamic time warping (DTW) similarity index to construct an energy synergy mapping matrix for quantitatively characterizing the synergy intensity between different energy sources.

[0012] S4. Collaborative output modeling and reconstruction: Calculate the collaborative score of each energy source based on the energy collaborative mapping matrix, and generate the fusion weight of each energy source at each time scale by combining the minimum variance constraint. Use the fusion weight to weight and superimpose the output characteristic modes of each energy source to reconstruct the system-level multi-energy collaborative output curve.

[0013] Specifically, the acquisition and preprocessing of multi-source power output and meteorological data includes acquiring power output sequences from wind power, photovoltaic, hydropower, and energy storage systems. and its corresponding meteorological driving data The data is normalized to a uniform time resolution. and through Outliers are eliminated using principles or box gating to obtain a standardized output sequence:

[0014] ;

[0015] In the formula, For time energy Normalized output; For time energy The original output; Energy The mean and standard deviation of the output.

[0016] Specifically, step S2 includes,

[0017] S21. Frequency domain identification is performed through EWT frequency band adaptive analysis;

[0018] First, perform a fast Fourier transform on the normalized output:

[0019] ;

[0020] The initial frequency band boundary is determined based on the local minima of the energy spectral density.

[0021] ;

[0022] In the formula, The Fourier spectrum of the signal; For the set of frequency band boundaries; For the first The dividing frequency of each frequency band; The initial number of decomposition layers determined for EWT;

[0023] S22. Determination of the number of adaptive decomposition layers driven by meteorological conditions;

[0024] The meteorological change rate is defined as:

[0025] ;

[0026] The formula for calculating the number of weather-driven adaptive layers is:

[0027] ;

[0028] In the formula, For energy Corresponding meteorological driving variables; For meteorological change rate; β is the initial decomposition layer number; β is the meteorological sensitivity parameter. For variance operators;

[0029] S23, VMD signal decomposition and mode selection;

[0030] Based on the determined number of layers Variational mode decomposition is used to decompose the normalized signal into several eigenmode functions with finite bandwidth:

[0031] ;

[0032] In the formula, For energy The decomposition yielded the first The intrinsic mode function represents the signal at the th eigenmode function. Fluctuation components at each time scale;

[0033] Find several mode functions with finite bandwidth in the time-frequency domain that minimize the total bandwidth:

[0034] ;

[0035] In the formula, Indicates an energy type index. These correspond to wind, solar, water, and storage, respectively. Indicates the modal layer index. ; The center angular frequency of this modal component; It is a Dirac impulse function; It is the imaginary unit; It is the time derivative operator; yes Norm;

[0036] After decomposition, the effective modes are screened using the spectral energy concentration index:

[0037] ;

[0038] In the formula, if If a mode is identified as noise, it is discarded; finally, the set of valid modes is obtained.

[0039] ;

[0040] In the formula, Modal components Fourier transform; For the first Energy percentage of each mode; Energy threshold; This represents the number of effective modes.

[0041] Specifically, step S3 includes the following steps:

[0042] S31, Modal scale setting;

[0043] For each type of energy Effective mode set Each mode This represents the energy in the Fluctuation components at each time scale;

[0044] S32, Calculation of the cooperative mapping matrix;

[0045] At each time scale Below, define the elements of the co-mapping matrix. for:

[0046] ;

[0047] In the formula, The correlation coefficient is an indicator. Mutual information indicators; The similarity index for Dynamic Time Warped (DTW); The weighting coefficients for the three indicators are: ;

[0048] S33. Calculation of Dynamic Time Warping (DTW) similarity index;

[0049] Define the local distance matrix The minimum cumulative distance is:

[0050] ;

[0051] In the formula, A set of time-aligned paths; For local Euclidean distance; The smaller the value, the stronger the collaboration.

[0052] The formula for calculating the Dynamic Time Warping (DTW) similarity index is as follows:

[0053] ;

[0054] In the formula, This is the distance attenuation coefficient;

[0055] S34, System Comprehensive Coordination Index;

[0056] Synergy matrix at various time scales Combined into system-level indicators:

[0057] ;

[0058] In the formula, It is a matrix The largest eigenvalue; It is a comprehensive system synergy index.

[0059] Specifically, in step S32, the correlation coefficient index for:

[0060] ;

[0061] In the formula, , Energy and energy In scale Next time The fluctuation components, Energy and energy In scale Modal mean under; Positive values ​​indicate positive collaboration, while negative values ​​indicate negative collaboration.

[0062] Mutual information indicators for:

[0063] ;

[0064] In the formula, For modality In the amplitude range The probability distribution within; For modality In the amplitude range The probability distribution within; It is the joint probability distribution of the two;

[0065] For normalization purposes, we define standardized mutual information:

[0066] ;

[0067] In the formula, This refers to information entropy.

[0068] Specifically, in step S32, let the first... The modal energy ratios at each time scale are:

[0069] ;

[0070] but, , and They are respectively:

[0071] ;

[0072] ;

[0073] ;

[0074] In the formula, For the first The proportion of system energy at each time scale; This is an adjustment coefficient, ranging from 0.1 to 0.3, used to balance linear and nonlinear components.

[0075] Specifically, step S4 includes,

[0076] S41. Calculate the basic weights of collaborative driving;

[0077] Based on the cooperative mapping matrix Computational Energy Collaborative score:

[0078] ;

[0079] Standardize it as a basic item for collaborative weights:

[0080] ;

[0081] In the formula, For energy Compared with other energy sources on time scales The sum of the synergistic strengths; For energy In time scale The basic weights driven by collaboration; The total number of energy types;

[0082] S42. Calculate the minimum variance combination weights;

[0083] In scale Internal Construct the covariance matrix:

[0084] ;

[0085] In the formula, For energy and In time scale Covariance on;

[0086] Based on the idea of ​​minimum variance combination, the combination weights that minimize the system variance are obtained:

[0087] ;

[0088] ;

[0089] In the formula, For energy The minimum variance combined weights; For dimension A vector of all 1s The vector represents the first Quantity; Let covariance be the matrix formed by all of them. composition.

[0090] S43, Energy and smoothness scale weight settings;

[0091] Define time scale System energy percentage:

[0092] ;

[0093] In the formula, Representing time scale The modal energy ratio, This represents the total number of effective modes, i.e., the number of time scales. For energy In scale The fluctuation component below; For energy In scale The fluctuation component below;

[0094] Define the total variance of the scale:

[0095] ;

[0096] Introducing a smoothness adjustment term:

[0097] ;

[0098] S44. In-scale weights that integrate multiple criteria;

[0099] At each time scale Internally, the synergistic enhancement and volatility suppression weights are integrated:

[0100] ;

[0101] In the formula, Indicates energy In scale The comprehensive integration weight, The fusion coefficient;

[0102] Based on scale energy and smoothness, the overall importance across scales is weighted to obtain the global weights used for the final reconstruction:

[0103] ;

[0104] In the formula, Time scale Smooth energy weights; For energy In time scale The final reconstructed weights;

[0105] S45, Cooperative output reconfiguration and constraint correction;

[0106] Obtain the final reconfiguration weights of each energy source at different time scales. Then, all modal components are weighted and superimposed to obtain the overall coordinated output of the system:

[0107] ;

[0108] In the formula, To contribute to the multi-functional synergy of the system;

[0109] Define the upper limit of the outbound channel capacity as The corrected output is:

[0110] ;

[0111] In the formula, This represents the actual output power of the system after constraint correction; This is the upper limit of the power transmission channel capacity; It is the proportional gain coefficient, used to correct the system output amplitude or match the target planning curve.

[0112] The present invention has the following advantages:

[0113] 1. This invention utilizes a combination of Empirical Wavelet Transform (EWT) and Variational Mode Decomposition (VMD) to decompose normalized multi-energy output signals. It introduces meteorological change rate as a driving factor to adaptively determine the number of decomposition layers and frequency band division, achieving precise extraction of energy power fluctuation characteristics under different meteorological conditions. This method can simultaneously capture short-term fluctuations and medium-to-long-term changes in wind, solar, hydro, and storage power output, achieving consistent modeling across time scales.

[0114] 2. This invention integrates linear correlation coefficients, mutual information, and dynamic time warping (DTW) indices to construct an energy synergy mapping matrix among multiple energy sources, which is used to quantitatively characterize the synergy strength and lag relationship among wind, solar, hydro, and storage at different time scales. Through modal energy weighting and scale-sensitive adjustment, dynamic adaptive characterization of the synergy relationship at different time resolutions is achieved.

[0115] 3. This invention proposes a weight generation method that combines synergistic enhancement and variance suppression. It extracts synergistic scores among energy sources through a synergistic mapping matrix and combines minimum variance steady-state constraints with scale-based energy smoothing weighting to form a final weight system integrating multiple criteria. This mechanism can adaptively adjust the contribution ratio of each energy source according to the system's operating state, achieving simultaneous optimization of output balance and fluctuation suppression.

[0116] 4. This invention uses multi-scale fusion weights to weight and superimpose various energy modes to form a system-level collaborative output curve; and on this basis, it introduces external transmission channel capacity constraints and proportional correction coefficients to ensure that the system output does not exceed the upper limit of power transmission safety, thereby realizing stable power transmission and friendly dispatch of the integrated "wind-solar-hydro-storage" system. Attached Figure Description

[0117] Figure 1 This is a schematic diagram of the wind-solar-hydro-storage synergistic power output modeling method of the present invention. Detailed Implementation

[0118] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only for explaining the invention and are not intended to limit the invention; that is, the described embodiments are merely some embodiments of the invention, and not all embodiments. The components of the embodiments of the invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0119] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0120] It should be noted that relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0121] The present invention will be further described below with reference to the accompanying drawings, but the scope of protection of the present invention is not limited to the following description.

[0122] like Figure 1 As shown, a wind-solar-hydro-storage collaborative power output modeling method based on multi-timescale feature decomposition is presented. This method includes the following steps:

[0123] S1. Acquisition and preprocessing of multi-source power output and meteorological data;

[0124] The acquisition and preprocessing of multi-source power output and meteorological data includes acquiring power output sequences from wind power, photovoltaic, hydropower, and energy storage systems. and its corresponding meteorological driving data The meteorological elements include wind speed, irradiance, temperature, water inflow, and rainfall. The data is then normalized to a uniform time resolution. and through Outliers are eliminated using principles or box gating to obtain a standardized output sequence:

[0125] ;

[0126] In the formula, For time energy Normalized output; For time energy The original output; Energy The mean and standard deviation of the output are determined. This ensures the comparability of multi-energy power sequences and provides a consistent data foundation for subsequent decomposition and synergistic analysis.

[0127] S2. Meteorological-driven adaptive multi-timescale decomposition, which includes constructing a hybrid decomposition framework of EWT, meteorological driving mechanism and VMD; the framework performs adaptive multi-scale decomposition of multi-energy power signals through a three-level nesting of frequency domain identification, meteorological regulation and time-frequency decomposition; in order to extract the fluctuation characteristics of each energy at different time scales, this invention adopts a hybrid decomposition framework of EWT + meteorological driving mechanism + VMD.

[0128] Step S2 specifically includes,

[0129] S21. Frequency domain identification is performed through EWT frequency band adaptive analysis;

[0130] First, perform a Fast Fourier Transform on the power signal:

[0131] ;

[0132] The initial frequency band boundary is determined based on the local minima of the energy spectral density.

[0133] ;

[0134] In the formula, The Fourier spectrum of the signal; For the set of frequency band boundaries; For the first The dividing frequency of each frequency band; The initial decomposition layer number is determined for EWT; this process can determine the main energy distribution region of the signal and provide frequency initialization conditions for VMD.

[0135] S22. Determination of the number of decomposition layers driven by meteorology; This invention introduces the rate of change of meteorological elements, including wind speed gradient, irradiance gradient, and rate of change of water inflow, to dynamically adjust the number of decomposition layers.

[0136] The meteorological change rate is defined as:

[0137] ;

[0138] The formula for calculating the number of weather-driven adaptive layers is:

[0139] ;

[0140] In the formula, For energy Corresponding meteorological driving variables; For meteorological change rate; β is the initial decomposition layer number; β is the meteorological sensitivity parameter. The variance operator is used; this mechanism allows the number of decomposition layers to change dynamically with meteorological fluctuations: when meteorological fluctuations intensify, the number of decomposition layers is automatically increased to capture more complex short-term fluctuation characteristics; conversely, the number of layers is reduced to simplify the model structure.

[0141] S23, VMD signal decomposition and mode selection;

[0142] Based on the determined number of layers Variational mode decomposition is used to decompose the normalized signal into several eigenmode functions with finite bandwidth:

[0143] ;

[0144] In the formula, For energy The decomposition yielded the first The intrinsic mode function represents the signal at the th eigenmode function. Fluctuation components at each time scale;

[0145] The optimization objective of VMD is to find several mode functions with finite bandwidth in the time-frequency domain that minimize the total bandwidth.

[0146] ;

[0147] In the formula, Indicates an energy type index. These correspond to wind, solar, water, and storage, respectively. Indicates the modal layer index. ; The center angular frequency of this modal component; It is a Dirac impulse function; It is the imaginary unit; It is the time derivative operator; yes Norm;

[0148] After decomposition, the effective modes are screened using the spectral energy concentration index:

[0149] ;

[0150] In the formula, if If a mode is identified as noise, it is discarded; finally, the set of valid modes is obtained.

[0151] ;

[0152] In the formula, Modal components Fourier transform; For the first Energy percentage of each mode; The energy threshold is typically between 0.01 and 0.05. The effective number of modes; the set of modes after filtering. This constitutes the output characteristics of the energy source at multiple time scales, which are used for subsequent energy coordination mapping and output reconstruction modeling.

[0153] This invention utilizes a combination of Empirical Wavelet Transform (EWT) and Variational Mode Decomposition (VMD) to decompose normalized multi-energy output signals. It also introduces meteorological variability as a driving factor to adaptively determine the number of decomposition layers and frequency band division, enabling precise extraction of energy power fluctuation characteristics under different meteorological conditions. This method can simultaneously capture both short-term fluctuations and medium-to-long-term changes in wind, solar, hydro, and energy storage outputs, achieving consistent modeling across time scales.

[0154] S3. Construction of the Energy Synergy Mapping Matrix: To quantitatively characterize the synergistic relationships of different energy sources across multiple time scales, this invention constructs an energy synergy mapping matrix. By integrating three feature indicators—correlation, mutual information, and dynamic time warping (DTW) similarity—and combining modal energy and scale weights, a comprehensive evaluation of the multi-dimensional synergistic strength between wind, solar, water, and storage can be achieved.

[0155] The specific steps of step S3 include:

[0156] S31, Modal scale setting;

[0157] For each type of energy Effective mode set Each mode This represents the energy in the Fluctuation components at various time scales, such as short-term, intraday, weekly, and seasonal;

[0158] S32, Calculation of the cooperative mapping matrix;

[0159] At each time scale Below, define the elements of the co-mapping matrix. for:

[0160] ;

[0161] In the formula, The correlation coefficient is an indicator. Mutual information indicators; The similarity index for Dynamic Time Warped (DTW); The weighting coefficients for the three indicators are: ;

[0162] Correlation coefficient index for:

[0163] ;

[0164] In the formula, , Energy and energy In scale The fluctuation components below, Energy and energy In scale Modal mean under; Positive values ​​indicate positive collaboration, while negative values ​​indicate negative collaboration.

[0165] Mutual Information Mutual information index reflects the nonlinear dependence of two modes on their probability distributions. for:

[0166] ;

[0167] In the formula, For modality In the amplitude range The probability distribution within; For modality In the amplitude range The probability distribution within; It is the joint probability distribution of the two;

[0168] For normalization purposes, we define standardized mutual information:

[0169] ;

[0170] In the formula, This refers to information entropy.

[0171] This invention uses modal energy weighting and scale-sensitive adjustment mechanisms to determine the weights. Let the first The modal energy ratios at each time scale are:

[0172] ;

[0173] but, , and They are respectively:

[0174] ;

[0175] ;

[0176] ;

[0177] In the formula, For the first The proportion of system energy at each time scale; The adjustment coefficient is set to 0.1-0.3 to balance linear and nonlinear components. This results in a larger weight for linear correlation at high energy (principal scale) and an increased weight for mutual information and DTW at low energy or noise-dominated scale, thus enhancing robustness.

[0178] S33. Calculation of Dynamic Time Warping (DTW) similarity index;

[0179] Define the local distance matrix The minimum cumulative distance is:

[0180] ;

[0181] In the formula, A set of time-aligned paths; For local Euclidean distance; The smaller the value, the stronger the collaboration.

[0182] The formula for calculating the Dynamic Time Warping (DTW) similarity index is as follows:

[0183] ;

[0184] In the formula, This is the distance attenuation coefficient, typically taken as 0.1–0.3;

[0185] S34, System Comprehensive Coordination Index;

[0186] Synergy matrix at various time scales Combined into system-level indicators:

[0187] ;

[0188] In the formula, It is a matrix The largest eigenvalue; It is a comprehensive system synergy index.

[0189] This invention integrates linear correlation coefficients, mutual information, and dynamic time warping (DTW) indices to construct an energy synergy mapping matrix among multiple energy sources, which quantitatively characterizes the synergy strength and lag relationship among wind, solar, hydro, and storage at different time scales. Through modal energy weighting and scale-sensitive adjustment, a dynamic adaptive characterization of the synergy relationship at different time resolutions is achieved.

[0190] S4. Collaborative output modeling and reconstruction: This invention constructs time-invariant or slowly time-varying weighting coefficients based on the collaborative relationship and fluctuation characteristics between various energy sources at multiple time scales, and reconstructs the collaborative output curve of the system to achieve the effects of fluctuation suppression and complementary enhancement.

[0191] Step S4 specifically includes:

[0192] S41. Calculate the basic weights of collaborative driving;

[0193] First at each time scale Based on the cooperative mapping matrix Computational Energy Collaborative score:

[0194] ;

[0195] Standardize it as a basic item for collaborative weights:

[0196] ;

[0197] In the formula, For energy Compared with other energy sources on time scales The sum of the synergistic strengths; For energy In time scale The basic weights driven by collaboration; The total number of energy types is usually four, including wind, solar, hydro, and energy storage.

[0198] S42. Calculate the minimum variance combination weights;

[0199] To further suppress system fluctuations, at the scale Internal Construct the covariance matrix:

[0200] ;

[0201] In the formula, For energy and In time scale The larger the covariance, the more consistent the fluctuations of the two, and the more significant the impact on the overall fluctuation.

[0202] Based on the idea of ​​minimum variance combination, the combination weights that minimize the system variance are obtained:

[0203] ;

[0204] ;

[0205] In the formula, For energy The minimum variance combined weights; For dimension A vector of all 1s The vector represents the first Quantity; Let covariance be the matrix formed by all of them. composition.

[0206] S43, Energy and smoothness scale weight settings;

[0207] To measure the overall contribution at different time scales, a time scale is defined. System energy percentage:

[0208] ;

[0209] In the formula, Representing time scale The modal energy proportion reflects the degree to which power fluctuations dominate at that time scale. This represents the total number of effective modes, i.e., the number of time scales. For energy In scale The fluctuation component below; For energy In scale The fluctuation component below;

[0210] Define the total variance of the scale:

[0211] ;

[0212] Introducing a smoothness adjustment term:

[0213] ;

[0214] in, The larger the value, the lower the volatility (the smoother) of that scale, and the higher its weight in the reconstruction.

[0215] S44. In-scale weights that integrate multiple criteria;

[0216] At each time scale Internally, the synergistic enhancement and volatility suppression weights are integrated:

[0217] ;

[0218] In the formula, Indicates energy In scale The comprehensive integration weight, The fusion coefficient; The larger the value, the more emphasis is placed on stability; The smaller the value, the greater the emphasis on energy synergy;

[0219] Then, based on scale energy and smoothness, the overall importance across scales is weighted to obtain the final global weights used for reconstruction:

[0220] ;

[0221] In the formula, Time scale Smooth energy weights; For energy In time scale The final reconstructed weights;

[0222] This invention proposes a weight generation method that combines synergistic enhancement and variance suppression. It extracts synergistic scores among energy sources through a synergistic mapping matrix and combines minimum variance steady-state constraints with scale-based energy smoothing weighting to form a final weighting system that integrates multiple criteria. This mechanism can adaptively adjust the contribution ratio of each energy source according to the system's operating state, achieving simultaneous optimization of output balance and fluctuation suppression.

[0223] S45, Cooperative output reconfiguration and constraint correction;

[0224] Obtain the final reconfiguration weights of each energy source at different time scales. Then, all modal components are weighted and superimposed to obtain the overall coordinated output of the system:

[0225] ;

[0226] In the formula, To contribute to the multi-functional synergy of the system;

[0227] To meet the requirements for safe operation of the power grid, the coordinated output of the system must be constrained by the maximum power transmission capacity. Therefore, the upper limit of the external transmission channel capacity is defined as follows: The corrected output is:

[0228] ;

[0229] In the formula, This represents the actual output power of the system after constraint correction; This is the upper limit of the power transmission channel capacity; It is the proportional gain coefficient, used to correct the system output amplitude or match the target planning curve.

[0230] This invention uses multi-scale fusion weights to weight and superimpose various energy modes to form a system-level collaborative output curve. On this basis, it introduces external transmission channel capacity constraints and proportional correction coefficients to ensure that the system output does not exceed the upper limit of power transmission safety, thereby realizing stable power transmission and friendly dispatch of the integrated "wind-solar-hydro-storage" system.

[0231] The above description is merely a preferred embodiment of the present invention and does not constitute any limitation on the present invention. Any person skilled in the art can make many possible variations and modifications to the technical solution of the present invention, or modify it into equivalent embodiments, without departing from the scope of the present invention. Therefore, any modifications, equivalent changes, and alterations made to the above embodiments based on the technology of the present invention without departing from the scope of the present invention are within the protection scope of the present invention.

Claims

1. A method for modeling the coordinated power output of wind, solar, hydro, and storage systems based on multi-timescale feature decomposition, characterized by: The method includes the following steps: S1. Acquisition and preprocessing of multi-source power output and meteorological data; acquire power output sequences of wind power, photovoltaic, hydropower and energy storage systems and their corresponding meteorological driving data, perform normalization and outlier removal to obtain standardized multi-energy power output sequences; S2. Meteorological-driven adaptive multi-timescale decomposition, which includes constructing a hybrid decomposition framework of EWT, meteorological driving mechanism and VMD; the framework performs adaptive multi-scale decomposition of multi-energy power signals through a three-level nesting of frequency domain identification, meteorological regulation and time-frequency decomposition. S3. Construct an energy synergy mapping matrix; For each time scale, integrate the correlation coefficient index, mutual information index and dynamic time warping (DTW) similarity index to construct an energy synergy mapping matrix for quantitatively characterizing the synergy intensity between different energy sources. S4. Collaborative output modeling and reconstruction: Calculate the collaborative score of each energy source based on the energy collaborative mapping matrix, and generate the fusion weight of each energy source at each time scale by combining the minimum variance constraint. Use the fusion weight to weight and superimpose the output characteristic modes of each energy source to reconstruct the system-level multi-energy collaborative output curve.

2. The wind-solar-hydro-storage synergistic power output modeling method based on multi-timescale feature decomposition according to claim 1, characterized in that: S1 specifically includes acquiring the power output sequences of wind power, photovoltaic, hydropower, and energy storage systems. and its corresponding meteorological driving data The data is normalized to a uniform time resolution. and through Outliers are eliminated using principles or box gating to obtain a standardized output sequence: ； In the formula, For time energy Normalized output; For time energy The original output; Energy The mean and standard deviation of the output.

3. The wind-solar-hydro-storage synergistic power output modeling method based on multi-timescale feature decomposition according to claim 1, characterized in that: Step S2 specifically includes, S21. Frequency domain identification is performed through EWT frequency band adaptive analysis; First, perform a fast Fourier transform on the normalized output: ； The initial frequency band boundary is determined based on the local minima of the energy spectral density. ； In the formula, The Fourier spectrum of the signal; For the set of frequency band boundaries; For the first The dividing frequency of each frequency band; The initial number of decomposition layers determined for EWT; S22. Determination of the number of adaptive decomposition layers driven by meteorological conditions; The meteorological change rate is defined as: ； The formula for calculating the number of weather-driven adaptive layers is: ； In the formula, For energy Corresponding meteorological driving variables; For meteorological change rate; β is the initial decomposition layer number; β is the meteorological sensitivity parameter. For variance operators; S23, VMD signal decomposition and mode selection; Based on the determined number of layers Variational mode decomposition is used to decompose the normalized signal into several eigenmode functions with finite bandwidth: ； In the formula, For energy The decomposition yielded the first The intrinsic mode function represents the signal at the th eigenmode function. Fluctuation components at each time scale; Find several mode functions with finite bandwidth in the time-frequency domain that minimize the total bandwidth: ； In the formula, Indicates an energy type index. These correspond to wind, solar, water, and storage, respectively. Indicates the modal layer index. ; The center angular frequency of this modal component; It is a Dirac impulse function; It is the imaginary unit; It is the time derivative operator; yes Norm; After decomposition, the effective modes are screened using the spectral energy concentration index: ； In the formula, if If a mode is identified as noise, it is discarded; finally, the set of valid modes is obtained. ； In the formula, Modal components Fourier transform; For the first Energy percentage of each mode; Energy threshold; This represents the number of effective modes.

4. The wind-solar-hydro-storage synergistic power output modeling method based on multi-timescale feature decomposition according to claim 3, characterized in that: The specific steps of step S3 include: S31, Modal scale setting; For each type of energy Effective mode set Each mode This represents the energy in the Fluctuation components at each time scale; S32, Calculation of the cooperative mapping matrix; At each time scale Below, define the elements of the co-mapping matrix. for: ； In the formula, The correlation coefficient is an indicator. Mutual information indicators; The similarity index for Dynamic Time Warped (DTW); The weighting coefficients for the three indicators are: ; S33. Calculation of Dynamic Time Warping (DTW) similarity index; Define the local distance matrix The minimum cumulative distance is: ； In the formula, A set of time-aligned paths; For local Euclidean distance; The smaller the value, the stronger the collaboration. The formula for calculating the Dynamic Time Warping (DTW) similarity index is as follows: ； In the formula, This is the distance attenuation coefficient; S34, System Comprehensive Coordination Index; Synergy matrix at various time scales Combined into system-level indicators: ； In the formula, It is a matrix The largest eigenvalue; It is a comprehensive system synergy index.

5. The wind-solar-hydro-storage synergistic power output modeling method based on multi-timescale feature decomposition according to claim 4, characterized in that: In step S32, the correlation coefficient index for: ； In the formula, , Energy and energy In scale The fluctuation components below, Energy and energy In scale Modal mean under; Positive values ​​indicate positive collaboration, while negative values ​​indicate negative collaboration. Mutual information indicators for: ； In the formula, For modality In the amplitude range The probability distribution within; For modality In the amplitude range The probability distribution within; It is the joint probability distribution of the two; For normalization purposes, we define standardized mutual information: ； In the formula, This refers to information entropy.

6. The wind-solar-hydro-storage synergistic power output modeling method based on multi-timescale feature decomposition according to claim 4, characterized in that: In step S32, let the first... The modal energy ratios at each time scale are: ； but, , and They are respectively: ； ； ； In the formula, For the first The proportion of system energy at each time scale; This is an adjustment coefficient, ranging from 0.1 to 0.3, used to balance linear and nonlinear components.

7. The wind-solar-hydro-storage synergistic power output modeling method based on multi-timescale feature decomposition according to claim 4, characterized in that: Step S4 specifically includes: S41. Calculate the basic weights of collaborative driving; Energy calculation based on cooperative mapping matrix Collaborative score: ； Standardize it as a basic item for collaborative weights: ； In the formula, For energy Compared with other energy sources on time scales The sum of the synergistic strengths; For energy In time scale The basic weights driven by collaboration; The total number of energy types; S42. Calculate the minimum variance combination weights; In scale Internal Construct the covariance matrix: ； In the formula, For energy and In time scale Covariance on; Based on the idea of ​​minimum variance combination, the combination weights that minimize the system variance are obtained: ； In the formula, For energy The minimum variance combined weights; For dimension A vector of all 1s The vector represents the first Quantity; Let covariance be the matrix formed by all of them. composition. S43, Energy and smoothness scale weight settings; Define time scale System energy percentage: ； In the formula, Representing time scale The modal energy ratio, This represents the total number of valid modes. For energy In scale The fluctuation component below; For energy In scale The fluctuation component below; Define the total variance of the scale: ； Introducing a smoothness adjustment term: ； S44. In-scale weights that integrate multiple criteria; At each time scale Internally, the synergistic enhancement and volatility suppression weights are integrated: ； In the formula, Indicates energy In scale The comprehensive integration weight, The fusion coefficient; Based on scale energy and smoothness, the overall importance across scales is weighted to obtain the global weights used for the final reconstruction: ； In the formula, Time scale Smooth energy weights; For energy In time scale The final reconstructed weights; S45, Cooperative output reconfiguration and constraint correction; Obtain the final reconfiguration weights of each energy source at different time scales. Then, all modal components are weighted and superimposed to obtain the overall coordinated output of the system: ； In the formula, To contribute to the multi-functional synergy of the system; Define the upper limit of the outbound channel capacity as The corrected output is: ； In the formula, This represents the actual output power of the system after constraint correction; This is the upper limit of the power transmission channel capacity; It is the proportional gain coefficient, used to correct the system output amplitude or match the target planning curve.

Images