A resource scheduling method and system for edge computing in power distribution networks

Abstract

This application discloses a resource scheduling method and system for distribution network edge computing, comprising: obtaining random characteristics corresponding to each power business scenario based on acquired historical wind power data, historical photovoltaic data, and historical load data, and obtaining resource requirements corresponding to each power business scenario based on the random characteristics; constructing a spring model corresponding to each power business scenario by using the resource utilization rate as the spring length and determining the elasticity coefficient based on the business flexibility value and decision risk value; determining the resource adjustment value corresponding to each power business scenario based on the resource requirements and the spring model, and performing resource scheduling for each power business scenario based on the resource adjustment value. This application can adapt to the resource requirements of different power businesses, improve resource scheduling capabilities in complex environments, and provide support for the elastic resource management of distribution network edge computing nodes.

Description

Technical Field

[0001] This application relates to the field of resource scheduling, and in particular to a resource scheduling method and system for edge computing in power distribution networks. Background Technology

[0002] Driven by both policy guidance and technological advancements, the power distribution network is undergoing an unprecedented transformation. Specifically, on the generation side, distributed generation of renewable energy sources such as photovoltaics has experienced explosive growth in the distribution network, a trend that has significantly exacerbated the uncertainty of current flow and the complexity of its dynamic characteristics. On the consumption side, electric vehicles, as a representative of new loads, have brought about a large amount of random charging behavior due to their large-scale grid connection. The superimposed characteristics of these new power sources and loads force edge computing nodes—the actual physical carriers performing tasks—to face an environment with more severe time variability, volatility, and uncertainty. Therefore, in this context, how can edge computing nodes explore and implement optimal strategies under limited computing resources and capacity constraints to ensure that various power service demands can be met in a timely and effective manner?

[0003] Existing edge computing nodes often use a fixed approach to configure power services and resources, which fails to fully adapt to the resource requirements of different power services and makes it difficult to truly reflect the correlation between power task resources and the distribution network system, resulting in insufficient resource scheduling capabilities in complex time-varying environments. Summary of the Invention

[0004] This application provides a resource scheduling method and system for edge computing in distribution networks, which can adapt to the resource needs of different power services, improve the resource scheduling capability in complex environments, and provide support for the elastic resource management of edge computing nodes in distribution networks.

[0005] Firstly, this application provides a resource scheduling method for edge computing in distribution networks, including:

[0006] Based on the historical wind power data, historical photovoltaic data and historical load data corresponding to each power business scenario, the random characteristics corresponding to each power business scenario are obtained, and the resource requirements corresponding to each power business scenario are obtained based on the random characteristics. The random characteristics include wind power prediction error, photovoltaic prediction error and load uncertainty.

[0007] Using the resource utilization rate as the spring length, the elasticity coefficient is determined based on the business flexibility value and the decision risk value, and a spring model corresponding to each power business scenario is constructed. The business flexibility value is the flexibility value of the edge computing node in resource adjustment, and the decision risk value is the risk value caused by the uncertainty of the power business scenario.

[0008] Based on the resource requirements and the spring model, the resource adjustment value corresponding to each power business scenario is determined, and resource scheduling is performed on each power business scenario based on the resource adjustment value.

[0009] This application's embodiments determine the corresponding resource requirements based on the random characteristics of each power business scenario, which can accurately determine the resource requirements of each power business scenario, facilitating subsequent resource scheduling for each power business scenario based on resource requirements. By using the resource utilization rate as the spring length and determining the elasticity coefficient based on the business flexibility value and decision risk value, a spring model corresponding to each power business scenario can be constructed. This allows for the construction of spring models that adapt to different resource scheduling requirements under the source-load fluctuation scenario of the distribution network, providing support for the elastic resource management of edge computing nodes in the distribution network. Through the spring model, the resource adjustment value corresponding to each power business scenario can be accurately determined, and resource scheduling for each power business scenario can be performed based on the resource adjustment value, adapting to the resource requirements of different power businesses and improving the resource scheduling capability in complex environments.

[0010] Furthermore, the process of obtaining the resource requirements corresponding to each power business scenario based on the random features specifically involves:

[0011] Construct a corresponding probability distribution function based on the aforementioned random features;

[0012] The source load data corresponding to the random feature is obtained through the probability distribution function;

[0013] By clustering the source-load data, the resource requirements corresponding to each power business scenario are obtained.

[0014] By obtaining the random features corresponding to each power business scenario, the corresponding probability distribution function is determined, and then the corresponding resource requirements are accurately determined through clustering. This allows for accurate determination of the resource requirements for each power business scenario, facilitating subsequent resource scheduling based on these requirements.

[0015] Furthermore, the formula for the probability distribution function is as follows:

[0016] ；

[0017] In the formula, Let be the joint probability distribution function. For wind power prediction error, For photovoltaic prediction error, For load uncertainty, , These are the distribution functions corresponding to the respective random features. For the Copula function corresponding to the random feature, It is a random number. It is the symbol for a partial derivative in mathematics.

[0018] Furthermore, by clustering the source-load data, the resource requirements corresponding to each power business scenario are obtained, specifically:

[0019] Based on the source load data, the initial resource requirements corresponding to each power business scenario are determined;

[0020] A similarity matrix is ​​established based on the similarity between any two initial resource requirements, and a corresponding diagonal matrix is ​​constructed based on the similarity matrix.

[0021] Based on the diagonal matrix and the similarity matrix, the Laplace matrix is ​​obtained;

[0022] Extract the eigenvalues ​​from the Laplacian matrix, and form a feature matrix based on the eigenvectors corresponding to the eigenvalues;

[0023] By performing k-means clustering on the feature matrix, the resource requirements corresponding to each power business scenario are obtained.

[0024] By clustering source and load data, the resource requirements for each power business scenario can be accurately determined, facilitating subsequent resource scheduling for each power business scenario based on these requirements.

[0025] Furthermore, the calculation formula for the resource requirements corresponding to each power business scenario is as follows:

[0026] ;

[0027] In the formula, For the set of resource requirements, , To determine the similarity between subsets of resource demand, , The first , A subset of resource requirements, Let be the probability density function. It is the set of resource requirements. The corresponding subset, yes The joint probability distribution, It is the symbol for a partial derivative in mathematics.

[0028] Furthermore, the evaluation formula for the business flexibility value is as follows:

[0029] ;

[0030] In the formula, This is the value representing the business flexibility. ·) indicates rounding up; , These represent the weights of task value and execution time, respectively, satisfying... ; For the value of the task; For time value, of which, , This represents the maximum static flexibility. Indicates task The deadline for adjustment, This refers to the current moment.

[0031] Furthermore, the determination of resource adjustment values ​​corresponding to each power business scenario based on the resource requirements and the spring model specifically involves:

[0032] ;

[0033] In the formula, For time t, the first The deformation of each spring represents the resource adjustment value for power business scenario i; This represents the resource utilization rate of edge computing nodes; For time t, the first The original length of each spring represents the resource requirements of power business scenario i. For time t, the first The spring constant of a spring, wherein the resource utilization rate The calculation formula is:

[0034] ;

[0035] In the formula: For power business scenarios In time The amount of resources allocated; This represents the average resource utilization rate. The number of power business scenarios; The total time period for resource adjustments; For edge computing nodes in time Resource utilization rate; For power business Total resource requirements; and These are the minimum and maximum resource usage values ​​of power service i at time t, respectively. For power business scenarios Resource allocation weights; This represents a value for business flexibility.

[0036] Furthermore, the spring model includes: an objective function and constraints. Based on the objective function and constraints, a spring model with optimal resource utilization is established, specifically as follows:

[0037] The objective function is expressed as:

[0038] ;

[0039] In the formula, This is the original length of the composite spring; For the composite spring in time The deformable variables represent resource adjustment values ​​for multiple power business scenarios; The average spring length represents the average resource requirement. As a symbolic variable, when the spring stretches... During compression ; This refers to the total adjustment time of the compound spring; Let be the spring constant at node t; For edge computing nodes in time The change in resource utilization rate; It is a constant; It is the minimum resource utilization rate of the edge computing node at time t; It represents the maximum resource utilization rate of the edge computing node at time t.

[0040] Secondly, this application provides a resource scheduling system for edge computing of distribution networks, comprising: a resource scheduling system for edge computing of distribution networks, characterized in that it includes: an acquisition module, a construction module and a scheduling module;

[0041] The acquisition module is used to obtain random features corresponding to each power business scenario based on the historical wind power data, historical photovoltaic data and historical load data obtained for each power business scenario, and to obtain the resource requirements corresponding to each power business scenario based on the random features. The random features include wind power prediction error, photovoltaic prediction error and load uncertainty.

[0042] The construction module is used to construct a spring model corresponding to each power business scenario by using the resource utilization rate as the spring length, determining the elasticity coefficient based on the business flexibility value and the decision risk value, wherein the business flexibility value is the flexibility value of the edge computing node in resource adjustment, and the decision risk value is the risk value caused by the uncertainty of the power business scenario;

[0043] The scheduling module is used to determine the resource adjustment value corresponding to each power business scenario based on the resource demand and the spring model, and to perform resource scheduling for each power business scenario based on the resource adjustment value.

[0044] This application's embodiments determine the corresponding resource requirements based on the random characteristics of each power business scenario, which can accurately determine the resource requirements of each power business scenario, facilitating subsequent resource scheduling for each power business scenario based on resource requirements. By using the resource utilization rate as the spring length and determining the elasticity coefficient based on the business flexibility value and decision risk value, a spring model corresponding to each power business scenario can be constructed. This allows for the construction of spring models that adapt to different resource scheduling requirements under the source-load fluctuation scenario of the distribution network, providing support for the elastic resource management of edge computing nodes in the distribution network. Through the spring model, the resource adjustment value corresponding to each power business scenario can be accurately determined, and resource scheduling for each power business scenario can be performed based on the resource adjustment value, adapting to the resource requirements of different power businesses and improving the resource scheduling capability in complex environments. Attached Figure Description

[0045] Figure 1 This is a flowchart illustrating an embodiment of a resource scheduling method for edge computing in a power distribution network provided in this application.

[0046] Figure 2 This is a schematic diagram of the area for adjusting edge computing resources used in power business scenarios provided in this application;

[0047] Figure 3 This is a flowchart illustrating one embodiment of a resource scheduling system for edge computing in a power distribution network provided in this application. Detailed Implementation

[0048] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of this application.

[0049] It should be understood that the step numbers used in the text are for ease of description only and are not intended to limit the order in which the steps are performed.

[0050] It should be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0051] The terms “comprising” and “including” indicate the presence of the described feature, whole, step, operation, element and / or component, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or collections thereof.

[0052] The term “and / or” refers to any combination of one or more of the associated listed items, as well as all possible combinations, and includes these combinations.

[0053] The power distribution network is undergoing an unprecedented transformation. Specifically, on the generation side, the surge in renewable energy sources such as photovoltaics has exacerbated the uncertainty of current flow and the complexity of its dynamic characteristics. On the consumption side, the large-scale integration of electric vehicles has brought about a large amount of random charging behavior. Therefore, edge computing nodes face more severe challenges related to time, fluctuations, and uncertainties. Against this backdrop, edge computing nodes need to explore optimal strategies with limited resources to meet the needs of power services, which has become a technical challenge. Existing technologies use fixed methods to allocate power services and resources, which cannot adapt to the needs of power services and have insufficient resource scheduling capabilities.

[0054] Next, the terms used in this application will be explained:

[0055] Clustering algorithms are core technologies in data mining and pattern recognition. They group objects in a dataset so that objects within the same group have high similarity, while objects in different groups have low similarity. These algorithms can be widely used in image processing, market analysis, bioinformatics, and other fields. Through clustering, people can understand the structure and characteristics of data, discover hidden patterns and regularities, and thus conduct more effective data analysis and decision-making. Common clustering algorithms include K-means clustering, hierarchical clustering, and DBSCAN.

[0056] Based on this, the embodiments of this application provide a resource scheduling method and system for distribution network edge computing, which can adapt to the resource needs of different power services, improve the resource scheduling capability in complex environments, and provide support for the elastic resource management of distribution network edge computing nodes.

[0057] This application provides a resource scheduling method for edge computing of distribution networks, which is specifically described through the following embodiments. First, the resource scheduling method for edge computing of distribution networks in this application is described.

[0058] The resource scheduling method for edge computing in power distribution networks provided in this application relates to the field of resource scheduling. This method can be applied to a terminal, a server, or software running on either a terminal or a server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, etc.; the server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms; the software can implement a component selection recommendation method, etc., but is not limited to the above forms.

[0059] This application can also be used in numerous general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics devices, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This application can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0060] Example 1

[0061] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating an embodiment of a resource scheduling method for edge computing in a power distribution network provided in this application, including steps S101 to S103;

[0062] Step S101: Based on the historical wind speed data, historical photovoltaic data and historical load data corresponding to each power business scenario, obtain the random characteristics corresponding to each power business scenario, and obtain the resource requirements corresponding to each power business scenario based on the random characteristics, wherein the random characteristics include wind power prediction error, photovoltaic prediction error and load uncertainty.

[0063] It is understandable that existing technologies can be used to obtain historical wind power data, historical photovoltaic data, and historical load data corresponding to various power business scenarios. Specifically, this can be achieved, but is not limited to, obtaining wind speed statistics by accessing energy meteorological data platforms, government research institutions, wind power companies, and industry associations; obtaining photovoltaic power generation data by accessing energy meteorological data platforms, government websites, public data platforms, photovoltaic companies, and research institutions; and collecting historical power load data by using load-side management platforms, among others. Once the historical wind speed data, historical photovoltaic data, and historical load data corresponding to various power business scenarios are obtained, data analysis can be performed on these data to determine the random characteristics such as wind power prediction errors, photovoltaic prediction errors, and load uncertainties during the data collection period. It should be noted that the methods of data analysis are not the focus of this application, and therefore will not be elaborated upon here.

[0064] After obtaining the random characteristics corresponding to each power business scenario, it is necessary to obtain the resource requirements corresponding to each power business scenario based on the random characteristics. Specifically, the following steps are taken: First, a corresponding probability distribution function is constructed based on the random characteristics, that is, considering the impact of demand-side load and renewable energy fluctuations on the uncertainty of power business. Scenario construction is performed using the probability density functions of wind power prediction error, photovoltaic prediction error, and load uncertainty, and correlation modeling is performed on the high-dimensional data. Finally, a joint probability distribution function of source-load scenarios considering the correlation of multi-dimensional variables is obtained. The probability density functions are all functions that follow their respective normal distributions, and the relevant formulas are as follows:

[0065] ；

[0066] In the formula, Let be the probability density function of wind power prediction error. For wind power prediction error, and These represent the mean and standard deviation of wind power forecasting errors, respectively. Let be the probability density function of photovoltaic prediction error. For photovoltaic prediction error, and These represent the mean and standard deviation of the photovoltaic prediction error, respectively. Let be the probability density function of the load uncertainty. For load uncertainty, and These are the mean and standard deviation of the demand-side load uncertainty, respectively.

[0067] It should be noted that the probability distribution function used in this application may be, but is not limited to, the vine structure Copula function. The specific construction method of the vine structure Copula function is as follows: (1) Calculate the correlation coefficient between the three random features (i.e., wind power prediction error, photovoltaic prediction error and load uncertainty) of all nodes in the first layer using Kendall rank correlation coefficient, and determine the selection of the node with the largest average correlation coefficient. The root node is selected from the nodes with the highest correlation coefficients. The tree structure is a D-vine structure, meaning it starts with the edge with the highest correlation coefficient and connects sequentially to create a sequential tree structure. The calculation formula is: In the formula, The correlation coefficient between two random features. The number of consecutive logarithms between two random features. The number of out-of-order logarithms between two random features. The total number of random features. It is a feature and Kendall rank correlation coefficient between them The closer the value is to 1, the stronger the correlation between random features. The closer the value is to 0, the weaker the correlation between random features. (2) For each edge of the first-level tree, the parameter value of the binary Copula function is estimated by maximizing the log-likelihood function, and the optimal binary Copula function is determined by using the Akaike information criteria (AIC). The calculation formula of the maximum likelihood estimation method is: In the formula, Let L be the maximum likelihood estimate, L be the maximum likelihood estimate of the Copula function, and k be the number of parameters in the model. The smaller the value, the higher the model fit. The smallest Copula function is taken as the optimal Copula function. (3) The Copula function obtained in step (2) is used to sample and obtain the node values ​​corresponding to the second-level tree. Repeat steps 2 and 3 to determine the second-level tree structure and the Copula function. The specific formula for sampling is as follows: In the formula, For the corresponding node value, It is the conditional inverse function of the Copula function. It is a random number drawn from the unit interval (0,1). (4) Repeat steps (1), (2) and (3) until all tree structures and Copula functions under the vine structure are obtained, and the vine Copula function model is obtained. (5) Calculate the CVM distance of each Copula function to evaluate the fitting effect of each model and determine the optimal vine structure Copula model. The relevant formula for calculating the CVM distance is: In the formula, For the constructed Copula function, The empirical Copula function is used, and the CvM distance is the distance between the target Copula function and the empirical Copula function. The smaller the distance, the higher the model's fitting accuracy.

[0068] It should be noted that the vine structure Copula function is a high-dimensional probability distribution modeling method that can well describe the correlation between variables.

[0069] It should be noted that CVM refers to the Cramér-Von Mises statistic, which is a statistical method used to test whether data follows a specific distribution (such as a normal distribution) or to evaluate the goodness of fit of a model.

[0070] Secondly, the source load data corresponding to the random feature is obtained through the probability distribution function; specifically, random numbers with values ​​in the range [0,1] that are independently and uniformly distributed are generated based on the Monte Carlo method. A vector composed of random numbers Based on We can consider the source load data with correlations and repeat them. Next, we can obtain The source payload data corresponding to the random features described in the group The formula for the probability distribution function is as follows:

[0071] ;

[0072] In the formula, Let be the joint probability distribution function. For wind power prediction error, For photovoltaic prediction error, For load uncertainty, , These are the distribution functions corresponding to the respective random features. For the Copula function corresponding to the random feature, It is a random number. It is the symbol for a partial derivative in mathematics.

[0073] Finally, by clustering the source-load data, the resource requirements corresponding to each power business scenario are obtained; specifically: based on the source-load data, the initial resource requirements corresponding to each power business scenario are determined; and a similarity matrix is ​​established based on the similarity between any two initial resource requirements. And construct the corresponding diagonal matrix based on the similarity matrix. Specifically, for the target area A partition subset Define the corresponding resource requirement membership matrices as follows: , No. Elements of the matrices Used to describe the resource requirements of each partitioned subset. Does it belong to a partition? : To obtain the maximum subset partition similarity, a method must be chosen to assign two nodes to the same subset. The proportion of two nodes assigned to the same subset across all subset partitions is defined as the node similarity matrix of that node pair. The similarity matrix can be obtained through... Solve for the equation, where... , is a concatenated matrix of the membership matrices of all nodes in the target region; based on the diagonal matrix and the similarity matrix , that is The Laplace matrix is ​​obtained. Extract the Laplacian matrix. The feature matrix is ​​composed of the feature values ​​in the matrix and the feature vectors corresponding to the feature values. ; By analyzing the feature matrix K-means clustering is performed to divide each resource demand into its corresponding partition, thereby obtaining the resource demand set corresponding to each power business scenario. Wherein, the resource demand set corresponding to each power business scenario is obtained. The calculation formula is as follows:

[0074] ;

[0075] In the formula, For the set of resource requirements, , To determine the similarity between subsets of resource demand, , The first , A subset of resource requirements, Let be the probability density function. It is the set of resource requirements. The corresponding subset, yes The joint probability distribution, It is the symbol for a partial derivative in mathematics.

[0076] By obtaining the random features corresponding to each power business scenario, the corresponding probability distribution function is determined, and then the corresponding resource requirements are accurately determined through clustering. This allows for accurate determination of the resource requirements for each power business scenario, facilitating subsequent resource scheduling based on these requirements.

[0077] Step S102: Using the resource utilization rate as the spring length, determine the elasticity coefficient based on the business flexibility value and the decision risk value, and construct the spring model corresponding to each power business scenario. The business flexibility value is the flexibility value of the edge computing node in resource adjustment, and the decision risk value is the risk value caused by the uncertainty of the power business scenario.

[0078] Understandably, after obtaining resource requirements, it's necessary to first calculate the resource utilization rate for the power business scenario. This utilization rate is then used as the spring length, along with subsequent elasticity coefficients, to construct a spring model corresponding to each power business scenario for resource scheduling. Specifically, after the power business enters the edge computing node processing queue, the controller can adjust its resource usage. While ensuring the safe operation of the power business, it participates in overall resource utilization optimization. The adjustment area for edge computing resources used by the power business scenario is as follows: Figure 2 As shown in the figure: For power business The start time when the device enters the processing queue of the edge computing node; For power business Total resource requirements ; For power business The fastest completion time; For power business To ensure the latest possible start time for fulfilling the requirements; For power business The cutoff operating time, the blue area abcda represents the time for power business. The maximum edge computing resource adjustment region is defined by the broken line abc, which represents the upper boundary of the maximum adjustment region and the fastest adjustment process of edge computing resources; the broken line adc represents the lower boundary of the maximum adjustment region and the slowest adjustment process of edge computing resources. The corresponding mathematical expression is as follows: In the formula, For power business Adjusting the response rate of edge computing resources; For power business The edge computing resource capacity. Using the above method, the resource adjustment area for all power services can be obtained, and dynamic resource adjustment of all individuals within the edge computing node can be achieved, physically similar to an equivalent machine. To ensure that resource utilization is as average as possible across time periods, the objective function for resource adjustment is defined as minimizing the variance of resource utilization across time periods, thereby determining the resource utilization rate, where the resource utilization rate... The calculation formula is:

[0079] ;

[0080] In the formula, For power business scenarios In time The allocated resources meet the resource adjustment area constraints; This represents the average resource utilization rate. The number of power business scenarios; The total time period for resource adjustments; For edge computing nodes in time Resource utilization rate; For power business Total resource requirements; and These are the minimum and maximum resource usage values ​​of power service i at time t, respectively. For power business scenarios The resource allocation weights are such that businesses with lower flexibility are more likely to be allocated resources. This represents a value for business flexibility.

[0081] Understandably, it is also necessary to first evaluate the flexibility of edge computing nodes in resource adjustment to determine the elasticity coefficient of the spring model. The evaluation formula for the business flexibility value is as follows:

[0082] ;

[0083] In the formula, This is the value representing the business flexibility. ·) indicates rounding up; , These represent the weights of task value and execution time, respectively, satisfying... ; For the value of the task; The time value indicates how close one is to the deadline. , This represents the maximum static flexibility. Indicates task The deadline for adjustment, This refers to the current moment.

[0084] It should be noted that the task value can be... The static flexibility set when initializing a task means that the higher the task's value, the lower the flexibility, while the time value... Compared to the current running time This is related to the task deadline adjustment time. Similarly, the shorter the deadline, the lower the task flexibility. When near hour, The larger the value, the better the result calculated in the formula. The smaller the value, the lower the flexibility; when Exceed At this time, the task does not participate in resource adjustment response, ensuring that the current task can be executed and completed within the delay; furthermore, , These two static parameters determine the flexibility of the task; because Using static flexibility values, the changes in the flexibility values ​​in the formula are mainly due to... The decision, and the coefficient This will control the range of flexibility adjustments. Obviously, when... When the user sets static flexibility, the resulting flexibility is the user-defined static flexibility. As the number of tasks gradually increases, the flexibility of the task is gradually determined by the earliest critical moment of operation. , The two weight parameters need to be set according to the specific application.

[0085] After assessing the business flexibility of edge computing nodes, it is also necessary to use CvaR to evaluate the revenue and risk values ​​caused by the uncertainty of power business scenarios as a basis for resource adjustment strategies. Representation and decision vector Related functions, The random characteristics of the uncertainties in the demand for resources in the power business are represented by the following probability density function: The confidence level is Risk Represented as: ;in, This represents the VaR value corresponding to a confidence level of α. To express the expectation value, this invention uses a function to facilitate direct solution. right Indicate: In the formula, For calculation Auxiliary variables, For random characteristics The set; the formula contains integral operations, and the expected value can be obtained by discretizing the integral terms, i.e. In the formula, For the first Random features The probability of occurrence; Discrete random characteristics The total number of samples.

[0086] Once the business flexibility value and decision risk value are determined, the elasticity coefficient can be comprehensively determined based on these two values. The resource utilization rate can then be used as the spring length to construct a single-time-point spring model for each power business scenario. The resource utilization curve under normal operation of a power business is known. For a single time node, the original length of the spring represents the resource utilization rate of the current service. The processing time of the entire service can be viewed as a combination of multiple different springs. In the formula, Let be the original length of the spring at node t. Let be the spring constant at time t, and β be the decision-maker's risk preference. A larger β indicates a higher degree of risk aversion, while β=0 indicates risk neutrality. This is the business flexibility value. This represents the CVaR value, or risk value, corresponding to a confidence level of α. Since there are errors between the actual scenario probability distribution and the initial scenario probabilities obtained from historical data, this application also needs to construct a probabilistic scenario fuzzy set of power business resource demand based on the 1-norm and ∞-norm. To determine the probability of spring length parameters in various typical scenarios The error range is constrained. In the formula, For probabilistic scene fuzzy sets, Let be the probability of a typical scenario s; Let be the initial probability of scenario s; and They represent for Positive and negative offset states; and They represent for Both the positive and negative offsets are positive values; and These represent the maximum deviation values ​​of typical scenario probabilities under 1-norm and ∞-norm constraints, respectively.

[0087] It is understandable that resource competition or cooperation exists between power business scenarios. Therefore, a single-time-point spring model cannot obtain accurate resource scheduling values. A composite spring model can be used to accurately simulate the resource scheduling process in real-world power business scenarios. Different spring elastic coefficients can express the intensity of competition or cooperation. The composite spring can generate any required nonlinear elastic deformation characteristics, thus representing the dynamic adjustment of resource acquisition strategies by power businesses at different time points. The composite spring consists of multiple springs, each horizontally connected regardless of its own weight. One end is connected to a wall, and the other end moves along the spring direction towards the resource boundary under the action of external forces. This means that each power business scenario strives to acquire the maximum resources. Each power business scenario can decide whether to compete for resources with other power business scenarios based on its own flexibility. Under the combined force, the springs reach a state of equilibrium under the current edge computing node resource allocation. Then, each spring further adjusts its resource allocation based on the completion status of the business scenario until the power business scenario is completed.

[0088] Step S103: Based on the resource requirements and the spring model, determine the resource adjustment value corresponding to each power business scenario, and perform resource scheduling for each power business scenario based on the resource adjustment value.

[0089] It is understood that resource adjustment values ​​corresponding to each power business scenario can be determined based on the resource requirements and the mathematical model of the spring model, wherein the mathematical model of the spring model is shown in the following formula:

[0090] ;

[0091] In the formula, For time t, the first The deformation of each spring represents the resource adjustment value for power business scenario i; This represents the resource utilization rate of edge computing nodes; For time t, the first The original length of the spring represents the resource utilization rate of power business scenario i. For time t, the first The spring constant of a spring.

[0092] It is understandable that during scheduling, it is also necessary to control the time of resource scheduling in power business scenarios. Elastic potential energy can be used to characterize the time delay constraints of power business scenarios. Since a power business scenario needs to obtain certain resources within the time delay to ensure completion, it corresponds to a spring model. The elastic potential energy of a composite spring composed of multiple different springs is constant, initially set to 0. For example, if the elastic potential energy of the spring increases at a certain time node, the spring at subsequent time nodes needs to be compressed to reduce the corresponding elastic potential energy. The elastic potential energy of the composite spring in this application can be negative. When the spring stretches at a certain time node, it means that the resource utilization rate after adjustment has increased, and the elastic potential energy is positive. When the spring is compressed at a certain time node, it means that the resource utilization rate after adjustment has decreased, and the elastic potential energy is negative. When the accumulated elastic potential energy reaches a certain negative value, it means that it no longer participates in resource adjustment, and the spring performs potential energy compensation to ensure the completion of the business. To ensure that resource utilization is as even as possible across different time periods, the objective function for the composite spring deformation is defined as minimizing the variance of the total spring length across different time periods. The spring model includes the objective function and constraints. Based on the objective function and constraints, a spring model with optimal resource utilization is established. The calculation process for the objective function is as follows:

[0093] ;

[0094] In the formula, This is the original length of the composite spring; For the composite spring in time The deformable variables represent resource adjustment values ​​for multiple power business scenarios; The average spring length represents the average resource demand. As a symbolic variable, when the spring stretches... During compression ; This refers to the total adjustment time of the compound spring; Let be the spring constant at node t; For edge computing nodes in time The change in resource utilization rate; It is a constant; It is the minimum resource utilization rate of the edge computing node at time t; It represents the maximum resource utilization rate of the edge computing node at time t.

[0095] This application's embodiments determine the corresponding resource requirements based on the random characteristics of each power business scenario, which can accurately determine the resource requirements of each power business scenario, facilitating subsequent resource scheduling for each power business scenario based on resource requirements. By using the resource utilization rate as the spring length and determining the elasticity coefficient based on the business flexibility value and decision risk value, a spring model corresponding to each power business scenario can be constructed. This allows for the construction of spring models that adapt to different resource scheduling requirements under the source-load fluctuation scenario of the distribution network, providing support for the elastic resource management of edge computing nodes in the distribution network. Through the spring model, the resource adjustment value corresponding to each power business scenario can be accurately determined, and resource scheduling for each power business scenario can be performed based on the resource adjustment value, adapting to the resource requirements of different power businesses and improving the resource scheduling capability in complex environments.

[0096] Example 2

[0097] Please refer to Figure 3 , Figure 3 This is a schematic diagram of the structure of an embodiment of a resource scheduling system for edge computing of a power distribution network provided in this application, including: an acquisition module 100, a construction module 200 and a scheduling module 300;

[0098] The acquisition module 100 is used to obtain random features corresponding to each power business scenario based on the historical wind power data, historical photovoltaic data and historical load data obtained for each power business scenario, and to obtain the resource requirements corresponding to each power business scenario based on the random features. The random features include wind power prediction error, photovoltaic prediction error and load uncertainty.

[0099] The construction module 200 is used to construct a spring model corresponding to each power business scenario by using the resource utilization rate as the spring length, determining the elasticity coefficient based on the business flexibility value and the decision risk value, wherein the business flexibility value is the flexibility value of the edge computing node in resource adjustment, and the decision risk value is the risk value caused by the uncertainty of the power business scenario.

[0100] The scheduling module 300 is used to determine the resource adjustment value corresponding to each power business scenario based on the resource demand and the spring model, and to perform resource scheduling for each power business scenario based on the resource adjustment value.

[0101] The information interaction and execution process between the modules in the above-mentioned resource scheduling system for edge computing of distribution networks are based on the same concept as the embodiment of the resource scheduling method for edge computing of distribution networks in the first aspect of the present invention, and the technical effects achieved are basically the same. For details, please refer to the description in the first embodiment of the method of the present invention, and will not be repeated here.

[0102] The device embodiments described above are merely illustrative. The modules described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of the method in this embodiment, depending on actual needs.

[0103] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of this application. It should be understood that the above descriptions are merely specific embodiments of this application and are not intended to limit the scope of protection of this application.

[0104] In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application for those skilled in the art.

Claims

1. A resource scheduling method for edge computing in distribution networks, characterized in that, include: Based on the historical wind power data, historical photovoltaic data and historical load data corresponding to each power business scenario, the random characteristics corresponding to each power business scenario are obtained, and the resource requirements corresponding to each power business scenario are obtained based on the random characteristics. The random characteristics include wind power prediction error, photovoltaic prediction error and load uncertainty. Using resource utilization rate as the spring length, and determining the elasticity coefficient based on business flexibility value and decision risk value, a spring model corresponding to each power business scenario is constructed. The business flexibility value is the flexibility value of edge computing nodes in resource adjustment, and the decision risk value is the risk value caused by the uncertainty of the power business scenario. Based on the resource requirements and the spring model, the resource adjustment value corresponding to each power business scenario is determined, and resource scheduling is performed on each power business scenario based on the resource adjustment value. The evaluation formula for the business flexibility value is as follows: ； In the formula, This represents the business flexibility value; ceil(·) indicates rounding up; , These represent the weights of task value and execution time, respectively, satisfying... ; For the value of the task; For time value, of which, , This represents the maximum static flexibility. Indicates task The deadline for adjustment, The current moment; Among them, the resource utilization rate The calculation formula is: ； In the formula, For power business scenarios In time The amount of resources allocated; This represents the average resource utilization rate. The number of power business scenarios; The total time period for resource adjustments; For edge computing nodes in time Resource utilization rate; For power business Total resource requirements; and These are the minimum and maximum resource usage values ​​of power service i at time t, respectively. For power business scenarios Resource allocation weights; This is the business flexibility value.

2. The resource scheduling method for edge computing of distribution networks according to claim 1, characterized in that, The process of obtaining the resource requirements corresponding to each power business scenario based on the random features is as follows: Construct a corresponding probability distribution function based on the aforementioned random features; The source payload data corresponding to the random feature is obtained through the probability distribution function, wherein the probability distribution function is a Copula function; By clustering the source-load data, the resource requirements corresponding to each power business scenario are obtained.

3. The resource scheduling method for edge computing of distribution networks according to claim 2, characterized in that, The formula for the probability distribution function is as follows: ； In the formula, Let be the joint probability distribution function. For wind power prediction error, For photovoltaic prediction error, For load uncertainty, , These are the distribution functions corresponding to the respective random features. For the Copula function corresponding to the random feature, It is a random number. It is the symbol for a partial derivative in mathematics.

4. The resource scheduling method for edge computing of distribution networks according to claim 2, characterized in that, By clustering the source-load data, the resource requirements corresponding to each power business scenario are obtained, specifically: Based on the source load data, the initial resource requirements corresponding to each power business scenario are determined; A similarity matrix is ​​established based on the similarity between any two initial resource requirements, and a corresponding diagonal matrix is ​​constructed based on the similarity matrix. Based on the diagonal matrix and the similarity matrix, the Laplace matrix is ​​obtained; Extract the eigenvalues ​​from the Laplacian matrix, and form a feature matrix based on the eigenvectors corresponding to the eigenvalues; By performing k-means clustering on the feature matrix, the resource requirements corresponding to each power business scenario are obtained.

5. The resource scheduling method for edge computing of distribution networks according to claim 4, characterized in that, The calculation formula for the resource requirements corresponding to each power business scenario is as follows: ； ； In the formula, For the set of resource requirements, , To determine the similarity between subsets of resource demand, , The first , A subset of resource requirements, Let be the probability density function. It is the set of resource requirements. The corresponding subset, yes The joint probability distribution, It is the symbol for a partial derivative in mathematics.

6. The resource scheduling method for edge computing of distribution networks according to claim 1, characterized in that, The determination of resource adjustment values ​​for each power business scenario based on the resource requirements and the spring model is as follows: ； In the formula, For time t, the first The deformation of each spring represents the resource adjustment value for power business scenario i; This represents the resource utilization rate of edge computing nodes; For time t, the first The original length of the spring represents the resource utilization rate of power business scenario i. For time t, the first The spring constant of a spring.

7. The resource scheduling method for edge computing of distribution networks according to claim 1, characterized in that, The spring model includes an objective function and constraints. Based on the objective function and constraints, a spring model with optimal resource utilization is established, specifically as follows: The objective function is expressed as: ； In the formula, This is the original length of the composite spring; For the composite spring in time The deformable variables represent resource adjustment values ​​for multiple power business scenarios; The average spring length represents the average resource requirement. As a symbolic variable, when the spring stretches... During compression ; This refers to the total adjustment time of the compound spring; Let be the spring constant at node t; For edge computing nodes in time The change in resource utilization rate; It is a constant; It is the minimum resource utilization rate of the edge computing node at time t; It represents the maximum resource utilization rate of the edge computing node at time t.

8. A resource scheduling system for edge computing in power distribution networks, characterized in that, include: Acquisition module, construction module, and scheduling module; The acquisition module is used to obtain random features corresponding to each power business scenario based on the historical wind power data, historical photovoltaic data and historical load data obtained for each power business scenario, and to obtain the resource requirements corresponding to each power business scenario based on the random features. The random features include wind power prediction error, photovoltaic prediction error and load uncertainty. The construction module is used to construct a spring model corresponding to each power business scenario by using resource utilization rate as the spring length, determining the elasticity coefficient based on business flexibility value and decision risk value, wherein the business flexibility value is the flexibility value of edge computing nodes in resource adjustment, and the decision risk value is the risk value caused by uncertainty in the power business scenario; The scheduling module is used to determine the resource adjustment value corresponding to each power business scenario based on the resource demand and the spring model, and to perform resource scheduling for each power business scenario based on the resource adjustment value. The evaluation formula for the business flexibility value is as follows: ； In the formula, This is the value representing the business flexibility. ·) indicates rounding up; , These represent the weights of task value and execution time, respectively, satisfying... ; For the value of the task; For time value, of which, , This represents the maximum static flexibility. Indicates task The deadline for adjustment, The current moment; Among them, the resource utilization rate The calculation formula is: ； In the formula, For power business scenarios In time The amount of resources allocated; This represents the average resource utilization rate. The number of power business scenarios; The total time period for resource adjustments; For edge computing nodes in time Resource utilization rate; For power business Total resource requirements; and These are the minimum and maximum resource usage values ​​of power service i at time t, respectively. For power business scenarios Resource allocation weights; This is the business flexibility value.

Images