A power grid dispatching method and system based on aggregation units

Abstract

The application discloses a power grid scheduling method and system based on aggregation units, and belongs to the field of power system scheduling. The method is as follows: by acquiring the dynamic characteristic vectors of various distributed resources in a power grid, a clustering algorithm is used to aggregate the dynamic characteristic vectors into aggregation matching units, a fast greedy multi-start selection algorithm is used to generate and screen candidate combinations, a final candidate combination is obtained, a scheduling optimization model is established according to the final candidate combination, and the optimal power instructions of the aggregation matching units are obtained by solving, and then the accurate scheduling of the power grid is realized. Therefore, by implementing the application, the problems of the prior art, such as static solidification of the aggregation model, difficulty in real-time adaptation to resource dynamic characteristics, high scheduling cost and low efficiency, can be solved.

Description

Technical Field

[0001] This invention relates to the field of power systems, and in particular to a power grid dispatching method and system based on aggregation units. Background Technology

[0002] In modern power systems, the rapid integration of distributed photovoltaic, wind power, energy storage, and flexible loads has resulted in a highly randomized, decentralized, and diversified system on both the source and load sides.

[0003] In existing technologies, aggregation methods are commonly used to combine geographically dispersed distributed resources with varying capacities into one or more larger, more manageable virtual entities, such as virtual power plants or virtual energy storage systems. These methods aim to transform a chaotic group of distributed resources into a small number of controllable, aggregated units similar to traditional power plants, which can then be uniformly interacted with and scheduled by the power grid master station. However, these existing aggregation models have fundamental limitations. Since they are mostly built based on offline historical statistical data or pre-set typical operating scenarios, they are inherently static and fixed. Because they cannot reflect the differences in the dynamic response characteristics of each unit within the resource group in real time and accurately, as well as the evolution of their collaborative potential at different time scales, there is a serious mismatch between the accuracy of the aggregation model and the requirements of real-time scheduling. Summary of the Invention

[0004] This invention provides a power grid dispatching method and system based on aggregation units, which can solve the problem in the prior art of reducing dispatching costs while ensuring power grid dispatching efficiency.

[0005] In a first aspect, embodiments of the present invention provide a power grid dispatching method based on aggregation units, comprising: Obtain the dynamic characteristic vectors corresponding to each distributed resource in the power grid; Based on the dynamic characteristic vectors corresponding to each of the distributed resources, a clustering algorithm is used to cluster each of the distributed resources to obtain each aggregation matching unit; Based on each of the aggregation matching units, a fast greedy multi-starting point selection algorithm is used to generate each candidate combination, and each candidate combination is filtered to obtain the final candidate combination; wherein, each candidate combination includes several aggregation matching units; A scheduling optimization model is established based on the final candidate combination, and the scheduling optimization model is solved to obtain the optimal power command corresponding to each aggregation matching unit in the final candidate combination. Then, the power grid is scheduled according to each optimal power command.

[0006] This application's embodiments acquire the dynamic characteristic vectors of each distributed resource in the power grid, enabling real-time and accurate capture of the dynamic response characteristics of each distributed resource, breaking the limitations of traditional static aggregation models that rely on offline data. By combining clustering algorithms to cluster distributed resources into aggregation matching units, geographically dispersed and characteristically diverse distributed resources can be scientifically and rationally classified and aggregated, significantly improving aggregation accuracy. A fast, greedy, multi-starting-point selection algorithm is used to generate and screen candidate combinations, efficiently exploring the collaborative potential of aggregation matching units and selecting final candidate combinations with greater optimization value. Finally, a scheduling optimization model is established based on the final candidate combinations, and the optimal power command is solved. This enables optimized control of power grid scheduling costs while satisfying scheduling variable constraints and power grid mapping constraints, while improving the scientific nature and execution efficiency of scheduling decisions, effectively balancing the efficiency and cost of power grid scheduling, and ensuring the reliability and flexibility of power grid operation.

[0007] As a preferred example of the first aspect, obtaining the dynamic characteristic vectors corresponding to each distributed resource in the power grid specifically involves: Obtain the operating data corresponding to each distributed resource in the power grid, and obtain the dynamic characteristic vector corresponding to each distributed resource based on the operating data corresponding to each distributed resource.

[0008] In this preferred example, by obtaining the operating data corresponding to each distributed resource of the power grid to generate its dynamic characteristic vector, the dynamic characteristic vector can be directly derived from the actual operating state of the resource, ensuring that the information carried by the vector is highly consistent with the real dynamic response of the distributed resource, and avoiding the vector information deviation that may be caused by relying on non-actual operating data.

[0009] As a preferred example of the first aspect, the step of clustering the distributed resources according to their dynamic characteristic vectors to obtain aggregate matching units using a clustering algorithm specifically involves: Based on the dynamic characteristic vectors corresponding to each of the distributed resources, the distributed resources are processed using a resource similarity formula and a clustering algorithm to obtain a cluster set; wherein, the cluster set includes several aggregation units; Based on the cluster set, the characteristic vector of each aggregation unit in the cluster set is obtained, and each aggregation unit in the cluster set and the characteristic vector of each aggregation unit are matched to obtain each aggregation matching unit.

[0010] In this preferred example, the distributed resources are first processed using the dynamic characteristic vectors of each distributed resource, combined with the resource similarity formula and clustering algorithm, to obtain a cluster set containing several aggregation units. The resource similarity formula can accurately quantify the degree of characteristic correlation between different distributed resources, and the clustering algorithm can efficiently group distributed resources with similar characteristics into one category, effectively improving the consistency of characteristics within the aggregation unit and avoiding the problem of insufficient aggregation accuracy caused by excessive differences in resource characteristics. Then, based on the cluster set, the aggregation unit characteristic vector corresponding to each aggregation unit is obtained, and the aggregation unit is matched with the corresponding characteristic vector to form an aggregation matching unit. This not only gives each aggregation unit a clear and quantifiable characteristic identifier, but also provides a clear characteristic basis for the subsequent selection, combination, and scheduling optimization of aggregation units, eliminating the need to analyze each dispersed distributed resource individually, greatly simplifying the subsequent processing flow. It also ensures that the invocation and control of aggregation units in subsequent stages are more in line with their actual characteristics, further guaranteeing the accuracy and rationality of power grid scheduling-related operations.

[0011] As a preferred example of the first aspect, the step of generating candidate combinations using a fast greedy multi-starting-point selection algorithm based on each of the aggregation matching units specifically involves: Randomly select several aggregation matching units from each of the aforementioned aggregation matching units as each initial unit; Traverse each initial unit, construct initialization candidate combinations based on the current initial unit, and generate candidate combinations corresponding to the current initial unit using a fast greedy multi-starting point selection algorithm based on the current initial unit and the current set of remaining units; wherein, the current set of remaining units is obtained based on each of the candidate combinations and the current initial unit. The candidate combinations corresponding to each initial unit are taken as each candidate combination.

[0012] In this preferred example, by randomly selecting several initial units from each aggregation matching unit, the limitation of local optima that are easily trapped in a single initial point can be effectively avoided. By using a multi-starting point layout, more potential combination exploration directions are covered, broadening the range of possible candidate combinations. Each initial unit is traversed and an initial candidate combination is constructed based on it. At the same time, combined with the set of remaining units determined by the candidate combination and the current initial unit, a fast greedy multi-starting point selection algorithm is used to generate corresponding candidate combinations. This not only improves the efficiency of candidate combination generation by leveraging its fast characteristics and reduces the time loss in the combination construction process, but also relies on greedy logic to select the unit with better current adaptability into the combination at each step, ensuring the basic quality of the combination under a single starting point. Furthermore, the introduction of the set of remaining units can avoid repeated unit calls, ensuring the independence and structural rationality of each candidate combination. Finally, the candidate combinations corresponding to each initial unit are summarized into an overall candidate combination, further enriching the quantity and diversity of candidate combinations. This provides sufficient and diverse selection criteria for the subsequent screening of final candidate combinations that meet the needs of power grid dispatching, reducing the risk of missing the optimal dispatching scheme due to insufficient candidate combination coverage. This lays a solid foundation for the accuracy and efficiency of subsequent power grid dispatching from the combination generation stage.

[0013] As a preferred example of the first aspect, the process of filtering each of the candidate combinations to obtain the final candidate combination specifically involves: Based on each aggregation matching unit in each candidate combination, the evaluation function value corresponding to each candidate combination is calculated using the evaluation function value calculation method. Based on the evaluation function values ​​corresponding to each candidate combination, the candidate combinations are sorted to obtain a sorting result, and the final candidate combination is determined based on the sorting result.

[0014] In this preferred example, the evaluation function value is obtained by calculating the evaluation function value of each aggregation matching unit within the candidate combination. This transforms the comprehensive performance of different candidate combinations into quantifiable objective indicators, avoiding the bias of relying on subjective experience judgment during the screening process. It provides a unified standard for measuring the merits of candidate combinations that were originally difficult to compare directly, fundamentally ensuring the fairness and accuracy of the screening. Furthermore, the candidate combinations are ranked and the final candidate combinations are determined based on the evaluation function values. This quickly identifies the priority order of each combination, eliminating the need for disordered one-by-one comparison of all combinations and greatly simplifying the screening process.

[0015] As a preferred example of the first aspect, the establishment of the scheduling optimization model based on the final candidate combination specifically includes: Based on the final candidate combinations, an objective function is established with the goal of minimizing grid dispatch costs. Based on the final candidate combinations, establish scheduling variable constraints and power grid mapping constraints; The scheduling optimization model is established based on the objective function, the scheduling variable constraints, and the power grid mapping constraints.

[0016] In this preferred example, an objective function is established based on the final candidate combination with the goal of minimizing grid dispatch costs. This provides a clear core orientation for the dispatch optimization model, ensuring that the optimization direction always focuses on cost control and avoiding the dispatch scheme from being biased towards non-critical indicators due to ambiguity of the objective. This provides guidance for reducing the economic losses of grid operation from the source. Furthermore, establishing dispatch variable constraints and grid mapping constraints based on the final candidate combination can regulate the reasonable value range of dispatch variables such as power commands in the model, effectively avoiding the problem that theoretical optimization solutions cannot be implemented due to deviation from the actual operating boundary of the grid, thus ensuring the feasibility of the dispatch scheme. Finally, the objective function and the two types of constraints are combined to construct a dispatch optimization model, which has both a clear cost optimization objective and is reasonably regulated by actual operating conditions. This achieves an organic balance between cost optimization and operational feasibility, providing a scientific and rigorous computational framework for the subsequent accurate solution of the optimal power commands of each aggregation matching unit, thereby promoting the efficient operation of grid dispatch while taking into account both economy and reliability.

[0017] Secondly, the present invention provides a power grid dispatching system based on aggregation units, comprising: a data acquisition module, a first processing module, a second processing module, and a dispatching module; The data acquisition module is used to acquire the dynamic characteristic vectors corresponding to each distributed resource in the power grid; The first processing module is used to cluster each of the distributed resources according to the dynamic characteristic vectors corresponding to each of the distributed resources using a clustering algorithm to obtain each aggregation matching unit; The second processing module is used to generate candidate combinations using a fast greedy multi-starting point selection algorithm based on each of the aggregation matching units, and to filter each of the candidate combinations to obtain the final candidate combination; wherein each candidate combination includes several aggregation matching units. The scheduling module is used to establish a scheduling optimization model based on the final candidate combination, solve the scheduling optimization model, obtain the optimal power command corresponding to each aggregation matching unit in the final candidate combination, and then schedule the power grid according to each optimal power command.

[0018] As a preferred example of the second aspect, obtaining the dynamic characteristic vectors corresponding to each distributed resource in the power grid specifically involves: Obtain the operating data corresponding to each distributed resource in the power grid, and obtain the dynamic characteristic vector corresponding to each distributed resource based on the operating data corresponding to each distributed resource.

[0019] As a preferred example of the second aspect, the step of clustering the distributed resources according to their dynamic characteristic vectors to obtain aggregate matching units using a clustering algorithm specifically involves: Based on the dynamic characteristic vectors corresponding to each of the distributed resources, the distributed resources are processed using a resource similarity formula and a clustering algorithm to obtain a cluster set; wherein, the cluster set includes several aggregation units; Based on the cluster set, the characteristic vector of each aggregation unit in the cluster set is obtained, and each aggregation unit in the cluster set and the characteristic vector of each aggregation unit are matched to obtain each aggregation matching unit.

[0020] As a preferred example of the second aspect, the step of generating candidate combinations using a fast greedy multi-starting-point selection algorithm based on each of the aggregation matching units specifically involves: Randomly select several aggregation matching units from each of the aforementioned aggregation matching units as each initial unit; Traverse each initial unit, construct initialization candidate combinations based on the current initial unit, and generate candidate combinations corresponding to the current initial unit using a fast greedy multi-starting point selection algorithm based on the current initial unit and the current set of remaining units; wherein, the current set of remaining units is obtained based on each of the candidate combinations and the current initial unit. The candidate combinations corresponding to each initial unit are taken as each candidate combination.

[0021] As a preferred example of the second aspect, the step of generating candidate combinations corresponding to the initial unit of the current traversal using a fast greedy multi-starting point selection algorithm based on the initial unit of the current traversal and the set of remaining units of the current traversal is specifically as follows: Based on the initial unit and the set of remaining units in the current traversal, a fast greedy multi-starting point selection algorithm is used to iteratively update the initialization candidate combination until the initialization candidate combination in the current iteration satisfies the first preset condition. The initialization candidate combination in the current iteration is then output as the candidate combination corresponding to the initial unit in the current traversal.

[0022] As a preferred example of the second aspect, the process of filtering each of the candidate combinations to obtain the final candidate combination specifically involves: Based on each aggregation matching unit in each candidate combination, the evaluation function value corresponding to each candidate combination is calculated using the evaluation function value calculation method. Based on the evaluation function values ​​corresponding to each candidate combination, the candidate combinations are sorted to obtain a sorting result, and the final candidate combination is determined based on the sorting result.

[0023] As a preferred example of the second aspect, the establishment of the scheduling optimization model based on the final candidate combination specifically includes: Based on the final candidate combinations, an objective function is established with the goal of minimizing grid dispatch costs. Based on the final candidate combinations, establish scheduling variable constraints and power grid mapping constraints; The scheduling optimization model is established based on the objective function, the scheduling variable constraints, and the power grid mapping constraints.

[0024] In summary, this application's embodiments, by acquiring the dynamic characteristic vectors of each distributed resource in the power grid, can capture the dynamic response characteristics of each distributed resource in real time and accurately, breaking the limitations of traditional static aggregation models that rely on offline data. By combining clustering algorithms to cluster distributed resources into aggregation matching units, geographically dispersed and characteristically diverse distributed resources can be scientifically and rationally classified and aggregated, significantly improving aggregation accuracy. By using a fast greedy multi-starting-point selection algorithm to generate and screen candidate combinations, the collaborative potential of aggregation matching units can be efficiently explored, selecting the final candidate combinations with greater optimization value. Finally, based on the final candidate combinations, a scheduling optimization model is established and the optimal power command is solved. This enables optimized control of power grid scheduling costs while satisfying scheduling variable constraints and power grid mapping constraints, while improving the scientific nature and execution efficiency of scheduling decisions, effectively balancing the efficiency and cost of power grid scheduling, and ensuring the reliability and flexibility of power grid operation.

[0025] Another embodiment of the present invention provides a terminal device, including: a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, it implements the steps of the power grid dispatching method based on aggregation unit of the present invention.

[0026] Another embodiment of the present invention provides a computer-readable storage medium item, including: a stored computer program, which, when the computer program is running, controls the device where the computer-readable storage medium is located to perform the steps of the power grid dispatching method based on aggregation unit of the present invention. Attached Figure Description

[0027] To more clearly illustrate the technical solution of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0028] Figure 1 This is a flowchart illustrating an embodiment of a power grid dispatching method based on aggregation units provided by the present invention. Figure 2This is a module structure diagram of an embodiment of a power grid dispatching system based on aggregation units provided by the present invention. Detailed Implementation

[0029] To make the objectives, technical solutions, and advantages of this application clearer, the technical solutions of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0030] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the application; the terms “comprising” and “having”, and any variations thereof, in the specification, claims, and foregoing description of the drawings are intended to cover non-exclusive inclusion.

[0031] In the description of the embodiments of this application, technical terms such as "first" and "second" are used only to distinguish different objects and should not be construed as indicating or implying relative importance or implicitly specifying the number, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, "multiple" means two or more, unless otherwise explicitly defined.

[0032] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0033] In the description of the embodiments in this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0034] In the description of the embodiments of this application, the term "multiple" refers to two or more (including two), similarly, "multiple sets" refers to two or more (including two sets), and "multiple pieces" refers to two or more (including two pieces).

[0035] In the description of the embodiments of this application, unless otherwise expressly specified and limited, technical terms such as "installation," "connection," "joining," and "fixing" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two components or the interaction between two components. For those skilled in the art, the specific meaning of the above terms in the embodiments of this application can be understood according to the specific circumstances.

[0036] Example 1 See Figure 1 To address the problem in existing technologies of simultaneously ensuring grid dispatch efficiency and reducing dispatch costs, an embodiment of the present invention provides a grid dispatch method based on aggregation units, comprising: S1. Obtain the dynamic characteristic vectors corresponding to each distributed resource in the power grid; In some embodiments of this application, obtaining the dynamic characteristic vector corresponding to each distributed resource in the power grid specifically involves: Obtain the operating data corresponding to each distributed resource in the power grid, and obtain the dynamic characteristic vector corresponding to each distributed resource based on the operating data corresponding to each distributed resource.

[0037] Specifically, obtaining the dynamic characteristic vectors corresponding to each distributed resource in the power grid can be implemented through the following preferred scheme: ① Definition of sets and indices. First, define the following finite sets: parent set Line collection Distributed resource collection Time step set ={1, ..., T}, where These represent wind power, solar power, battery energy storage, and flexible load aggregation, respectively, with each resource... Connected to the busbar All sets are finite sets, and the index is taken from the corresponding set.

[0038] ②Resource dynamic model and constraint expression. For any Define its power variable .

[0039] Wind / solar (renewable) power output: in, The amount of wind / solar curtailment represents the unutilized output of renewable energy. , The available power of resource r within time period t is determined by meteorological conditions or forecasts. To inject power for eventual grid connection.

[0040] Energy dynamics of an energy storage unit (ESS): in, The energy level of the energy storage unit at time t, and satisfying , , These are charging power and discharging power, respectively. These are the charge / discharge efficiencies, To schedule the time step, These are the upper and lower bounds of energy, respectively.

[0041] Flexible load The power is expressed as: in, , This represents the baseline load demand for time period t. For flexible adjustment, This is the maximum adjustable range.

[0042] And all resources satisfy the adjustment rate constraint: in, This represents the maximum climbing rate.

[0043] ③ The dynamic characteristic vector is defined as follows: in, To maximize available output, Maximum climbing speed, It is a typical time delay. As a measure of output uncertainty, These represent the upper and lower limits of energy for energy storage resources.

[0044] S2. Based on the dynamic characteristic vectors corresponding to each of the distributed resources, a clustering algorithm is used to cluster each of the distributed resources to obtain each aggregation matching unit; In some embodiments of this application, the step of clustering the distributed resources according to their dynamic characteristic vectors to obtain aggregate matching units specifically involves: Based on the dynamic characteristic vectors corresponding to each of the distributed resources, the distributed resources are processed using a resource similarity formula and a clustering algorithm to obtain a cluster set; wherein, the cluster set includes several aggregation units; Based on the cluster set, the characteristic vector of each aggregation unit in the cluster set is obtained, and each aggregation unit in the cluster set and the characteristic vector of each aggregation unit are matched to obtain each aggregation matching unit.

[0045] Specifically, to fully explain the above steps, the following scheme will be used as an example: ① Clustering and Grouping Based on the dynamic characteristic vector obtained in step S1, resource similarity is calculated using the following formula: in, Let be the weighted Euclidean distance between resource i and resource j, used to measure the similarity of their dynamic characteristics. The weights for the k-th feature dimension are... , Let be the attribute values ​​of resource i and resource j in the k-th feature dimension.

[0046] Then, clustering algorithms are used to... Clustering, forming cluster sets .

[0047] ② Equivalent modeling of aggregate units For each cluster set Define aggregation unit The specific formula is as follows: in, Aggregation unit The maximum productive output is the sum of the maximum productive outputs of all resources of that type. To maximize the active power output of resource r. Aggregation unit The maximum climbing rate indicates the fastest speed at which the overall power can be adjusted. The maximum ramp rate for a single resource r.

[0048] Aggregate unit characteristic vector The definition formula is as follows: in, Aggregation unit The minimum active power, These are the upper and lower limits of the energy constraint for the aggregation unit. The dynamic response time constant reflects the inertia / delay characteristics of the aggregation unit to scheduling commands, and represents the fastest speed at which the overall power can be adjusted. This is an uncertainty parameter, typically the standard deviation of the prediction error, reflecting the reproducibility and randomness of the unit.

[0049] ③Equivalent network impact in, Let t be the line power flow. Aggregation unit PTDF vector, The active power injection of the aggregation unit at time t. Inject power into conventional power sources or external power grids that are not aggregated in the system.

[0050] ④ Model library generation All aggregation units Along with its characteristics Stored in the aggregate model library (That is, each of the aforementioned aggregation matching units).

[0051] S3. Based on each of the aggregation matching units, a fast greedy multi-starting point selection algorithm is used to generate each candidate combination, and each candidate combination is filtered to obtain the final candidate combination; wherein, each candidate combination includes several aggregation matching units. In some embodiments of this application, the step of generating candidate combinations using a fast greedy multi-starting-point selection algorithm based on each of the aggregation matching units specifically involves: Randomly select several aggregation matching units from each of the aforementioned aggregation matching units as each initial unit; Traverse each initial unit, construct initialization candidate combinations based on the current initial unit, and generate candidate combinations corresponding to the current initial unit using a fast greedy multi-starting point selection algorithm based on the current initial unit and the current set of remaining units; wherein, the current set of remaining units is obtained based on each of the candidate combinations and the current initial unit. The candidate combinations corresponding to each initial unit are taken as each candidate combination.

[0052] In some embodiments of this application, the step of generating candidate combinations corresponding to the initial unit of the current traversal using a fast greedy multi-starting point selection algorithm based on the initial unit of the current traversal and the set of remaining units of the current traversal specifically involves: In some embodiments of this application, the step of filtering each of the candidate combinations to obtain the final candidate combination specifically involves: Based on each aggregation matching unit in each candidate combination, the evaluation function value corresponding to each candidate combination is calculated using the evaluation function value calculation method. Based on the evaluation function values ​​corresponding to each candidate combination, the candidate combinations are sorted to obtain a sorting result, and the final candidate combination is determined based on the sorting result.

[0053] Based on the initial unit and the set of remaining units in the current traversal, a fast greedy multi-starting point selection algorithm is used to iteratively update the initialization candidate combination until the initialization candidate combination in the current iteration satisfies the first preset condition. The initialization candidate combination in the current iteration is then output as the candidate combination corresponding to the initial unit in the current traversal.

[0054] Specifically, to fully explain the above steps, the following scheme will be used as an example: ① Candidate solution generation At runtime t, from Generate candidate combinations Each It is a combination of several units.

[0055] Candidate Combinations The fast greedy multi-starting-point selection algorithm is used for generation, and the specific steps are as follows: (1) Starting point set setting: from the aggregation unit Select several initial units according to resource type. , forming the starting set : in, Set the preset number of candidates (typically 3–10).

[0056] (2) For each starting point Initialize candidate combinations: In each iteration, select units from the remaining units that maximize the overall marginal return. Largest unit join in ,in The expression is as follows: in, The weighting factor is the response speed, capacity, and latency. The system's average ramp rate and average power are given. This is a penalty weight used to balance the cost of uncertainty.

[0057] Select the one that satisfies and update .

[0058] (3) If the total regulation capacity or response speed of the current combination meets the requirements Terminate this round of greedy construction and output candidate combinations. .

[0059] (4) For all starting points Repeat steps (2) and (3) to obtain the candidate combination set. .

[0060] ② Calculation of performance evaluation indicators (1) The formula for calculating the group response speed is as follows: in, For combination The group response rate index at time t, The maximum ramp rate of polymer unit U. This represents the maximum available active power of the aggregation unit U.

[0061] (2) The formula for calculating flexibility is as follows: in, For combination The flexibility index at time t, The output of aggregation unit U that has been scheduled at time t. This represents the maximum available active power of the aggregation unit U.

[0062] (3) The formula for calculating control accuracy is as follows: in, For combination The control accuracy index at time t This represents the average power tracking deviation of the combined output, i.e., the average difference between the dispatch command and the actual output. The standard deviation of the combined power deviation It is the standard normal distribution function.

[0063] ③ The formula for calculating the overall score is as follows: in, Candidate combinations The overall score at time t, These are the weighting coefficients, These are normalized indices for group response speed, flexibility, and control precision, respectively.

[0064] The combination with the highest score will be selected. As the final candidate combination.

[0065] S4. Establish a scheduling optimization model based on the final candidate combination, solve the scheduling optimization model, obtain the optimal power command corresponding to each aggregation matching unit in the final candidate combination, and then schedule the power grid according to each optimal power command.

[0066] In some embodiments of this application, the step of establishing a scheduling optimization model based on the final candidate combination specifically includes: Based on the final candidate combinations, an objective function is established with the goal of minimizing grid dispatch costs. Based on the final candidate combinations, establish scheduling variable constraints and power grid mapping constraints; The scheduling optimization model is established based on the objective function, the scheduling variable constraints, and the power grid mapping constraints.

[0067] Specifically, to fully explain the above steps, the following scheme will be used as an example: The optimal aggregation combination output in step S3 is directly used as the scheduling variable set of the optimization model in step S4, and its member parameters are used to limit the adjustable boundary and dynamic constraints. Through optimization, the optimal power command of each aggregation unit in the optimal combination in step S3 is obtained, realizing real-time interaction between the aggregation layer and the grid layer. The proposal process of the scheduling optimization model is as follows: ①Scheduling variable constraints in, Let be the active power output / schedule of aggregation unit U at time t. Let be the minimum and maximum active power output boundaries of the aggregation unit U at time t, respectively. The maximum ramp rate of the polymer unit U.

[0068] ② Power grid mapping constraints in, Let be the injected power vector of each node in the system at time t. This is the power flow distribution factor matrix, reflecting the linear impact of nodal power injection on the power flow of each transmission line. For the power flow on each transmission line at time t, This is the maximum permissible power flow capacity of the transmission line.

[0069] ③ Objective function in, Let be the power generation cost function of a conventional generating unit at time t, with variables . Indicates the unit's output or status. Let be the scheduling cost function of aggregation unit U at time t.

[0070] Finally, commercial solvers can be used to optimize the solution of the scheduling optimization model.

[0071] In summary, this application's embodiments, by acquiring the dynamic characteristic vectors of each distributed resource in the power grid, can capture the dynamic response characteristics of each distributed resource in real time and accurately, breaking the limitations of traditional static aggregation models that rely on offline data. By combining clustering algorithms to cluster distributed resources into aggregation matching units, geographically dispersed and characteristically diverse distributed resources can be scientifically and rationally classified and aggregated, significantly improving aggregation accuracy. By using a fast greedy multi-starting-point selection algorithm to generate and screen candidate combinations, the collaborative potential of aggregation matching units can be efficiently explored, selecting the final candidate combinations with greater optimization value. Finally, based on the final candidate combinations, a scheduling optimization model is established and the optimal power command is solved. This enables optimized control of power grid scheduling costs while satisfying scheduling variable constraints and power grid mapping constraints, while improving the scientific nature and execution efficiency of scheduling decisions, effectively balancing the efficiency and cost of power grid scheduling, and ensuring the reliability and flexibility of power grid operation.

[0072] Example 2 like Figure 2 As shown, based on the above method embodiments, corresponding device embodiments are provided; An embodiment of the present invention provides a power grid dispatching system based on aggregation units, including: a data acquisition module 21, a first processing module 22, a second processing module 23, and a dispatching module 24; Data acquisition module 21 is used to acquire the dynamic characteristic vectors corresponding to each distributed resource in the power grid; The first processing module 22 is used to cluster each of the distributed resources according to the dynamic characteristic vectors corresponding to each of the distributed resources, and obtain each aggregation matching unit. The second processing module 23 is used to generate candidate combinations according to each of the aggregation matching units using a fast greedy multi-starting point selection algorithm, and to filter each of the candidate combinations to obtain the final candidate combination; wherein each candidate combination includes several aggregation matching units. The scheduling module 24 is used to establish a scheduling optimization model based on the final candidate combination, solve the scheduling optimization model, obtain the optimal power command corresponding to each aggregation matching unit in the final candidate combination, and then schedule the power grid according to each optimal power command.

[0073] In some embodiments of this application, obtaining the dynamic characteristic vector corresponding to each distributed resource in the power grid specifically involves: Obtain the operating data corresponding to each distributed resource in the power grid, and obtain the dynamic characteristic vector corresponding to each distributed resource based on the operating data corresponding to each distributed resource.

[0074] As a preferred example of the second aspect, the step of clustering the distributed resources according to their dynamic characteristic vectors to obtain aggregate matching units using a clustering algorithm specifically involves: Based on the dynamic characteristic vectors corresponding to each of the distributed resources, the distributed resources are processed using a resource similarity formula and a clustering algorithm to obtain a cluster set; wherein, the cluster set includes several aggregation units; Based on the cluster set, the characteristic vector of each aggregation unit in the cluster set is obtained, and each aggregation unit in the cluster set and the characteristic vector of each aggregation unit are matched to obtain each aggregation matching unit.

[0075] In some embodiments of this application, the step of generating candidate combinations using a fast greedy multi-starting-point selection algorithm based on each of the aggregation matching units specifically involves: Randomly select several aggregation matching units from each of the aforementioned aggregation matching units as each initial unit; Traverse each initial unit, construct initialization candidate combinations based on the current initial unit, and generate candidate combinations corresponding to the current initial unit using a fast greedy multi-starting point selection algorithm based on the current initial unit and the current set of remaining units; wherein, the current set of remaining units is obtained based on each of the candidate combinations and the current initial unit. The candidate combinations corresponding to each initial unit are taken as each candidate combination.

[0076] In some embodiments of this application, the step of generating candidate combinations corresponding to the initial unit of the current traversal using a fast greedy multi-starting point selection algorithm based on the initial unit of the current traversal and the set of remaining units of the current traversal specifically involves: Based on the initial unit and the set of remaining units in the current traversal, a fast greedy multi-starting point selection algorithm is used to iteratively update the initialization candidate combination until the initialization candidate combination in the current iteration satisfies the first preset condition. The initialization candidate combination in the current iteration is then output as the candidate combination corresponding to the initial unit in the current traversal.

[0077] In some embodiments of this application, the step of filtering each of the candidate combinations to obtain the final candidate combination specifically involves: Based on each aggregation matching unit in each candidate combination, the evaluation function value corresponding to each candidate combination is calculated using the evaluation function value calculation method. Based on the evaluation function values ​​corresponding to each candidate combination, the candidate combinations are sorted to obtain a sorting result, and the final candidate combination is determined based on the sorting result.

[0078] In some embodiments of this application, the step of establishing a scheduling optimization model based on the final candidate combination specifically includes: Based on the final candidate combinations, an objective function is established with the goal of minimizing grid dispatch costs. Based on the final candidate combinations, establish scheduling variable constraints and power grid mapping constraints; The scheduling optimization model is established based on the objective function, the scheduling variable constraints, and the power grid mapping constraints.

[0079] For more detailed steps and working principles of this embodiment, please refer to the relevant description in Embodiment 1, but not limited to these descriptions.

[0080] In summary, this application's embodiments, by acquiring the dynamic characteristic vectors of each distributed resource in the power grid, can capture the dynamic response characteristics of each distributed resource in real time and accurately, breaking the limitations of traditional static aggregation models that rely on offline data. By combining clustering algorithms to cluster distributed resources into aggregation matching units, geographically dispersed and characteristically diverse distributed resources can be scientifically and rationally classified and aggregated, significantly improving aggregation accuracy. By using a fast greedy multi-starting-point selection algorithm to generate and screen candidate combinations, the collaborative potential of aggregation matching units can be efficiently explored, selecting the final candidate combinations with greater optimization value. Finally, based on the final candidate combinations, a scheduling optimization model is established and the optimal power command is solved. This enables optimized control of power grid scheduling costs while satisfying scheduling variable constraints and power grid mapping constraints, while improving the scientific nature and execution efficiency of scheduling decisions, effectively balancing the efficiency and cost of power grid scheduling, and ensuring the reliability and flexibility of power grid operation.

[0081] It is understood that the above-described device embodiments correspond to the method embodiments of the present invention, and can implement the power grid dispatching method based on aggregation unit provided by any of the above-described method embodiments of the present invention.

[0082] It should be noted that the device embodiments described above are merely illustrative, and some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Furthermore, in the accompanying drawings of the device embodiments provided by this invention, the connection relationships between modules indicate that they have communication connections, which can specifically be implemented as one or more communication buses or signal lines. Those skilled in the art can understand and implement this without any creative effort.

[0083] Example 3 Based on the above embodiments of the grid dispatching method based on aggregation units, another embodiment of the present invention provides a terminal device, which includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the grid dispatching method based on aggregation units of any embodiment of the present invention.

[0084] For example, in this embodiment, the computer program can be divided into one or more modules, which are stored in the memory and executed by the processor to complete the present invention. The one or more modules may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of the computer program in the terminal device.

[0085] The terminal device may be a desktop computer, laptop, handheld computer, or cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.

[0086] The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor. The processor is the control center of the terminal device, connecting all parts of the terminal device via various interfaces and lines.

[0087] Example 4 Based on the above-described method embodiments, another embodiment of the present invention provides a computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device where the computer-readable storage medium is located to execute the power grid dispatching method based on aggregation units as described in any of the above-described method embodiments of the present invention.

[0088] The modules / units integrated in the device / terminal equipment, if implemented as software functional units and sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, all or part of the processes in the above embodiments of the present invention can also be implemented by a computer program instructing related hardware. The computer program can be stored in a computer-readable storage medium, and when executed by a processor, it can implement the steps of the various method embodiments described above. The computer program includes computer program code, which can be in the form of source code, object code, executable files, or certain intermediate forms. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium, etc.

[0089] The above description represents the preferred embodiments of the present invention. It should be noted that those skilled in the art can make various improvements and modifications without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention.

Claims

1. A power grid dispatching method based on aggregation units, characterized in that, include: Obtain dynamic characteristic vectors corresponding to each distributed resource in the power grid to characterize the real-time dynamic response characteristics of the distributed resources; wherein, the elements of the dynamic characteristic vectors include maximum available output, maximum ramp rate, typical time delay, output uncertainty metric, upper limit of energy storage resources, and lower limit of energy storage resources; Based on the dynamic characteristic vectors corresponding to each of the distributed resources, a clustering algorithm is used to cluster each of the distributed resources to obtain each aggregation matching unit; Based on each of the aggregation matching units, a fast greedy multi-starting point selection algorithm is used to generate each candidate combination, and each candidate combination is filtered to obtain the final candidate combination; wherein, each candidate combination includes several aggregation matching units; A scheduling optimization model is established based on the final candidate combination, and the scheduling optimization model is solved to obtain the optimal power command corresponding to each aggregation matching unit in the final candidate combination. Then, the power grid is scheduled according to each optimal power command. The step of generating candidate combinations using a fast greedy multi-starting-point selection algorithm based on each of the aggregation matching units is as follows: Randomly select several aggregation matching units from each of the aforementioned aggregation matching units as each initial unit; Traverse each initial unit, construct initialization candidate combinations based on the current initial unit, and generate candidate combinations corresponding to the current initial unit using a fast greedy multi-starting point selection algorithm based on the current initial unit and the current set of remaining units; wherein, the current set of remaining units is obtained based on each of the candidate combinations and the current initial unit. The candidate combinations corresponding to each initial unit are taken as each candidate combination.

2. The power grid dispatching method based on aggregation units as described in claim 1, characterized in that, The specific steps for obtaining the dynamic characteristic vectors corresponding to each distributed resource in the power grid are as follows: Obtain the operating data corresponding to each distributed resource in the power grid, and obtain the dynamic characteristic vector corresponding to each distributed resource based on the operating data corresponding to each distributed resource.

3. The power grid dispatching method based on aggregation units as described in claim 1, characterized in that, The step of clustering each distributed resource based on its dynamic characteristic vector to obtain each aggregation matching unit is as follows: Based on the dynamic characteristic vectors corresponding to each of the distributed resources, the distributed resources are processed using a resource similarity formula and a clustering algorithm to obtain a cluster set; wherein, the cluster set includes several aggregation units; Based on the cluster set, the characteristic vector of each aggregation unit in the cluster set is obtained, and each aggregation unit in the cluster set and the characteristic vector of each aggregation unit are matched to obtain each aggregation matching unit.

4. The power grid dispatching method based on aggregation units as described in claim 1, characterized in that, The process of generating candidate combinations corresponding to the initial unit of the current traversal using a fast greedy multi-starting point selection algorithm, based on the initial unit and the remaining unit set of the current traversal, is as follows: Based on the initial unit and the set of remaining units in the current traversal, a fast greedy multi-starting point selection algorithm is used to iteratively update the initialization candidate combination until the initialization candidate combination in the current iteration satisfies the first preset condition. The initialization candidate combination in the current iteration is then output as the candidate combination corresponding to the initial unit in the current traversal.

5. The power grid dispatching method based on aggregation units as described in claim 1, characterized in that, The process of filtering each candidate combination to obtain the final candidate combination is as follows: Based on each aggregation matching unit in each candidate combination, the evaluation function value corresponding to each candidate combination is calculated using the evaluation function value calculation method. Based on the evaluation function values ​​corresponding to each candidate combination, the candidate combinations are sorted to obtain a sorting result, and the final candidate combination is determined based on the sorting result.

6. The power grid dispatching method based on aggregation units as described in claim 1, characterized in that, The step of establishing a scheduling optimization model based on the final candidate combination specifically involves: Based on the final candidate combinations, an objective function is established with the goal of minimizing grid dispatch costs. Based on the final candidate combinations, establish scheduling variable constraints and power grid mapping constraints; The scheduling optimization model is established based on the objective function, the scheduling variable constraints, and the power grid mapping constraints.

7. A power grid dispatching system based on aggregation units, characterized in that, include: The system comprises a data acquisition module, a first processing module, a second processing module, and a scheduling module. The data acquisition module is used to acquire dynamic characteristic vectors corresponding to each distributed resource in the power grid, which characterize the real-time dynamic response characteristics of the distributed resources; wherein, the elements of the dynamic characteristic vectors include maximum available output, maximum ramp rate, typical time delay, output uncertainty metric, upper limit of energy storage resources, and lower limit of energy storage resources. The first processing module is used to cluster each of the distributed resources according to the dynamic characteristic vectors corresponding to each of the distributed resources using a clustering algorithm to obtain each aggregation matching unit; The second processing module is used to generate candidate combinations using a fast greedy multi-starting point selection algorithm based on each of the aggregation matching units, and to filter each of the candidate combinations to obtain the final candidate combination; wherein each candidate combination includes several aggregation matching units. The scheduling module is used to establish a scheduling optimization model based on the final candidate combination, solve the scheduling optimization model, obtain the optimal power command corresponding to each aggregation matching unit in the final candidate combination, and then schedule the power grid according to each optimal power command. The step of generating candidate combinations using a fast greedy multi-starting-point selection algorithm based on each of the aggregation matching units is as follows: Randomly select several aggregation matching units from each of the aforementioned aggregation matching units as each initial unit; Traverse each initial unit, construct initialization candidate combinations based on the current initial unit, and generate candidate combinations corresponding to the current initial unit using a fast greedy multi-starting point selection algorithm based on the current initial unit and the current set of remaining units; wherein, the current set of remaining units is obtained based on each of the candidate combinations and the current initial unit. The candidate combinations corresponding to each initial unit are taken as each candidate combination.

8. A power grid dispatching system based on aggregation units as described in claim 7, characterized in that, The specific steps for obtaining the dynamic characteristic vectors corresponding to each distributed resource in the power grid are as follows: Obtain the operating data corresponding to each distributed resource in the power grid, and obtain the dynamic characteristic vector corresponding to each distributed resource based on the operating data corresponding to each distributed resource.

9. A power grid dispatching system based on aggregation units as described in claim 7, characterized in that, The step of clustering each distributed resource based on its dynamic characteristic vector to obtain each aggregation matching unit is as follows: Based on the dynamic characteristic vectors corresponding to each of the distributed resources, the distributed resources are processed using a resource similarity formula and a clustering algorithm to obtain a cluster set; wherein, the cluster set includes several aggregation units; Based on the cluster set, the characteristic vector of each aggregation unit in the cluster set is obtained, and each aggregation unit in the cluster set and the characteristic vector of each aggregation unit are matched to obtain each aggregation matching unit.

Images