Collaborative regulation and control optimization method and device for integrated energy system, and electronic equipment

Abstract

The invention provides a collaborative regulation and control optimization method and device for an integrated energy system and electronic equipment, and relates to the technical field of micro-grids. The method comprises the steps of obtaining source load historical data, and generating a large-scale scene set based on the source load historical data; reducing scenes in the large-scale scene set by adopting a backward reduction method to obtain a source-load combined scene set; based on the source-load combined scene set, performing optimization solution by taking the minimum daily comprehensive cost as an objective function to obtain a primary optimization regulation and control scheme; correcting the power of the super capacitor and the power of the lithium battery based on the primary optimization regulation and control scheme by adopting a fuzzy control strategy to obtain a target optimization regulation and control scheme; wherein the target optimization regulation and control scheme comprises an equipment capacity configuration scheme. According to the method, scenes are reduced, the calculation complexity is greatly reduced while key working conditions are covered, and a scheduling scheme is made to adapt to various actual operation working conditions; and meanwhile, two-stage optimization is adopted, so that the optimization difficulty is reduced, and the optimization efficiency is improved.

Description

Technical Field

[0001] This invention relates to the field of microgrid technology, and in particular to a method, apparatus and electronic equipment for coordinated regulation and optimization of integrated energy systems. Background Technology

[0002] Because renewable energy sources are characterized by significant intermittency, volatility, and randomness, and because integrated energy systems involve diverse user load demands, a single energy storage method can hardly simultaneously meet the system's requirements for energy efficiency, cost-effectiveness, and operational flexibility. A multi-energy storage system, composed of energy storage technologies with different equipment characteristics, can, to some extent, compensate for the shortcomings of a single energy storage technology. Multi-energy storage systems can achieve complementary advantages through multi-energy collaborative storage.

[0003] In existing technologies, the application of multi-energy storage in various energy forms to integrated energy systems and distributed energy systems, such as electricity storage-cold / heat storage and electricity storage-cold-heat storage, mostly focuses on the coordination between electrochemical or mechanical energy storage, without fully considering the coordinated control of heterogeneous energy flow devices in multi-energy coupling scenarios of electricity-heat-cold, which limits the full utilization of the flexible adjustment capabilities of renewable energy. Summary of the Invention

[0004] This invention provides a method, apparatus, and electronic device for coordinated regulation and optimization of an integrated energy system, in order to solve the problem that existing control methods do not fully consider the coordinated control of various devices in multiple scenarios, thus limiting the utilization of renewable energy capabilities.

[0005] In a first aspect, embodiments of the present invention provide a method for coordinated regulation and optimization of an integrated energy system, comprising: Acquire historical source load data, and generate a large-scale scene set based on the historical source load data; The back-reduction method is used to reduce the scenes in the large-scale scene set to obtain a source-load joint scene set; Based on the aforementioned source-load joint scenario set, optimization is performed with the objective function of minimizing daily comprehensive cost to obtain a primary optimized control scheme. A fuzzy control strategy is adopted to correct the power of the supercapacitor and the power of the lithium battery based on the primary optimization control scheme, thereby obtaining a target optimization control scheme; wherein, the target optimization control scheme includes a device capacity configuration scheme.

[0006] Secondly, embodiments of the present invention provide a comprehensive energy system coordinated regulation and optimization device, comprising: The initial scene generation module is used to acquire historical source load data and generate a large-scale scene set based on the historical source load data; The scene reduction module is used to reduce the scenes in the large-scale scene set using the backward reduction method to obtain the source-load joint scene set; The initial scheme generation module is used to optimize and solve the initial optimized control scheme based on the source-load joint scenario set, with the objective function of minimizing the daily comprehensive cost. The scheme correction module is used to use a fuzzy control strategy to correct the power of the supercapacitor and the power of the lithium battery based on the primary optimization control scheme to obtain a target optimization control scheme; wherein, the target optimization control scheme includes a device capacity configuration scheme.

[0007] Thirdly, embodiments of the present invention provide an electronic device, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the integrated energy system coordinated regulation and optimization method as described in the first aspect or any possible implementation of the first aspect.

[0008] This invention provides a method, apparatus, and electronic device for the coordinated regulation and optimization of an integrated energy system. The method includes: acquiring historical source-load data; generating a large-scale scenario set based on the historical source-load data; reducing the scenarios in the large-scale scenario set using a backward reduction method to obtain a joint source-load scenario set; optimizing the solution based on the joint source-load scenario set with the objective function of minimizing daily comprehensive cost to obtain a primary optimized regulation scheme; and using a fuzzy control strategy to correct the power of the supercapacitor and the lithium battery based on the primary optimized regulation scheme to obtain a target optimized regulation scheme; wherein the target optimized regulation scheme includes an equipment capacity configuration scheme. This invention reduces the number of scenarios while covering key operating conditions, significantly reducing computational complexity and making the scheduling scheme adaptable to various actual operating conditions, meeting practical application requirements. Simultaneously, the two-stage optimization reduces optimization difficulty, improves optimization efficiency, and makes the optimization scheme more reasonable. Attached Figure Description

[0009] Figure 1 This is a flowchart illustrating the implementation of a comprehensive energy system coordinated regulation and optimization method provided in an embodiment of the present invention; Figure 2 This is a Beta distribution diagram of illumination radiation under different shape parameters provided in the embodiments of the present invention; Figure 3 This is a schematic diagram of the integrated energy system coordinated regulation and optimization device provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0010] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0011] See Figure 1 The flowchart illustrating the implementation of the integrated energy system coordinated regulation and optimization method provided in this embodiment of the invention is described in detail below: The aforementioned integrated energy system coordinated regulation and optimization methods include: S101: Obtain historical source load data and generate a large-scale scene set based on the historical source load data; First, a wind-solar hybrid integrated energy system combining multiple energy storage components is constructed. The renewable energy equipment unit includes photovoltaic arrays and wind turbines, while the multiple energy storage unit includes lithium-ion batteries, supercapacitors, and thermal storage tanks. The energy conversion unit is a heat pump, used to realize the conversion of electricity into heat / cold energy flow to meet the diversified energy needs of users. The multiple energy storage system absorbs and mitigates the wind and solar power output of the system.

[0012] This application considers a multi-element energy storage system that combines energy storage (lithium batteries), power storage (supercapacitors), and thermal / cold storage. The system relies on renewable energy to meet the user's electricity, heat, and cold load requirements, without relying on external power grids or heating networks, which greatly reduces fossil energy consumption and carbon emissions.

[0013] The system's uncertainty parameters mainly originate from photovoltaic (PV) power and electricity, heat, and gas loads, which exhibit random fluctuations. To quantify these uncertainties, a photovoltaic-load uncertainty model is first established. PV power output is affected by meteorological conditions such as solar radiation, while electricity, heat, and cooling loads depend on user energy consumption behavior. A probability distribution model fitted based on the statistical regularities of a large amount of historical data can characterize these uncertainties.

[0014] (1) Uncertainty model of light radiation Numerous studies have shown that the randomness of light radiation follows a Beta distribution, and its probability density function at time t can be expressed as:

[0015] in, This is the normalized value of the light radiation; It is the Gamma function; and The two shape parameters that determine the Beta distribution are calculated as follows, and the distribution shape for different values ​​is shown in the figure. Figure 2 As shown.

[0016]

[0017]

[0018] in, and denoted as the mean and standard deviation of light radiation at different times t.

[0019] (2) Uncertainty model of electrical, heat and cooling loads Theoretically, when a large amount of historical load data is available, the volatility of multi-energy loads approximately follows a traditional normal distribution. However, in reality, because loads have maximum and minimum values, truncating the tail portion of the traditional normal distribution beyond the upper and lower load boundaries yields a truncated normal distribution describing the volatility of multi-energy loads, as shown in the following equation:

[0020] in, , and Let be the load value of the kl-th energy source at time t, and its maximum and minimum values, respectively. and denoted as the mean and standard deviation of the kl-th energy type under different loads at time t.

[0021] Based on the above system, this application uses the annual historical scenes of solar radiation and electrical, heat and cooling loads at the building site as source load historical data for mining, providing basic scene support for subsequent analysis.

[0022] In one possible implementation, S101 may include: S1011: Normalize the source load historical data to obtain normalized source load historical data; Since the wind speed, solar radiation, and electrical, heat, and gas loads in the historical source load data have different units and large numerical differences, which increases the complexity of subsequent iterative calculations, this application first normalizes the historical source load data to improve the algorithm's convergence speed and eliminate errors between data.

[0023] Specifically, the Min-Max normalization method can be used to unify the dimensions of various data types. The formula for normalizing 365 days of source-load historical data is shown below:

[0024] in, , , This provides the historical data, maximum and minimum values ​​for scenario s (including five scenarios: wind speed, solar radiation, electrical load, heat load and gas load) at time t on day d. This represents the normalized source load historical data at time t on day d.

[0025] S1012: Perform daily scene segmentation on the normalized source-load historical data to form a historical daily scene matrix; Daily scene segmentation was performed on the long-term historical time series of the normalized source-load historical data to obtain the same number of daily scene matrices for each variable of the source-load. The hourly data of source and load variables for the same day are treated as a vector and then merged into a multi-dimensional vector (n*24 dimensions), where n is the number of source and load variables. This yields a historical daily scene matrix with source-load temporal correlation. .

[0026] S1013: Cluster the historical day scene matrix to obtain multiple clusters; We conduct feature analysis on historical daily scenarios with source-load temporal correlations, such as weather changes and the distinct seasonal characteristics of building energy consumption types, for example, summer is dominated by cooling loads and winter by heating loads. Based on human experience and building location, we categorize the annual data, for example, March-May and September-November are transitional seasons, June-August is summer, and December-February is winter.

[0027] Therefore, clustering the historical day scene matrix yields multiple clusters.

[0028] Specifically, the K-means algorithm can be used for clustering.

[0029] The K-means algorithm is an unsupervised learning algorithm primarily used for data clustering. Determining the value of K, i.e., the number of clusters to be formed, is a crucial issue.

[0030] The similarity within a cluster increases with the number of clusters, but too many clusters can reduce the differences between clusters, making it impossible to distinguish between categories. Therefore, it is necessary to set an appropriate number of clusters.

[0031] This application uses the elbow method to determine the optimal number of clusters and utilizes the sum of squared Euclidean distances to measure the clustering effect. As the number of clusters k increases, there is an inflection point in the change of the sum of squared distances; the optimal number of clusters is considered to be reached when the rate of decrease suddenly slows down. The formula for calculating the sum of squared Euclidean distances is as follows:

[0032] in, Indicates a cluster, Indicates the number of cluster centers. For a sample within a certain cluster, It is the center of mass.

[0033] The specific steps of the K-means algorithm are as follows: iteratively calculate k from 1 to 10, and calculate after each clustering. ; The number of clusters gradually decreases, and there will be an inflection point during the process. When the rate of decrease suddenly slows down, it is considered to be the optimal number of clusters.

[0034] S1014: Based on multiple clusters, probability distribution modeling and correlation modeling are performed to obtain the source load probability distribution model and correlation model; To further quantify the temporal patterns and inter-cluster associations of each cluster, probability distribution modeling and correlation modeling need to be carried out for each of the multiple clusters obtained.

[0035] In one possible implementation, S1014 may include: 1. Determine the temporal probability distribution parameters of each cluster based on maximum likelihood estimation to form a source load probability distribution model; The output of photovoltaic power depends on meteorological conditions such as solar radiation, while the load depends on user energy consumption behavior. The source-load probability distribution model, fitted based on the statistical regularities of a large amount of historical data, can characterize its uncertainty. Numerous studies have shown that short-term intraday variations in solar irradiance approximately follow a beta distribution, and variations in multi-energy load can be considered to approximately follow a normal distribution.

[0036] 2. Determine the autocorrelation matrix of each cluster and the cross-correlation matrix between each cluster. The autocorrelation matrix of each cluster and the cross-correlation matrix between each cluster form a correlation model.

[0037] Since the internal data of each cluster are correlated and the external data are significantly different, internal autocorrelation analysis and external cross-correlation analysis are performed on each cluster.

[0038] Specifically, the Spearman correlation coefficient method can be used for correlation analysis. As an effective method for analyzing variables that do not follow a normal distribution, the Spearman correlation coefficient method is suitable for random scenarios where the data follows a Beta distribution, a Weibull distribution, or a truncated normal distribution.

[0039] For a given cluster, the autocorrelation coefficient of its internal scenes The calculation method is shown in the formula. For any scenario on day k and day n in this cluster, the closer the days are, the stronger the correlation. Specifically, when k=n, the correlation coefficient between the two scenarios is 1, therefore the correlation coefficient matrix... The diagonal element is 1, and the specific calculation formula is as follows:

[0040] in, and These are the data at time i on day k and the mean of the data on day k, respectively. and These are the data at time j on day n and the mean of the data on day n, respectively.

[0041] For any two clusters, the cross-correlation coefficient between their scenarios The calculation method is as follows:

[0042] in, and These are the scene vector on day i in the u-th cluster and the scene vector corresponding to its cluster center, respectively. and , respectively, are the scene vectors for day j in the v-th cluster and the scene vectors corresponding to their cluster centers; each of these four vectors contains the amount of data for one day, so the dimension is 24; b is the number of scene days contained in the cluster, and both clusters u and v have M days of scenes.

[0043] The cross-correlation matrix formed As shown in the following formula:

[0044] S1015: Based on the source load probability distribution model and correlation model, a large-scale scene set is obtained by using the Latin hypercube sampling method based on Cholesky decomposition.

[0045] This application leverages the efficiency of Latin hypercube sampling and the ability of Cholesky decomposition to handle correlations to generate large-scale scene sets.

[0046] In one possible implementation, S1015 may include: 1. Construct a joint correlation matrix based on the autocorrelation matrix and the cross-correlation matrix; By combining the autocorrelation matrix and the cross-correlation matrix, a comprehensive joint correlation matrix is ​​formed, which fully characterizes the correlation between source and load variables.

[0047] 2. Perform Cholesky decomposition on the joint correlation matrix to obtain a lower triangular matrix; Cholesky decomposition is typically used to process random number matrices with specific correlations. This application decomposes the joint correlation matrix into a lower triangular matrix, preparing for the subsequent introduction of correlations.

[0048] 3. Based on the source load probability distribution model, the Latin hypercube sampling method is used to perform stratified sampling of the source load variables of each cluster to generate an initial sampling matrix. The Latin hypercube method avoids the uneven data distribution caused by the traditional Monte Carlo method, which concentrates sampling in the center of the sampling interval and has less sampling at the end of the interval, by uniformly dividing the sampling interval.

[0049] Based on the probability distribution models of the source and load respectively, stratified sampling is performed on the variables of each cluster to generate an initial sample matrix, which ensures the uniform distribution of samples in the probability space of each variable and reflects the advantages of Latin hypercube sampling.

[0050] 4. Multiply the initial sampling matrix with the lower triangular matrix to obtain the relevant sample matrix and map it to obtain a large-scale scene set.

[0051] By multiplying the initial sampling matrix with the lower triangular matrix, the originally independent samples are given a pre-defined correlation structure. After mapping and transformation, a large-scale scene set that conforms to the source load probability distribution and correlation characteristics is finally obtained.

[0052] This application retains the efficiency of Latin hypercube sampling while accurately reflecting the complex correlation between source and load, making it suitable for constructing high-quality power system scenario sets.

[0053] S102: The back reduction method is used to reduce the scenes in the large-scale scene set to obtain the source-load joint scene set; The backward reduction method is used to reduce the number of scenes in a large-scale scene set, which significantly reduces the complexity of subsequent optimization calculations while retaining key probabilistic features.

[0054] In one possible implementation, S102 includes: S1021: Calculate the probabilistic distance between any two scenes; Probabilistic distance, also known as Wasserstein distance, is a core metric for measuring the difference between two probability distributions. Its key advantage lies in its ability to effectively quantify the difference even when the support sets (regions with non-zero probabilities) of the two distributions do not overlap, and it is more robust to local changes in the distributions. Therefore, it is widely used in fields such as machine learning, signal processing, and power system scene generation.

[0055] S1022: Based on the probability distance between any two scenes, the backward reduction method is used to reduce the number of scenes, resulting in an initial source-load joint scene set and a discard set; The specific steps for scene reduction using the backward reduction method are as follows: Initialization: Let the discard set J = The initial source-load joint scene set R = {all initial scenes}; Scenario selection: In the k-th iteration, select the scenario that minimizes the following expression. Add to the deprecation set:

[0056] in, For the scene The probability of occurrence; Update set: Deprecated set formula: The formula for retaining sets is: ; Termination condition: Stop when the retained set contains only the preset number of scenes.

[0057] S1023: Merge the probabilities of each scene in the discard set into the scene in the nearest initial source-load joint scene set to obtain the source-load joint scene set.

[0058] For each scene i in the discard set J, transfer its probability quality to the nearest scene j in the retention set:

[0059] in, This is the set of all abandoned scenarios with J as the nearest neighbor.

[0060] The source-load joint scene set generated based on the above steps takes into account temporal sequence, correlation and computational complexity. It can not only reflect the random characteristics of historical scenes with a small number of scenes, but also improve the computational speed of subsequent scheme optimization algorithms.

[0061] S103: Based on the source-load joint scenario set, the optimization solution is obtained by minimizing the daily comprehensive cost as the objective function. In one possible implementation, S103 may include: S1031: Based on the source-load joint scenario set, establish an optimization model with the first formula as the objective function; S1032: The genetic algorithm is used to solve the optimization model to obtain the primary optimization control scheme; The first formula can be:

[0062] in, The total cost per day The daily comprehensive cost of lithium batteries, The daily comprehensive cost of supercapacitors, The total daily cost of the heating (cooling) system, The daily comprehensive cost of the wind turbine, The daily comprehensive cost of a photovoltaic power generation system, Daily operating penalty costs.

[0063] Select the number of fans (N) WT ), number of photovoltaic modules (N) PV ), lithium battery rated capacity (E) bat (N, kWh), rated power of lithium battery (P) bat (N, kW), rated capacity of supercapacitor (E) SC,N (kWh), thermal storage tank capacity (H)tank,N (kWh) and heat pump rated power (P) ep,max The system's daily economic efficiency is evaluated using the daily comprehensive cost (kW) as the optimization variable. The objective function is to minimize the daily comprehensive cost, and a genetic algorithm is used to solve for the primary optimal control scheme.

[0064] For example, the genetic algorithm population size is set to 100, the number of iterations is set to 800, the mutation probability is set to 0.1, and the crossover probability is set to 0.8.

[0065] S104: A fuzzy control strategy is adopted to correct the power of the supercapacitor and the power of the lithium battery based on the primary optimization control scheme, so as to obtain the target optimization control scheme; wherein, the target optimization control scheme includes the equipment capacity configuration scheme.

[0066] After reconstructing the high, medium, and low frequency components of the rated capacity of each device based on VMD, and considering the characteristics of supercapacitors as power storage devices with low energy density and high charging and discharging frequency, in order to avoid overcharging and over-discharging of supercapacitors, a fuzzy control output power correction coefficient is used to correct the power of supercapacitors and lithium batteries based on the real-time state of charge of supercapacitors and the initial power allocation.

[0067] In one possible implementation, S104 may include: S1041: The power correction coefficient is obtained by solving the problem using the state of charge of the supercapacitor as the first input variable of the fuzzy control and the initial power distribution of the supercapacitor as the second input variable of the fuzzy control. The state of charge of the supercapacitor is used as input variable 1 for fuzzy control, with a fuzzy universe of discourse of [0, 1] and a fuzzy set of {VS (very small), S (small), M (medium), B (large), VB (very large)}. The initial power distribution of the supercapacitor is used as input variable 2, with a fuzzy universe of discourse of [-1, 1] and a corresponding fuzzy set of {NB (negative large), NS (negative small), PS (positive small), PB (positive large)}. The fuzzy universe of discourse of the power correction coefficient is [0, 1] and the corresponding fuzzy set of {VS (very small), S (small), M (medium), B (large), VB (very large)}.

[0068] S1042: The power of the supercapacitor and the power of the lithium battery are corrected by using a power correction factor to obtain the corrected power of the supercapacitor and the corrected power of the lithium battery. In one possible implementation, S1042 may include: 1. Using a power correction coefficient and combining it with the second formula, the power of the supercapacitor and the power of the lithium battery are corrected to obtain the corrected power of the supercapacitor and the corrected power of the lithium battery. The second formula may include:

[0069]

[0070] in, This is the corrected supercapacitor power. The initial power distribution to the supercapacitor This is the power correction factor. This is the corrected lithium battery power. To initially allocate power to the lithium battery, This refers to the power returned to the lithium battery via the electrothermal conversion coordination control layer.

[0071] S1043: The primary optimization control scheme is updated by using the corrected supercapacitor power and the corrected lithium battery power to obtain the target optimization control scheme.

[0072] This application employs a two-stage optimization approach, enabling supercapacitors to achieve long-term fluctuation mitigation and reduce lifespan loss. Overcharging and over-discharging phenomena in both supercapacitors and lithium batteries are significantly reduced, leading to lower penalties for power shortages, heat loss, and waste, thereby further reducing overall daily costs. Furthermore, compared to single-stage optimization in existing technologies, this approach is less prone to getting trapped in local optima and improves optimization efficiency.

[0073] The integrated energy system coordinated regulation and optimization provided in this application completes the coordinated regulation and optimization of multiple types of energy, including electricity, heat / cooling, and absorbs renewable energy output under the optimal economic cost to meet diverse load demands.

[0074] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0075] The following are device embodiments of the present invention. For details not described in detail, please refer to the corresponding method embodiments described above.

[0076] Figure 3 A schematic diagram of the integrated energy system coordinated regulation and optimization device provided in an embodiment of the present invention is shown. For ease of explanation, only the parts related to the embodiment of the present invention are shown, and are described in detail below: like Figure 3 As shown, the integrated energy system coordinated regulation and optimization device includes: The initial scene generation module 21 is used to acquire historical source load data and generate a large-scale scene set based on the historical source load data. Scene reduction module 22 is used to reduce the scenes in a large-scale scene set using the backward reduction method to obtain a source-load joint scene set; The initial scheme generation module 23 is used to optimize and solve the initial optimized control scheme based on the source-load joint scenario set with the objective function of minimizing the daily comprehensive cost. The scheme correction module 24 is used to adopt a fuzzy control strategy to correct the power of the supercapacitor and the power of the lithium battery based on the primary optimization control scheme to obtain the target optimization control scheme; wherein, the target optimization control scheme includes the equipment capacity configuration scheme.

[0077] In one possible implementation, the initial scene generation module 21 may include: The normalization unit is used to normalize the source load historical data to obtain the normalized source load historical data. The daily scene segmentation unit is used to segment the normalized source load historical data into daily scenes to form a historical daily scene matrix. Clustering units are used to cluster the historical day scene matrix to obtain multiple cluster groups; The modeling unit is used to perform probability distribution modeling and correlation modeling based on multiple clusters to obtain the source load probability distribution model and correlation model. The scene set generation unit is used to obtain a large-scale scene set by adopting the Latin hypercube sampling method based on Cholesky decomposition, according to the source load probability distribution model and correlation model.

[0078] In one possible implementation, the modeling unit includes: The output sub-unit of the probability distribution model is used to determine the temporal probability distribution parameters of each cluster based on maximum likelihood estimation, thus forming the source load probability distribution model; The correlation model output sub-units are used to determine the autocorrelation matrix of each cluster and the cross-correlation matrix between each cluster. The autocorrelation matrix of each cluster and the cross-correlation matrix between each cluster form the correlation model.

[0079] In one possible implementation, the scene set generation unit may include: The joint matrix construction sub-unit is used to construct the joint correlation matrix based on the autocorrelation matrix and the cross-correlation matrix; The lower triangular matrix determines the sub-units, which are used to perform Cholesky decomposition on the joint correlation matrix to obtain the lower triangular matrix. The initial sampling matrix output sub-unit is used to perform stratified sampling of the source load variables of each cluster according to the source load probability distribution model using the Latin hypercube sampling method to generate the initial sampling matrix. The scene set output sub-unit is used to multiply the initial sampling matrix with the lower triangular matrix to obtain the relevant sample matrix and map it to obtain a large-scale scene set.

[0080] In one possible implementation, the scene reduction module 22 may include: The distance calculation unit is used to calculate the probabilistic distance between any two scenes. The scene reduction unit is used to reduce scenes based on the probability distance between any two scenes using the backward reduction method, to obtain the initial source-load joint scene set and the discard set; The probability merging unit is used to merge the probabilities of each scene in the discard set into the scene in the nearest initial source-load joint scene set to obtain the source-load joint scene set.

[0081] In one possible implementation, the initial scheme generation module 23 may include: The optimization model building unit is used to build an optimization model with the first formula as the objective function based on the source-load joint scenario set. The model solving unit is used to solve the optimization model using a genetic algorithm to obtain a primary optimization control scheme. The first formula is:

[0082] in, The total cost per day The daily comprehensive cost of lithium batteries, The daily comprehensive cost of supercapacitors, The total daily cost of the heating (cooling) system, The daily comprehensive cost of the wind turbine, The daily comprehensive cost of a photovoltaic power generation system, Daily operating penalty costs.

[0083] In one possible implementation, the scheme correction module 24 may include: The modified parameter model solving unit is used to solve the power correction coefficient by taking the state of charge of the supercapacitor as the first input variable of fuzzy control and the initial power distribution of the supercapacitor as the second input variable of fuzzy control. The power correction unit is used to correct the power of the supercapacitor and the power of the lithium battery using a power correction coefficient, so as to obtain the corrected power of the supercapacitor and the corrected power of the lithium battery. The scheme update unit is used to update the primary optimized control scheme by adopting the corrected supercapacitor power and the corrected lithium battery power, so as to obtain the target optimized control scheme.

[0084] In one possible implementation, the power correction unit can be specifically used for: By using a power correction factor and combining it with the second formula, the power of the supercapacitor and the power of the lithium battery are corrected to obtain the corrected power of the supercapacitor and the corrected power of the lithium battery. The second formula may include:

[0085]

[0086] in, This is the corrected supercapacitor power. The initial power distribution to the supercapacitor This is the power correction factor. This is the corrected lithium battery power. To initially allocate power to the lithium battery, This refers to the power returned to the lithium battery via the electrothermal conversion coordination control layer.

[0087] Figure 4 This is a schematic diagram of an electronic device provided in an embodiment of the present invention. Figure 4 As shown, the electronic device 3 of this embodiment includes a processor 30 and a memory 31. The memory 31 stores a computer program 32. When the processor 30 executes the computer program 32, it implements the steps in the various method embodiments described above. Alternatively, when the processor 30 executes the computer program 32, it implements the functions of each module / unit in the various device embodiments described above.

[0088] For example, computer program 32 may be divided into one or more modules / units, which are stored in memory 31 and executed by processor 30 to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of computer program 32 in electronic device 3.

[0089] Electronic device 3 may include, but is not limited to, processor 30 and memory 31. Those skilled in the art will understand that... Figure 4 This is merely an example of electronic device 3 and does not constitute a limitation on electronic device 3. It may include more or fewer components than shown, or combine certain components, or different components. For example, electronic device 3 may also include input / output devices, network access devices, buses, etc.

[0090] For the sake of simplicity and clarity, only the above-described functional modules / units are used as examples. In practical applications, the functions described above can be assigned to different functional modules / units as needed. These modules / units can be implemented in hardware, software, or a combination of both.

[0091] In the above embodiments, the descriptions of each embodiment have their own emphasis. Parts not detailed or described in a particular embodiment can be referred to in the relevant descriptions of other embodiments. Unless otherwise specified or in conflict with logic, the terminology and / or descriptions between different embodiments are consistent and can be referenced interchangeably. Technical features in different embodiments can be combined to form new embodiments based on their inherent logical relationships.

[0092] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for integrated energy system coordinated regulation optimization, characterized in that, The method comprises the following steps: obtaining source-load historical data, and generating a large-scale scenario set based on the source-load historical data; performing reduction on scenarios in the large-scale scenario set by using a backward reduction method to obtain a source-load joint scenario set; performing optimization and solving based on the source-load joint scenario set with a minimum daily comprehensive cost as an objective function to obtain a primary optimization and control scheme; performing correction on super capacitor power and lithium battery power based on the primary optimization and control scheme by using a fuzzy control strategy to obtain a target optimization and control scheme; wherein the target optimization and control scheme comprises a device capacity configuration scheme. 2.The integrated energy system coordinated regulation optimization method of claim 1, wherein, The method of generating the large-scale scenario set based on the source-load historical data comprises the following steps: normalizing the source-load historical data to obtain normalized source-load historical data; performing daily scenario segmentation on the normalized source-load historical data to form a historical daily scenario matrix; clustering the historical daily scenario matrix to obtain a plurality of clustering groups; performing probability distribution modeling and correlation modeling based on the plurality of clustering groups to obtain a source-load probability distribution model and a correlation model; obtaining the large-scale scenario set by using a Latin hypercube sampling method based on Cholesky decomposition according to the source-load probability distribution model and the correlation model. 3.The integrated energy system coordinated regulation optimization method of claim 2, wherein, The method of performing probability distribution modeling and correlation modeling based on the plurality of clustering groups to obtain a source-load probability distribution model and a correlation model comprises the following steps: determining time sequence probability distribution parameters of each clustering group based on maximum likelihood estimation to form the source-load probability distribution model; determining an autocorrelation matrix of each clustering group and a cross-correlation matrix between each clustering group, wherein the autocorrelation matrix of each clustering group and the cross-correlation matrix between each clustering group form the correlation model.

4. The method of claim 3, wherein, The method of obtaining the large-scale scenario set by using a Latin hypercube sampling method based on Cholesky decomposition according to the source-load probability distribution model and the correlation model comprises the following steps: constructing a joint correlation matrix according to the autocorrelation matrix and the cross-correlation matrix; performing Cholesky decomposition on the joint correlation matrix to obtain a lower triangular matrix; performing stratified sampling on source-load variables of each clustering group by using a Latin hypercube sampling method according to the source-load probability distribution model to generate an initial sampling matrix; multiplying the initial sampling matrix and the lower triangular matrix to obtain a correlation sample matrix and map to obtain the large-scale scenario set. 5.The integrated energy system coordinated regulation optimization method according to any one of claims 1 to 4, characterized in that, The method of performing reduction on scenarios in the large-scale scenario set by using a backward reduction method to obtain a source-load joint scenario set comprises the following steps: calculating a probability distance between any two scenarios; performing scenario reduction by using a backward reduction method based on the probability distance between any two scenarios to obtain an initial source-load joint scenario set and a discard set; merging probabilities of scenarios in the discard set to scenarios in the initial source-load joint scenario set closest to the scenarios to obtain the source-load joint scenario set. 6.The integrated energy system coordinated regulation optimization method according to any one of claims 1 to 4, characterized in that, The method of performing optimization and solving based on the source-load joint scenario set with a minimum daily comprehensive cost as an objective function to obtain a primary optimization and control scheme comprises the following steps: establishing an optimization model with a first formula as the objective function based on the source-load joint scenario set; Solving the optimization model by using a genetic algorithm to obtain the preliminary optimal regulation scheme; The first formula is: wherein, is the daily integrated cost for the lithium battery, is the daily integrated cost for the lithium battery, is the daily integrated cost for the supercapacitor, is the daily integrated cost for the heating / cooling system, is the daily integrated cost for the fan, is the daily integrated cost for the photovoltaic system, is the daily operational penalty cost.

7. The method of claim 1 to 4, wherein, The fuzzy control strategy is used to correct the super capacitor power and the lithium battery power based on the preliminary optimal regulation scheme to obtain a target optimal regulation scheme, including: The state of charge of the super capacitor is taken as a first input variable of the fuzzy control, and the initial allocation power of the super capacitor is taken as a second input variable of the fuzzy control to obtain a power correction coefficient; The super capacitor power and the lithium battery power are corrected by using the power correction coefficient to obtain corrected super capacitor power and corrected lithium battery power; The preliminary optimal regulation scheme is updated by using the corrected super capacitor power and the corrected lithium battery power to obtain the target optimal regulation scheme. 8.The integrated energy system coordinated regulation optimization method of claim 7, wherein, The super capacitor power and the lithium battery power are corrected by using the power correction coefficient to obtain corrected super capacitor power and corrected lithium battery power, including: The super capacitor power and the lithium battery power are corrected by using the power correction coefficient in combination with a second formula to obtain corrected super capacitor power and corrected lithium battery power; The second formula includes: wherein, is the modified super capacitor power, is the super capacitor initial allocation power, is the power modification coefficient, is the modified lithium battery power, is the lithium battery initial allocation power, is the power returned to the lithium battery by the electro-thermal conversion coordination control layer.

9. An integrated energy system coordinated regulation optimization device, characterized in that, including: An initial scenario generation module is configured to acquire source load historical data and generate a large-scale scenario set based on the source load historical data; A scenario reduction module is configured to reduce scenarios in the large-scale scenario set by using a backward reduction method to obtain a source load joint scenario set; An initial scheme generation module is configured to perform optimization solving based on the source load joint scenario set and taking minimum daily comprehensive cost as an objective function to obtain a preliminary optimal regulation scheme; A scheme correction module is configured to correct the super capacitor power and the lithium battery power based on the preliminary optimal regulation scheme by using a fuzzy control strategy to obtain a target optimal regulation scheme; the target optimal regulation scheme includes a device capacity configuration scheme.

10. An electronic device, comprising: The integrated energy system collaborative regulation optimization method includes a memory and a processor, the memory stores a computer program, and the processor implements the integrated energy system collaborative regulation optimization method when executing the computer program.

Images