Power distribution area resource collaborative optimization method and system considering source-load matching characteristics

By collecting data, building models, and optimizing the configuration of distributed resources in the distribution area, the mismatch between source, load, and storage controllable resources was solved, achieving efficient power balance and flexible operation, and improving power quality and resource utilization efficiency.

CN118971177APending Publication Date: 2024-11-15GUIZHOU POWER GRID CO LTD
View PDF 0 Cites 13 Cited by

Patent Information

Application Number
CN202410884260.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-03
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

The dispersed access, power randomness, and control nonlinearity of distributed resources exacerbate the power disturbance, imbalance, and harmonic injection in distribution substations, affecting power quality. Existing technologies are insufficient to effectively optimize the utilization of source-load-storage controllable resources in distribution substations.

Method used

By collecting historical data from distribution substations, using clustering algorithms to extract typical daily operating data, constructing a probability distribution model for new energy sources and loads, calculating source-load similarity coefficients, establishing a multi-objective optimization model, and solving it using the NSGA-II algorithm, the optimal compromise solution is selected by combining fuzzy membership functions, thus optimizing the configuration of wind, solar, and energy storage.

Benefits of technology

It has improved the power energy balance and flexible operation of the distribution substation, enhanced resource utilization efficiency, reduced system operating costs, improved power quality and system reliability, and achieved a balance between economy and environmental protection.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118971177A_ABST
    Figure CN118971177A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of collaborative optimization, in particular to a power distribution area resource collaborative optimization method and system considering source-load matching characteristics. Through the steps of data preprocessing, probability distribution modeling, source-load similarity coefficient calculation, multi-objective optimization and the like, efficient collaborative optimization of power distribution area resources is realized. According to the method, the utilization efficiency of distributed energy and stored energy is remarkably improved, the flexibility and reliability of the system are enhanced, and the balance of economy and environmental protection is realized. According to the method, a source load matching evaluation method of Euclidean distance and cosine similarity is innovatively combined, and an optimization solution process of an NSGA-II algorithm and a fuzzy membership function is adopted, so that the calculation efficiency is improved, and a plurality of high-quality selectable schemes are provided for decision makers. The optimization method comprehensively considering the source-load matching characteristics provides a powerful support tool for planning and operation of the intelligent power distribution network.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of collaborative optimization, and in particular to a method and system for collaborative optimization of distribution station area resources taking into account source-load matching characteristics. Background Art

[0002] The distribution substations represented by distributed photovoltaic, energy storage, data centers, electric vehicles, variable frequency air conditioners, etc. have diverse elements. The distribution substations have gradually evolved from a single "load" nature to a "source-load-storage" coexistence nature, which poses a challenge to the efficient use of controllable resources of source, load and storage in the distribution substations. In addition, distributed resources represented by electric vehicles and new energy have significant uncertainty characteristics, which will increase the uncertainty and complexity of the substation operation status. It is necessary to optimize the coordinated control method of the distribution substations to improve the power and energy balance capability and flexible operation level of the distribution substations. Summary of the invention

[0003] In view of the problems existing in the prior art, the present invention is proposed.

[0004] Therefore, the problem to be solved by the present invention is that the characteristics of distributed resource decentralized access, power randomness and control nonlinearity have aggravated the power disturbance, imbalance, harmonic injection, etc. in the distribution station area, affecting the power quality of the distribution station area. This patent takes into account the source-load matching characteristics and proposes a method for collaborative optimization of distribution station area resources to improve the power energy balance capability and flexible operation level of the distribution station area.

[0005] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0006] In a first aspect, an embodiment of the present invention provides a method for collaborative optimization of distribution network area resources taking into account source-load matching characteristics, which includes collecting and preprocessing historical operation data of distribution network areas, and extracting typical daily operation data using a clustering algorithm;

[0007] Construct a probability distribution model of renewable energy output and load, and calculate the source-load similarity coefficient to evaluate the matching characteristics;

[0008] Establish a multi-objective optimization model with the goal of minimizing the source-load matching index and the daily average planning cost, while considering the constraints of distributed generation power, ramping and energy storage capacity;

[0009] The NSGA-II algorithm is used to solve the optimization problem, which is solved iteratively through the steps of population initialization, non-dominated sorting, crowding distance calculation, genetic operation and elite strategy.

[0010] The fuzzy membership function is used to select the optimal compromise solution from the solution set, and the optimal capacity configuration of wind power, solar energy and energy storage and other decision variable values ​​are output.

[0011] As a preferred solution of the method for collaborative optimization of distribution substation resources taking into account the source-load matching characteristics of the present invention, wherein: based on collecting and preprocessing the historical operation data of the distribution substation, a clustering algorithm is used to extract typical daily operation data, specifically the following steps are adopted;

[0012] Based on the historical data of the distribution area, pre-processing steps are implemented for wind power and photovoltaic power generation output and regional load;

[0013] Using k-means clustering technology, the raw data of a whole year are divided into 8 data clusters. These categories represent the typical wind power generation, solar power generation and load levels in 12 months of the year. During the processing, the raw data of wind power and solar power generation are first normalized;

[0014] This study chooses Euclidean distance as an indicator to measure data similarity. By calculating the sum of the squares of the distances J(C) between each original data point and the cluster center of its category, the data point is classified into the cluster center closest to it; if multiple original records have the shortest distance to the same cluster center, these records are classified into the same category;

[0015]

[0016] Where: u mn is a binary variable, u mn =1 means the nth sample belongs to the mth class, u mn =0 means it does not belong to this category; d(c m ,x n ) is the distance from the sample to the cluster center of its category; c m is the sample cluster center; x n For other sample data in the class;

[0017] Based on the above classification results, this paper classifies k cluster centers c m Iterative optimization was performed. Since the calculation of Euclidean distance is based on the square root of the sum of the squares of the differences between each data point and the cluster center, this study introduced the J(C) constraint condition using the Lagrange multiplier method, and selected the square term of the Euclidean distance as the objective function under the criterion of minimizing the square error for optimization. Through the optimization process, the centers of the k categories were iteratively adjusted according to the following formula;

[0018]

[0019] To verify whether the prerequisites for convergence are met, an evaluation is first performed. If the evaluation results show that the convergence criteria are not met, the iterative process based on the aforementioned classification results will be repeated;

[0020] Once the convergence conditions are met, the study will output a set of 24×8 typical daily data, which represents representative conditions in different months of the year.

[0021] As a preferred solution of the distribution station area resource collaborative optimization method for calculating the source-load matching characteristics of the present invention, wherein: the new energy output model and the load probability distribution model are constructed, and the source-load similarity coefficient is calculated to evaluate the matching characteristics. The new energy output model constructed includes the output probability distribution covering wind power and solar power generation. The probability distribution of wind speed and solar radiation intensity is first calculated by a parameterization method. Using these distribution models, the output probability model of wind power and solar power generation is derived through the relationship between wind speed and wind energy conversion efficiency and the relationship between solar radiation intensity and photovoltaic conversion efficiency;

[0022] The power output of new energy is constrained by wind speed, solar radiation, wind turbines, and photovoltaic panels, which makes the model based on the parameterized method have errors. The non-parametric method is used to use historical measured data and the kernel density estimation technology to establish a probability characteristic model. Its expression is as follows:

[0023]

[0024] Where: h is the smoothing coefficient and h>0; K(〃) is the kernel function; x i is a sample of the random variable x; integrating f(x) gives its cumulative distribution function.

[0025] The load probability distribution model is constructed, and the probability characteristics of the load approximately obey the normal distribution:

[0026]

[0027] Where: is the load demand; σ l and u l are the mean and standard deviation respectively. Based on the historical data of load demand, the maximum likelihood estimation method is used to obtain the probability distribution model of load demand.

[0028] 5. As a preferred solution of the method for collaborative optimization of distribution station area resources taking into account source-load matching characteristics of the present invention, wherein: the source-load similarity coefficient is calculated to evaluate the matching characteristics, and the source-load similarity coefficient is based on the similarity of time series;

[0029] When evaluating the similarity of time series, the distance metric is used as the evaluation criterion; the similarity between any two time series is determined by comparing their distances;

[0030] If the distance between two time series is small, they are considered to have a higher similarity; conversely, if the distance is large, the similarity is low;

[0031] At present, according to the Euclidean distance principle, the formula for measuring source-charge similarity can be expressed as:

[0032]

[0033] Where: D E is the Euclidean distance; W(t) is the grid load at time t; P c (t) is the power grid cluster output at time t;

[0034] In the process of coordinated dispatching of power systems, attention should be paid to the absolute distance between the output curves and load curves of different power generation equipment, and the similarity in their shapes should also be considered;

[0035] In order to achieve synchronous adjustment of power generation and power consumption curves, the similarity measurement should not only reflect the direct distance and morphological matching between the curves, but also include cosine similarity in the evaluation of time series similarity. The cosine similarity evaluates the direction difference of two vectors by calculating the cosine value of the angle between them in the vector space. The application method of cosine similarity in the similarity measurement of source-load curves is as follows:

[0036]

[0037] Where: D W is the source-charge similarity coefficient; W max is the maximum value of the grid load at the time scale under investigation or at a longer time scale; α and β are the Euclidean distance weight and cosine similarity weight, respectively, both of which are taken as 0.5 in this paper. The smaller the source-load similarity coefficient, the stronger the matching characteristics between the new energy output and the load.

[0038] As a preferred solution of the method for collaborative optimization of distribution station area resources taking into account the source-load matching characteristics of the present invention, wherein: the optimization is solved by using the NSGA-II algorithm, and the solution is iteratively solved through the steps of population initialization, non-dominated sorting, crowding distance calculation, genetic operation and elite strategy;

[0039] Population initialization: Generate a group of individuals through random methods to form the initial population;

[0040] Fast non-dominated sorting: Use non-dominated relations to sort the initial population to distinguish the superior and inferior levels among individuals;

[0041] Maintaining diversity: Using the concept of crowding distance to protect population diversity, this method avoids parameter adjustment and improves computational efficiency;

[0042] Individual selection: Based on the results of non-dominated sorting and crowding distance, through the league selection mechanism, individuals with higher fitness are screened out to enter the next generation;

[0043] Genetic operation: crossover and mutation operations are performed on the selected individuals to generate a new offspring population;

[0044] Application of elite strategy: merge the newly generated offspring with the original parent generation, and use non-dominated sorting and crowding distance evaluation again to select the next generation of population;

[0045] Iterative execution: Repeatedly follow the established algorithm process until the preset termination condition is met, usually reaching the preset maximum number of iterations;

[0046] Result output: After a series of iterations, the algorithm finally produces a set of solutions that are considered non-dominated in the Pareto sense, forming a Pareto optimal solution set, which represents the optimal compromise between multiple objectives.

[0047] As a preferred solution of the method for collaborative optimization of distribution station area resources taking into account the source-load matching characteristics of the present invention, wherein: the multi-objective optimization model with the source-load matching index and the daily average planning cost minimization as the goal is established, while considering the distributed power source power, ramp climbing and energy storage capacity constraints, which adopts the following steps;

[0048] A multi-objective optimization model is established with the source-load matching index and the minimum user charging cost as the optimization objectives to optimize the wind and solar power output and electric vehicle charging behavior in the region;

[0049] (1) The source-load matching index (F1) is the smallest:

[0050] The source-load matching coefficient and other source-load matching evaluation indicators are used as optimization targets:

[0051] minF1

[0052] (2) The daily average planning cost (F2) is the smallest,

[0053] The daily average planning cost includes investment cost and operating cost. The investment cost includes the installed capacity cost of wind power, photovoltaic power and electrochemical energy storage. The operating cost includes the purchase cost of electricity and the operation, maintenance and depreciation costs of wind, photovoltaic and electrochemical energy storage.

[0054] minF2=f1+f2

[0055]

[0056] Where: F2, f1, f2 are daily average planning cost, daily average investment cost, and daily average operating cost respectively; s is the discount rate, r is the number of years for discount; c w 、c s 、c b are the unit capacity costs of wind power, photovoltaic power, and electrochemical energy storage respectively; Kw , K s , K b are the optimized installed capacities of wind power, photovoltaic power and electrochemical energy storage respectively; Y, D and T are the total number of days in a year, the total number of optimized days and the total number of optimized time periods respectively; P g (d, t) is the power generation of conventional units; P ch (d,t) is the energy storage charging power;

[0057] Power constraints of each distributed power source:

[0058] P i,min ≤P i ≤P i,max

[0059] Where P i is the power output of the i-th distributed generation; P i,min is the minimum power output of the i-th distributed power source; P i,max is the maximum power output of the i-th distributed power source;

[0060] Distributed power supply ramp constraints:

[0061] R i,down Δt≤P i,t -P i,t-1 ≤R i,up Δt

[0062] Where P i,t , P i,t-1 are the output power of the i-th distributed generation at time t and t-1 respectively; R i,down is the lower limit of output active power; R i,up The upper limit of output active power;

[0063] Energy storage capacity constraints:

[0064] SOC min ≤SOC≤SOC max

[0065] SOC start =SOC end

[0066] In the formula, SOC is the energy storage charge state SOC min SOC is the minimum state of charge for energy storage; max SOC is the maximum state of charge of energy storage; start SOC is the initial state of charge of energy storage; end Terminating the state of charge for energy storage;

[0067] The multi-objective optimization model comprises the following steps:

[0068] A group of non-dominated solutions is adopted, and the Pareto solution set represents the non-dominated solutions; the fuzzy membership function is adopted to evaluate the satisfaction of each Pareto solution, and the optimal compromise solution is identified from the Pareto solution set.

[0069] By comparing the satisfaction values ​​of each solution, the optimal compromise solution can be selected; the fuzzy membership function is constructed as follows:

[0070]

[0071] Where: f i is the value of the i-th objective function; f imax and f imin are the upper and lower limits of the objective function value range respectively; m is the number of optimization objectives in the model; M is the satisfaction level corresponding to the i-th objective function value. i When it is 0 or 1, it means that the person is completely dissatisfied or completely satisfied with the value of the i-th objective function; μ is the standardized satisfaction after averaging the satisfaction values ​​of each item. The solution with the largest standardized satisfaction value is the optimal compromise solution.

[0072] As a preferred solution of the method for collaborative optimization of distribution network area resources taking into account the source-load matching characteristics of the present invention, wherein: the optimization is solved by using the NSGA-II algorithm, and the collaborative optimization algorithm for distribution network area resources taking into account the source-load matching characteristics is solved by using the NSGA-II algorithm, and the steps are as follows:

[0073] The historical data of load and renewable energy in the selected distribution area within a specific time period are preprocessed to retain the basic characteristics of the data. For negative values ​​in the data, they are directly set to zero; for over-generation, garbled or missing values, the power values ​​at adjacent time points are used to adjust and supplement them;

[0074] Using the K-means clustering analysis method, the typical daily output sequences of wind energy, solar energy and load are extracted from the historical data to form 8 typical daily data sets of 24×8.

[0075] According to the typical daily output sequence data, the new energy output probability distribution model and load probability distribution model are constructed;

[0076] With the goal of minimizing the source-load matching index and the daily average planning cost, a resource collaborative optimization model for distribution substations that includes wind, solar, and storage and source-load matching characteristics is established by considering the constraints of distributed generation power constraints, distributed generation ramp constraints, and energy storage capacity constraints.

[0077] The obtained clustering data is used as the input of the model, and the boundary parameters of the power grid are set. The NSGA-II algorithm is used to solve the distribution station area resource collaborative optimization model to obtain the optimal capacity configuration of wind power, solar energy and energy storage and the values ​​of other decision variables.

[0078] In a second aspect, an embodiment of the present invention provides a distribution substation resource collaborative optimization system taking into account source-load matching characteristics, which includes a collection module for collecting and preprocessing historical operation data of distribution substations and extracting typical daily operation data using a clustering algorithm;

[0079] Build a module to construct a probability distribution model of renewable energy output and load, and calculate the source-load similarity coefficient to evaluate the matching characteristics;

[0080] Constraint module, which establishes a multi-objective optimization model with the goal of minimizing the source-load matching index and the daily average planning cost, while considering the constraints of distributed power generation power, ramping and energy storage capacity;

[0081] The iterative module uses the NSGA-II algorithm to solve the optimization problem, and iterates through the steps of population initialization, non-dominated sorting, crowding distance calculation, genetic operation and elite strategy;

[0082] The output module uses the fuzzy membership function to select the optimal compromise solution from the solution set and outputs the optimal capacity configuration of wind energy, solar energy and energy storage and other decision variable values.

[0083] In a third aspect, an embodiment of the present invention provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: when the computer program instructions are executed by the processor, the steps of the method for collaborative optimization of distribution station area resources taking into account source-load matching characteristics as described in the first aspect of the present invention are implemented.

[0084] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program instructions are executed by a processor, the steps of the method for collaborative optimization of distribution station area resources taking into account source-load matching characteristics as described in the first aspect of the present invention are implemented.

[0085] The beneficial effects of the present invention are: Data collection and preprocessing of this scheme:

[0086] By collecting and preprocessing the historical operation data of the distribution substation area and using the k-means clustering algorithm to extract typical daily operation data, the simplification and dimensionality reduction of a large amount of complex data are achieved. This step not only improves the efficiency of subsequent analysis, but also retains the representative characteristics of the data, laying the foundation for accurately evaluating the source-load matching characteristics.

[0087] Probability distribution model construction:

[0088] By constructing a probability distribution model of renewable energy output and load, this method achieves a quantitative description of system uncertainty. In particular, the use of kernel density estimation technology to construct a renewable energy output model overcomes the limitations of traditional parameterization methods and improves the accuracy and adaptability of the model.

[0089] Calculation of source-charge similarity coefficient:

[0090] The source-load similarity coefficient is calculated by innovatively combining Euclidean distance and cosine similarity, realizing a comprehensive evaluation of the source-load matching characteristics. This method, which comprehensively considers the curve shape and absolute distance, can more accurately reflect the matching degree between the output of renewable energy and the load, and provides a reliable evaluation index for subsequent optimization.

[0091] Multi-objective optimization model establishment:

[0092] By establishing a multi-objective optimization model with the goal of minimizing the source-load matching index and the daily average planning cost, and taking into account the constraints such as the power of distributed generation, ramping and energy storage capacity, a comprehensive coordinated optimization of distribution substation resources is achieved. This multi-objective optimization method can ensure good performance of source-load matching while improving the economy of the system.

[0093] NSGA-II algorithm solution:

[0094] The NSGA-II algorithm is used to solve multi-objective optimization problems, and efficient iterative solutions are achieved through steps such as population initialization, non-dominated sorting, crowding distance calculation, genetic operations, and elite strategies. This advanced optimization algorithm can not only quickly converge to the Pareto optimal solution set, but also maintain the diversity of solutions, providing decision makers with multiple high-quality solutions.

[0095] Fuzzy membership function selects the optimal solution:

[0096] The fuzzy membership function is used to select the optimal compromise solution from the Pareto solution set, achieving a reasonable trade-off between multiple objectives. This method takes into account the decision maker's preferences and can find the solution that best meets actual needs among many non-inferior solutions.

[0097] Improve the efficiency of resource utilization in distribution station areas:

[0098] By considering the source-load matching characteristics for collaborative optimization, this method significantly improves the utilization efficiency of distributed energy, energy storage and other resources in the distribution area. This not only reduces the system operation cost, but also improves the absorption capacity of new energy, which helps to achieve a low-carbon transformation of the energy structure.

[0099] Enhance system flexibility and reliability:

[0100] By optimizing the configuration of distributed power sources and energy storage, this method enhances the power regulation and energy balancing capabilities of the distribution area. This increased flexibility helps to cope with fluctuations in load and renewable energy output, and improves system reliability and power quality.

[0101] Achieve a balance between economy and environmental protection:

[0102] Through multi-objective optimization, this method reduces planning costs while improving the utilization rate of new energy. This balance not only brings economic benefits, but also helps reduce carbon emissions, achieving a win-win situation for economic and environmental benefits.

[0103] Providing decision support:

[0104] By generating a Pareto optimal solution set and selecting the optimal compromise solution using a fuzzy membership function, this method provides decision makers with multiple options and a scientific basis for decision making, which helps to make more reasonable planning decisions in complex practical situations.

[0105] Highly adaptable:

[0106] This method has strong adaptability and scalability by considering the source-load matching characteristics and multiple constraints. It can adapt to different types of distribution substations and various distributed resources, providing a powerful tool for the planning and operation of future smart distribution networks.

[0107] Improve computational efficiency:

[0108] By adopting k-means clustering and NSGA-II algorithm, this method significantly improves the computational efficiency while ensuring the optimization quality, which enables the method to handle large-scale complex systems and meet the needs of practical engineering applications.

[0109] In summary, this invention has achieved a comprehensive improvement in system efficiency, reliability, economy and environmental protection by innovatively introducing source-load matching characteristics into the coordinated optimization of distribution network resources. Compared with the existing technology, this method has significant advantages in the comprehensiveness of optimization objectives, the efficiency of algorithms and the practicality of results. This unexpected comprehensive performance improvement makes this invention have important application value and promotion prospects in the field of distribution network planning and operation optimization. BRIEF DESCRIPTION OF THE DRAWINGS

[0110] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0111] Figure 1 A flow chart of the coordinated optimization method for distribution area resources taking into account source-load matching characteristics;

[0112] Figure 2 Computer equipment diagram for the collaborative optimization method of distribution station area resources taking into account source-load matching characteristics. DETAILED DESCRIPTION

[0113] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the accompanying drawings.

[0114] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0115] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or selective embodiment that is mutually exclusive with other embodiments.

[0116] Example 1

[0117] Reference Figure 1-2 , which is the first embodiment of the present invention, and provides a method for collaborative optimization of distribution station area resources taking into account source-load matching characteristics, including:

[0118] S100: Collect and pre-process the historical operation data of the distribution station area, and use the clustering algorithm to extract the typical daily operation data;

[0119] S101: Based on the collection and preprocessing of the historical operation data of the distribution station area, the clustering algorithm is used to extract the typical daily operation data, which specifically adopts the following steps;

[0120] Based on the historical data of the distribution area, pre-processing steps are implemented for wind power and photovoltaic power generation output and regional load;

[0121] Using k-means clustering technology, the raw data of a whole year are divided into 8 data clusters. These categories represent the typical wind power generation, solar power generation and load levels in 12 months of the year. During the processing, the raw data of wind power and solar power generation are first normalized;

[0122] This study chooses Euclidean distance as an indicator to measure data similarity. By calculating the sum of the squares of the distances J(C) between each original data point and the cluster center of its category, the data point is classified into the cluster center closest to it; if multiple original records have the shortest distance to the same cluster center, these records are classified into the same category;

[0123]

[0124] Where: u mn is a binary variable, u mn =1 means the nth sample belongs to the mth class, u mn =0 means it does not belong to this category; d(c m ,x n ) is the distance from the sample to the cluster center of its category; c m is the sample cluster center; x n For other sample data in the class;

[0125] Based on the above classification results, this paper classifies k cluster centers c m Iterative optimization was performed. Since the calculation of Euclidean distance is based on the square root of the sum of the squares of the differences between each data point and the cluster center, this study introduced the J(C) constraint condition using the Lagrange multiplier method, and selected the square term of the Euclidean distance as the objective function under the criterion of minimizing the square error for optimization. Through the optimization process, the centers of the k categories were iteratively adjusted according to the following formula;

[0126]

[0127] To verify whether the prerequisites for convergence are met, an evaluation is first performed. If the evaluation results show that the convergence criteria are not met, the iterative process based on the aforementioned classification results will be repeated;

[0128] Once the convergence conditions are met, the study will output a set of 24×8 typical daily data, which represents representative conditions in different months of the year.

[0129] S200: Construct a probability distribution model of renewable energy output and load, and calculate the source-load similarity coefficient to evaluate matching characteristics;

[0130] S201: Based on constructing a new energy output model and a load probability distribution model, and calculating the source-load similarity coefficient to evaluate the matching characteristics, a new energy output model is constructed, including the output probability distribution of wind and solar power generation. Through a parameterized method, the probability distribution of wind speed and solar radiation intensity is first calculated. Using these distribution models, the output probability models of wind and solar power generation are derived through the relationship between wind speed and wind energy conversion efficiency and between solar radiation intensity and photovoltaic conversion efficiency;

[0131] The power output of new energy is constrained by wind speed, solar radiation, wind turbines, and photovoltaic panels, which makes the model based on the parameterized method have errors. The non-parametric method is used to use historical measured data and the kernel density estimation technology to establish a probability characteristic model. Its expression is as follows:

[0132]

[0133] Where: h is the smoothing coefficient and h>0; K(〃) is the kernel function; x i is a sample of the random variable x; integrating f(x) gives its cumulative distribution function.

[0134] Construct a load probability distribution model, and the probability characteristics of the load approximately obey the normal distribution:

[0135]

[0136] Where: is the load demand; σ l and u l are the mean and standard deviation respectively. Based on the historical data of load demand, the maximum likelihood estimation method is used to obtain the probability distribution model of load demand.

[0137] 6. S202: Calculate the source-load similarity coefficient to evaluate the matching characteristics, where the source-load similarity coefficient is based on the similarity of the time series;

[0138] When evaluating the similarity of time series, the distance metric is used as the evaluation criterion; the similarity between any two time series is determined by comparing their distances;

[0139] If the distance between two time series is small, they are considered to have a higher similarity; conversely, if the distance is large, the similarity is low;

[0140] At present, according to the Euclidean distance principle, the formula for measuring source-charge similarity can be expressed as:

[0141]

[0142] Where: D E is the Euclidean distance; W(t) is the grid load at time t; P c (t) is the power grid cluster output at time t;

[0143] In the process of coordinated dispatching of power systems, attention should be paid to the absolute distance between the output curves and load curves of different power generation equipment, and the similarity in their shapes should also be considered;

[0144] In order to achieve synchronous adjustment of power generation and power consumption curves, the similarity measurement should not only reflect the direct distance and morphological matching between the curves, but also include cosine similarity in the evaluation of time series similarity. Cosine similarity evaluates the direction difference of two vectors by calculating the cosine value of the angle between them in the vector space. The application method of cosine similarity in the similarity measurement of source-load curves is as follows:

[0145]

[0146] Where: D W is the source-charge similarity coefficient; W max is the maximum value of the grid load at the time scale under investigation or at a longer time scale; α and β are the Euclidean distance weight and cosine similarity weight, respectively, both of which are taken as 0.5 in this paper. The smaller the source-load similarity coefficient, the stronger the matching characteristics between the new energy output and the load.

[0147] S300: Establish a multi-objective optimization model with the goal of minimizing the source-load matching index and the daily average planning cost, while considering the constraints of distributed generation power, ramping and energy storage capacity;

[0148] S301: Establish a multi-objective optimization model with the goal of minimizing the source-load matching index and the daily average planning cost, while considering the distributed power generation power, ramping and energy storage capacity constraints, which adopts the following steps;

[0149] A multi-objective optimization model is established with the source-load matching index and the minimum user charging cost as the optimization objectives to optimize the wind and solar power output and electric vehicle charging behavior in the region;

[0150] (1) The source-load matching index (F1) is the smallest:

[0151] The source-load matching coefficient and other source-load matching evaluation indicators are used as optimization targets:

[0152] minF1

[0153] (2) The daily average planning cost (F2) is the smallest,

[0154] The daily average planning cost includes investment cost and operating cost. The investment cost includes the installed capacity cost of wind power, photovoltaic power and electrochemical energy storage. The operating cost includes the purchase cost of electricity and the operation, maintenance and depreciation costs of wind, photovoltaic and electrochemical energy storage.

[0155] minF2=f1+f2

[0156]

[0157] Where: F2, f1, f2 are daily average planning cost, daily average investment cost, and daily average operating cost respectively; s is the discount rate, r is the number of years for discount; c w、c s 、c b are the unit capacity costs of wind power, photovoltaic power, and electrochemical energy storage respectively; K w , K s , K b are the optimized installed capacities of wind power, photovoltaic power and electrochemical energy storage respectively; Y, D and T are the total number of days in a year, the total number of optimized days and the total number of optimized time periods respectively; P g (d, t) is the power generation of conventional units; P ch (d,t) is the energy storage charging power;

[0158] Power constraints of each distributed power source:

[0159] P i,min ≤P i ≤P i,max

[0160] Where P i is the power output of the i-th distributed generation; P i,min is the minimum power output of the i-th distributed power source; P i,max is the maximum power output of the i-th distributed power source;

[0161] Distributed power supply ramp constraints:

[0162] R i,down Δt≤P i,t -P i,t-1 ≤R i,up Δt

[0163] Where P i,t , P i,t-1 are the output power of the i-th distributed generation at time t and t-1 respectively; R i,down is the lower limit of output active power; R i,up The upper limit of output active power;

[0164] Energy storage capacity constraints:

[0165] SOC min ≤SOC≤SOC max

[0166] SOC start =SOC end

[0167] In the formula, SOC is the energy storage charge state SOC min SOC is the minimum state of charge for energy storage; max SOC is the maximum state of charge of energy storage; start SOC is the initial state of charge of energy storage; end Terminating the state of charge for energy storage;

[0168] The multi-objective optimization model includes the following steps:

[0169] A set of non-dominated solutions is used, and the Pareto solution set represents the non-dominated solutions. The fuzzy membership function is used to evaluate the satisfaction of each Pareto solution, and the optimal compromise solution is identified from the Pareto solution set.

[0170] By comparing the satisfaction values ​​of each solution, the optimal compromise solution can be selected; the fuzzy membership function is constructed as follows:

[0171]

[0172] Where: f i is the value of the i-th objective function; f imax and f imin are the upper and lower limits of the objective function value range respectively; m is the number of optimization objectives in the model; M is the satisfaction level corresponding to the i-th objective function value. i When it is 0 or 1, it means that the person is completely dissatisfied or completely satisfied with the value of the i-th objective function; μ is the standardized satisfaction after averaging the satisfaction values ​​of each item. The solution with the largest standardized satisfaction value is the optimal compromise solution.

[0173] S400: The optimization is solved using the NSGA-II algorithm, which is solved iteratively through the steps of population initialization, non-dominated sorting, crowding distance calculation, genetic operation and elite strategy;

[0174] S401: Use NSGA-II algorithm to solve the optimization, and iterate through the steps of population initialization, non-dominated sorting, crowding distance calculation, genetic operation and elite strategy;

[0175] Population initialization: Generate a group of individuals through random methods to form the initial population;

[0176] Fast non-dominated sorting: Use non-dominated relations to sort the initial population to distinguish the superior and inferior levels among individuals;

[0177] Maintaining diversity: Using the concept of crowding distance to protect population diversity, this method avoids parameter adjustment and improves computational efficiency;

[0178] Individual selection: Based on the results of non-dominated sorting and crowding distance, through the league selection mechanism, individuals with higher fitness are screened out to enter the next generation;

[0179] Genetic operation: crossover and mutation operations are performed on the selected individuals to generate a new offspring population;

[0180] Application of elite strategy: merge the newly generated offspring with the original parent generation, and use non-dominated sorting and crowding distance evaluation again to select the next generation of population;

[0181] Iterative execution: Repeatedly follow the established algorithm process until the preset termination condition is met, usually reaching the preset maximum number of iterations;

[0182] Result output: After a series of iterations, the algorithm finally produces a set of solutions that are considered non-dominated in the Pareto sense, forming a Pareto optimal solution set, which represents the optimal compromise between multiple objectives.

[0183] S500: The fuzzy membership function is used to select the optimal compromise solution from the solution set, and the optimal capacity configuration of wind energy, solar energy and energy storage and other decision variable values ​​are output.

[0184] S501: Using the NSGA-II algorithm to solve the optimization, the NSGA-II algorithm is used to solve the distribution area resource collaborative optimization algorithm taking into account the source-load matching characteristics, and the steps are as follows:

[0185] The historical data of load and renewable energy in the selected distribution area within a specific time period are preprocessed to retain the basic characteristics of the data. For negative values ​​in the data, they are directly set to zero; for over-generation, garbled or missing values, the power values ​​at adjacent time points are used to adjust and supplement them;

[0186] Using the K-means clustering analysis method, the typical daily output sequences of wind energy, solar energy and load are extracted from the historical data to form 8 typical daily data sets of 24×8.

[0187] According to the typical daily output sequence data, the new energy output probability distribution model and load probability distribution model are constructed;

[0188] With the goal of minimizing the source-load matching index and the daily average planning cost, a resource collaborative optimization model for distribution substations that includes wind, solar, and storage and source-load matching characteristics is established by considering the constraints of distributed generation power constraints, distributed generation ramp constraints, and energy storage capacity constraints.

[0189] The obtained clustering data is used as the input of the model, and the boundary parameters of the power grid are set. The NSGA-II algorithm is used to solve the distribution station area resource collaborative optimization model to obtain the optimal capacity configuration of wind power, solar energy and energy storage and the values ​​of other decision variables.

[0190] Furthermore, this embodiment also provides a distribution network area resource collaborative optimization system taking into account source-load matching characteristics, including:

[0191] The collection module collects and pre-processes the historical operation data of the distribution station area and uses the clustering algorithm to extract the typical daily operation data;

[0192] Build a module to construct a probability distribution model of renewable energy output and load, and calculate the source-load similarity coefficient to evaluate the matching characteristics;

[0193] Constraint module, which establishes a multi-objective optimization model with the goal of minimizing the source-load matching index and the daily average planning cost, while considering the constraints of distributed power generation power, ramping and energy storage capacity;

[0194] The iterative module uses the NSGA-II algorithm to solve the optimization problem, and iterates through the steps of population initialization, non-dominated sorting, crowding distance calculation, genetic operation and elite strategy;

[0195] The output module uses the fuzzy membership function to select the optimal compromise solution from the solution set and outputs the optimal capacity configuration of wind energy, solar energy and energy storage and other decision variable values.

[0196] This embodiment also provides a computer device, which is suitable for the case of a collaborative optimization method for distribution station area resources taking into account source-load matching characteristics, including a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement the collaborative optimization method for distribution station area resources taking into account source-load matching characteristics as proposed in the above embodiment.

[0197] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0198] This embodiment further provides a storage medium on which a computer program is stored. When the program is executed by a processor, the method for collaboratively optimizing distribution station area resources taking into account source-load matching characteristics as proposed in the above embodiment is implemented.

[0199] In summary, data collection and preprocessing:

[0200] By collecting and preprocessing the historical operation data of the distribution substation area and using the k-means clustering algorithm to extract typical daily operation data, the simplification and dimensionality reduction of a large amount of complex data are achieved. This step not only improves the efficiency of subsequent analysis, but also retains the representative characteristics of the data, laying the foundation for accurately evaluating the source-load matching characteristics.

[0201] Probability distribution model construction:

[0202] By constructing a probability distribution model of renewable energy output and load, this method achieves a quantitative description of system uncertainty. In particular, the use of kernel density estimation technology to construct a renewable energy output model overcomes the limitations of traditional parameterization methods and improves the accuracy and adaptability of the model.

[0203] Calculation of source-charge similarity coefficient:

[0204] The source-load similarity coefficient is calculated by innovatively combining Euclidean distance and cosine similarity, realizing a comprehensive evaluation of the source-load matching characteristics. This method, which comprehensively considers the curve shape and absolute distance, can more accurately reflect the matching degree between the output of renewable energy and the load, and provides a reliable evaluation index for subsequent optimization.

[0205] Multi-objective optimization model establishment:

[0206] By establishing a multi-objective optimization model with the goal of minimizing the source-load matching index and the daily average planning cost, and taking into account the constraints such as the power of distributed generation, ramping and energy storage capacity, a comprehensive coordinated optimization of distribution substation resources is achieved. This multi-objective optimization method can ensure good performance of source-load matching while improving the economy of the system.

[0207] NSGA-II algorithm solution:

[0208] The NSGA-II algorithm is used to solve multi-objective optimization problems, and efficient iterative solutions are achieved through steps such as population initialization, non-dominated sorting, crowding distance calculation, genetic operations, and elite strategies. This advanced optimization algorithm can not only quickly converge to the Pareto optimal solution set, but also maintain the diversity of solutions, providing decision makers with multiple high-quality solutions.

[0209] Fuzzy membership function selects the optimal solution:

[0210] The fuzzy membership function is used to select the optimal compromise solution from the Pareto solution set, achieving a reasonable trade-off between multiple objectives. This method takes into account the decision maker's preferences and can find the solution that best meets actual needs among many non-inferior solutions.

[0211] Improve the efficiency of resource utilization in distribution station areas:

[0212] By considering the source-load matching characteristics for collaborative optimization, this method significantly improves the utilization efficiency of distributed energy, energy storage and other resources in the distribution area. This not only reduces the system operation cost, but also improves the absorption capacity of new energy, which helps to achieve a low-carbon transformation of the energy structure.

[0213] Enhance system flexibility and reliability:

[0214] By optimizing the configuration of distributed power sources and energy storage, this method enhances the power regulation and energy balancing capabilities of the distribution area. This increased flexibility helps to cope with fluctuations in load and renewable energy output, and improves system reliability and power quality.

[0215] Achieve a balance between economy and environmental protection:

[0216] Through multi-objective optimization, this method reduces planning costs while improving the utilization rate of new energy. This balance not only brings economic benefits, but also helps reduce carbon emissions, achieving a win-win situation for economic and environmental benefits.

[0217] Providing decision support:

[0218] By generating a Pareto optimal solution set and selecting the optimal compromise solution using a fuzzy membership function, this method provides decision makers with multiple options and a scientific basis for decision making, which helps to make more reasonable planning decisions in complex practical situations.

[0219] Highly adaptable:

[0220] This method has strong adaptability and scalability by considering the source-load matching characteristics and multiple constraints. It can adapt to different types of distribution substations and various distributed resources, providing a powerful tool for the planning and operation of future smart distribution networks.

[0221] Improve computational efficiency:

[0222] By adopting k-means clustering and NSGA-II algorithm, this method significantly improves the computational efficiency while ensuring the optimization quality, which enables the method to handle large-scale complex systems and meet the needs of practical engineering applications.

[0223] In summary, this invention has achieved a comprehensive improvement in system efficiency, reliability, economy and environmental protection by innovatively introducing source-load matching characteristics into the coordinated optimization of distribution network resources. Compared with the existing technology, this method has significant advantages in the comprehensiveness of optimization objectives, the efficiency of algorithms and the practicality of results. This unexpected comprehensive performance improvement makes this invention have important application value and promotion prospects in the field of distribution network planning and operation optimization.

[0224] Example 2

[0225] Reference Figure 1 - Figure 2 , which is the second embodiment of the present invention, and this embodiment provides a method for collaborative optimization of distribution station area resources taking into account source-load matching characteristics. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0226] In order to verify the effectiveness of the resource collaborative optimization method of the distribution area taking into account the source-load matching characteristics proposed in this invention, a typical distribution area in a certain region was selected as the research object for experimental analysis. The distribution area covers an area of ​​about 10 square kilometers, including 1,000 residential users, 50 commercial users and 5 small industrial users, with an annual electricity consumption of about 15 million kWh. There is a photovoltaic power generation system with an installed capacity of 1MW and a wind turbine generator set with an installed capacity of 0.5MW in the area, but it faces problems such as the mismatch between the output of new energy and the load and the high system operation cost.

[0227] First, the historical operation data of the distribution substation in the past three years were collected, including load data at 15-minute intervals, new energy output data, and related meteorological data. The original data was preprocessed to remove outliers and complete missing data. Then the improved K-means clustering algorithm was used to divide the data of one year into 8 typical days. Compared with the traditional K-means algorithm, this experiment adopted a density-based initial clustering center selection method and introduced an adaptive learning rate to improve the clustering effect and convergence speed.

[0228] Next, the new energy output and load probability distribution model is constructed based on typical daily data. This experiment uses the kernel density estimation method to construct the wind power and photovoltaic output probability model, overcoming the limitations of the parameterized method. For the load probability distribution, considering its obvious time series characteristics, the conditional kernel density estimation method is used to introduce the time factor as a conditional variable into the model, which improves the accuracy of the load probability model.

[0229] In terms of source-load similarity coefficient calculation, this experiment innovatively proposed a comprehensive evaluation method based on dynamic time warping (DTW) distance and Pearson correlation coefficient. DTW distance can better capture the morphological similarity of time series, while Pearson correlation coefficient reflects the consistency of sequence change trend. The combination of the two takes into account both the absolute matching degree of source-load curves and the coordination of change trends.

[0230] The optimization model was established using a multi-objective weighted adaptive method. The weights of the source-load matching index and the daily average planning cost were dynamically adjusted according to the decision maker's preferences and the real-time operating status of the system. At the same time, considering the intermittent characteristics of distributed power sources, the power limit of distributed power sources based on probability constraints was introduced to improve the practicality of the model.

[0231] In terms of the solution algorithm, this experiment improved the NSGA-II algorithm. The adaptive crossover and mutation operations based on crowding were introduced to enhance the local search ability of the algorithm. At the same time, the ε-constraint method was used to handle the constraints to improve the feasibility of the solution. When selecting the optimal compromise solution, the TOPSIS method and the fuzzy membership function were innovatively combined to better balance multiple objectives.

[0232] The experiment uses the traditional method (without considering the source-load matching characteristics) and the method of the present invention to optimize the configuration and compare it with the original configuration. The optimization objectives include the source-load matching index, the daily average planning cost, the new energy consumption rate and the system reliability index SAIDI (system average power outage time index). The optimization results are shown in the following table:

[0233]

[0234] Through comparative analysis, it can be seen that the method of the present invention is significantly superior to the traditional method and the original configuration in many aspects:

[0235] Source-load matching index: The method of the present invention reduces the source-load matching index from the original 0.68 to 0.42, which is 28.8% lower than the 0.59 of the traditional method. This shows that the method can better coordinate the matching relationship between the output of new energy and the load demand, and effectively reduce the power fluctuation and energy imbalance problems.

[0236] Daily average planning cost: The daily average planning cost of the method of the present invention is 18,200 yuan, which is 15.3% lower than the original configuration and 6.7% lower than the traditional method. This shows that the method not only improves system performance, but also optimizes economic efficiency, bringing considerable economic benefits to the distribution company.

[0237] New energy consumption rate: The method of the present invention increases the new energy consumption rate to 93.7%, which is 15.2 percentage points higher than the original configuration and 8.4 percentage points higher than the traditional method. This significant improvement reflects the advantages of the method in promoting the efficient use of new energy and helps to achieve a low-carbon transformation of the energy structure.

[0238] System reliability: The method of the present invention reduces SAIDI to 2.1 hours / year, which is 34.4% lower than the original configuration and 25% lower than the traditional method. This shows that the method significantly improves the reliability of the distribution system and reduces the power outage time of users through optimized configuration and coordinated control.

[0239] Optimization of resource allocation: The method of the present invention has increased the configuration of wind power, photovoltaic and energy storage capacity, but the increase is more reasonable. In particular, the significant increase in energy storage capacity (1.5MWh) reflects that the method fully considers the source-load matching characteristics, and uses the flexible adjustment ability of the energy storage system to smooth the fluctuation of new energy output and improve the overall performance of the system.

[0240] Synergistic effect: By comparison, it can be found that the improvement of the method of the present invention in various indicators far exceeds the improvement that can be brought about by simply increasing the equipment capacity. This shows that the method realizes the coordinated optimization of various resources by considering the source-load matching characteristics, and produces a system effect of "1+1>2".

[0241] Adaptability and robustness: Although not directly reflected in the table, it can be inferred from the significant improvement in the source-load matching index that the system optimized by this method has a stronger adaptability to the fluctuations in renewable energy output and load demand, and the robustness of the system operation has been improved.

[0242] Environmental benefits: The substantial increase in the consumption rate of new energy means that more clean energy is used, thereby reducing the use of conventional energy. Based on the calculation of 0.5 kg of carbon dioxide emissions per kilowatt-hour, this method reduces carbon dioxide emissions by about 630 tons per year compared with traditional methods, with significant environmental benefits.

[0243] In summary, the method of the present invention has achieved significant improvements in economy, reliability, environmental protection, and system performance by innovatively introducing source-load matching characteristics into the coordinated optimization of distribution station resources. This comprehensive optimization effect not only solves the problems of mismatch between new energy and load and low system efficiency in the prior art, but also provides new ideas and tools for the planning and operation of future smart distribution networks. The implementation of this method is expected to promote the development of distribution systems in a cleaner, more efficient, reliable, and economical direction, and has broad application prospects.

[0244] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for collaborative optimization of distribution area resources taking into account source-load matching characteristics, characterized in that: include, Collect and pre-process the historical operation data of the distribution station area, and use clustering algorithms to extract typical daily operation data; Construct a probability distribution model of renewable energy output and load, and calculate the source-load similarity coefficient to evaluate matching characteristics; Establish a multi-objective optimization model with the goal of minimizing the source-load matching index and the daily average planning cost, while considering the constraints of distributed generation power, ramping and energy storage capacity; The NSGA-II algorithm is used to solve the optimization problem, which is solved iteratively through the steps of population initialization, non-dominated sorting, crowding distance calculation, genetic operation and elite strategy. The fuzzy membership function is used to select the optimal compromise solution from the solution set, and the optimal capacity configuration of wind power, solar energy and energy storage and other decision variable values ​​are output.

2. The method for collaborative optimization of distribution area resources taking into account source-load matching characteristics as claimed in claim 1, characterized in that: The method is based on collecting and preprocessing the historical operation data of the distribution station area, and using a clustering algorithm to extract typical daily operation data, specifically using the following steps; Based on the historical data of the distribution area, pre-processing steps are implemented for wind power and photovoltaic power generation output and regional load; Using k-means clustering technology, the raw data of a whole year are divided into 8 data clusters. These categories represent the typical wind power generation, solar power generation and load levels in 12 months of the year. During the processing, the raw data of wind power and solar power generation are first normalized; This study selects Euclidean distance as an indicator to measure data similarity. By calculating the sum of the squares of the distances J(C) between each original data point and the cluster center of its category, the data point is classified into the cluster center that is closest to it. If multiple original records have the shortest distance to the same cluster center, these records are classified into the same category. Where: u mn is a binary variable, u mn =1 means the nth sample belongs to the mth class, u mn =0 means it does not belong to this category; d(c m ,x n ) is the distance from the sample to the cluster center of its category; c m is the sample cluster center; x n For other sample data in the class; Based on the above classification results, this paper classifies k cluster centers c m Iterative optimization was performed. Since the calculation of Euclidean distance is based on the square root of the sum of the squares of the differences between each data point and the cluster center, this study introduced the J(C) constraint condition using the Lagrange multiplier method and selected the square term of the Euclidean distance as the objective function under the criterion of minimizing the square error for optimization. Through the optimization process, the centers of the k categories were iteratively adjusted according to the following formula; In order to verify whether the prerequisites for convergence are met, an evaluation is first performed. If the evaluation results show that the convergence criteria have not been met, the iterative process based on the aforementioned classification results will be repeated; Once the convergence conditions are met, the study will output a set of 24×8 typical daily data, which represents representative conditions in different months of the year.

3. The method for collaborative optimization of distribution area resources taking into account source-load matching characteristics as claimed in claim 2, characterized in that: The method is based on constructing a new energy output model and a load probability distribution model, and calculating the source-load similarity coefficient to evaluate the matching characteristics. The new energy output model includes the output probability distribution covering wind power and solar power generation. Through a parameterized method, the probability distribution of wind speed and solar radiation intensity is first calculated. Using these distribution models, the output probability models of wind power and solar power generation are derived through the relationship between wind speed and wind energy conversion efficiency and between solar radiation intensity and photovoltaic conversion efficiency. The power output of new energy is constrained by wind speed, solar radiation, wind turbines, and photovoltaic panels, which makes the model based on the parameterized method have errors. The non-parametric method is used to use historical measured data and the kernel density estimation technology to establish a probability characteristic model. Its expression is as follows: Where: h is the smoothing coefficient and h>0; K(〃) is the kernel function; x i is a sample of the random variable x; integrating f(x) yields its cumulative distribution function, The load probability distribution model is constructed, and the probability characteristics of the load approximately obey the normal distribution: Where: xlx l is the load demand; σ l and u l are the mean and standard deviation respectively. Based on the historical data of load demand, the maximum likelihood estimation method is used to obtain the load demand probability distribution model.

4. The method for collaborative optimization of distribution area resources taking into account source-load matching characteristics as claimed in claim 3, characterized in that: and calculating a source-load similarity coefficient to evaluate matching characteristics, wherein the source-load similarity coefficient is based on the similarity of the time series; When evaluating the similarity of time series, the distance metric is used as the evaluation criterion; the similarity between any two time series is determined by comparing their distances; If the distance between two time series is small, they are considered to have a higher similarity; conversely, if the distance is large, the similarity is low; At present, according to the Euclidean distance principle, the formula for measuring source-charge similarity can be expressed as: Where: D E is the Euclidean distance; W(t) is the grid load at time t; P c (t) is the power grid cluster output at time t; In the process of coordinated dispatching of power systems, attention should be paid to the absolute distance between the output curves and load curves of different power generation equipment, and the similarity in their shapes should also be considered; In order to achieve synchronous adjustment of power generation and power consumption curves, the similarity measurement should not only reflect the direct distance and morphological matching between the curves, but also include cosine similarity in the evaluation of time series similarity. The cosine similarity evaluates the direction difference of two vectors by calculating the cosine value of the angle between them in the vector space. The application method of cosine similarity in the similarity measurement of source-load curves is as follows: Where: D W is the source-charge similarity coefficient;W max is the maximum value of the grid load under the investigation time scale or longer time scale; α and β are the Euclidean distance weight and cosine similarity weight, respectively, both of which are taken as 0.5 in this paper. The smaller the source-load similarity coefficient, the stronger the matching characteristics between the new energy output and the load.

5. The method for collaborative optimization of distribution area resources taking into account source-load matching characteristics as claimed in claim 4, characterized in that: The optimization is solved by using the NSGA-II algorithm, which is solved iteratively through the steps of population initialization, non-dominated sorting, crowding distance calculation, genetic operation and elite strategy; Population initialization: Generate a group of individuals through random methods to form the initial population; Fast non-dominated sorting: Use non-dominated relations to sort the initial population to distinguish the superior and inferior levels among individuals; Maintaining diversity: Using the concept of crowding distance to protect population diversity, this method avoids parameter adjustment and improves computational efficiency; Individual selection: Based on the results of non-dominated sorting and crowding distance, through the league selection mechanism, individuals with higher fitness are screened out to enter the next generation; Genetic operation: crossover and mutation operations are performed on the selected individuals to generate a new offspring population; Application of elite strategy: merge the newly generated offspring with the original parent generation, and use non-dominated sorting and crowding distance evaluation again to select the next generation of population; Iterative execution: Repeatedly follow the established algorithm flow until the preset termination condition is met, usually reaching the preset maximum number of iterations; Result output: After a series of iterations, the algorithm finally produces a set of solutions that are considered non-dominated in the Pareto sense, forming a Pareto optimal solution set, which represents the optimal compromise between multiple objectives.

6. The method for collaborative optimization of distribution area resources taking into account source-load matching characteristics as claimed in claim 5, characterized in that: The multi-objective optimization model is established with the goal of minimizing the source-load matching index and the daily average planning cost, while considering the distributed power supply power, ramp climbing and energy storage capacity constraints, which adopts the following steps; A multi-objective optimization model is established with the source-load matching index and the minimum user charging cost as the optimization objectives to optimize the wind and solar power output and electric vehicle charging behavior in the region; (1) The source-charge matching index (F1) is the smallest: The source-load matching coefficient and other source-load matching evaluation indicators are used as optimization targets: min F1 (2) The daily average planning cost (F2) is the smallest, The daily average planning cost includes investment cost and operating cost. The investment cost includes the installed capacity cost of wind power, photovoltaic power and electrochemical energy storage. The operating cost includes the purchase cost of electricity and the operation, maintenance and depreciation costs of wind, photovoltaic and electrochemical energy storage. min F2=f1+f2 Where: F2, f1, f2 are daily average planning cost, daily average investment cost, and daily average operating cost respectively; s is the discount rate, r is the number of years for discount; c w 、c s 、c b are the unit capacity costs of wind power, photovoltaic power, and electrochemical energy storage respectively; K w , K s , K b are the optimized installed capacities of wind power, photovoltaic power and electrochemical energy storage respectively; Y, D and T are the total number of days in a year, the total number of optimized days and the total number of optimized time periods respectively; P g (d, t) is the power generation of conventional units; P ch (d,t) is the energy storage charging power; Power constraints of each distributed power source: P i,min ≤P i ≤P i,max Where P i is the power output of the i-th distributed generation; P i,min is the minimum power output of the i-th distributed power source; P i,max is the maximum power output of the i-th distributed power source; Distributed power supply ramp constraints: R i,down Δt≤P i,t -P i,t-1 ≤R i,up Δt Where P i,t , P i,t-1 are the output power of the i-th distributed generation at time t and t-1 respectively; R i,down is the lower limit of output active power; R i,up The upper limit of output active power; Energy storage capacity constraints: SOC min ≤SOC≤SOC max SOCIETY start =SOC end In the formula, SOC is the energy storage charge state SOC min SOC is the minimum state of charge for energy storage; max SOC is the maximum state of charge of energy storage; start SOC is the initial state of charge of energy storage; end Terminating the state of charge for energy storage; The multi-objective optimization model comprises the following steps: A group of non-dominated solutions is adopted, and the Pareto solution set represents the non-dominated solutions; the fuzzy membership function is adopted to evaluate the satisfaction of each Pareto solution, and the optimal compromise solution is identified from the Pareto solution set. By comparing the satisfaction values ​​of each solution, the optimal compromise solution can be selected; the fuzzy membership function is constructed as follows: Where: f i is the value of the i-th objective function; f imax and f imin are the upper and lower limits of the objective function value range respectively; m is the number of optimization objectives in the model; M is the satisfaction level corresponding to the i-th objective function value. i When it is 0 or 1, it means that the person is completely dissatisfied or completely satisfied with the value of the i-th objective function; μ is the standardized satisfaction after averaging the satisfaction values ​​of each item. The solution with the largest standardized satisfaction value is the optimal compromise solution.

7. The method for collaborative optimization of distribution area resources taking into account source-load matching characteristics as claimed in claim 6, characterized in that: The optimization is solved by using the NSGA-II algorithm. The steps of solving the distribution station area resource collaborative optimization algorithm taking into account the source-load matching characteristics by using the NSGA-II algorithm are as follows: The historical data of load and renewable energy in the selected distribution area within a specific time period are preprocessed to retain the basic characteristics of the data. For negative values ​​in the data, they are directly set to zero; for over-generation, garbled or missing values, the power values ​​at adjacent time points are used to adjust and supplement them; Using the K-means clustering analysis method, the typical daily output sequences of wind power, solar power and load are extracted from the historical data to form 8 typical daily data sets of 24×8. According to the typical daily output sequence data, the new energy output probability distribution model and load probability distribution model are constructed; With the goal of minimizing the source-load matching index and the daily average planning cost, a resource collaborative optimization model for distribution substations that includes wind, solar, and storage and source-load matching characteristics is established by considering the constraints of distributed generation power constraints, distributed generation ramp constraints, and energy storage capacity constraints. The obtained clustering data is used as the input of the model, and the boundary parameters of the power grid are set. The NSGA-II algorithm is used to solve the distribution station area resource collaborative optimization model to obtain the optimal capacity configuration of wind power, solar energy and energy storage and the values ​​of other decision variables.

8. A system for collaborative optimization of distribution network area resources taking into account source-load matching characteristics, based on the method for collaborative optimization of distribution network area resources taking into account source-load matching characteristics as claimed in any one of claims 1 to 7, characterized in that: Also includes, The collection module collects and pre-processes the historical operation data of the distribution station area and uses the clustering algorithm to extract the typical daily operation data; Build a module to construct a probability distribution model of renewable energy output and load, and calculate the source-load similarity coefficient to evaluate the matching characteristics; Constraint module, which establishes a multi-objective optimization model with the goal of minimizing the source-load matching index and the daily average planning cost, while considering the constraints of distributed power generation power, ramping and energy storage capacity; The iterative module uses the NSGA-II algorithm to solve the optimization problem, and iterates through the steps of population initialization, non-dominated sorting, crowding distance calculation, genetic operation and elite strategy; The output module uses the fuzzy membership function to select the optimal compromise solution from the solution set and outputs the optimal capacity configuration of wind energy, solar energy and energy storage and other decision variable values.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for collaborative optimization of distribution station area resources taking into account source-load matching characteristics as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for collaborative optimization of distribution station area resources taking into account source-load matching characteristics as described in any one of claims 1 to 7 are implemented.

Citation Information

Cited By

  • Chain park energy consumption load management optimization method and system based on clustering algorithm

    CN119226833A

  • Distribution transformer district adjustable and controllable resource capacity evaluation method and system based on multilayer clustering strategy

    CN120262434A

  • A method and system for evaluating adjustable resource capacity of a distribution transformer area based on a multi-layer clustering strategy

    CN120262434B

  • Method and system for optimizing distributed resources of multiple transformer areas of power distribution network

    CN120262575A

  • Carbon electricity energy storage power control strategy analysis method and system considering multi-source-load cooperation

    CN120433305A