A user-side energy storage participating in auxiliary service method and system

Abstract

The application discloses a user-side energy storage participating auxiliary service method and belongs to the technical field of power system dispatching control. The method comprises the following steps: performing load prediction on a prediction day by adopting a pre-constructed load prediction clustering method to obtain a load prediction value of each adjustable resource; adopting a pre-constructed user-side energy storage physical-behavior model to calculate the adjustment cost of each adjustable resource on the prediction day, and then obtaining the willingness of the user to participate in power grid aggregation; and according to the adjustment cost of each adjustable resource, the willingness of the user to participate in power grid aggregation, the load prediction value of each adjustable resource and pre-acquired auxiliary service transaction announcement information, the adjustable resource participating in peak shaving auxiliary service is declared in quantity and price with the target of maximizing the user benefits. The method can improve the power grid asset utilization rate, delay or reduce the power grid investment, and can also create additional economic value for the user-side energy storage.

Description

TECHNICAL FIELD

[0001] The application provides a user-side energy storage participation auxiliary service method and system, and belongs to the technical field of power system dispatching control. BACKGROUND

[0002] Under the application scenarios of enhancing the power supply guarantee capacity of weak areas of the power grid, delaying and replacing the investment in power transmission and transformation facilities, improving the system emergency guarantee capacity, supporting the construction of distributed energy supply systems, providing customized energy use services, and improving the flexible adjustment capacity of users, the distributed electrochemical energy storage on the distribution network side and the user side has developed rapidly. How to evaluate the potential of user-side energy storage participating in the power grid and how user-side energy storage participates in the market have become problems that need to be solved.

[0003] There are still some deficiencies in the evaluation of the sharing potential of user-side energy storage at present, which are embodied in the following aspects: (1) The existing technology is mostly qualitative analysis of user adjustable capacity, and there is less quantitative evaluation research on the adjustable capacity of user load under different power consumption modes. Moreover, the adjustable capacity of user load is affected by multiple factors such as self, society and environment, and how to improve the accuracy of quantitative evaluation of user load adjustable capacity is also a problem that needs to be further studied.

[0004] (2) When the existing technology evaluates the sharing potential of user-side energy storage, it is mostly based on historical big data of user-side energy storage, but does not consider the influence of the willingness of user-side energy storage itself to participate in the power grid and the market state on the evaluation benefit.

[0005] Such deficiencies make the evaluation of the sharing potential of user-side energy storage not accurate enough, and thus the utilization rate of power grid assets is low. SUMMARY

[0006] To solve the above technical problems, the application provides a user-side energy storage participation auxiliary service method and system, which can improve the utilization rate of power grid assets, delay or reduce the investment in the power grid, and create additional economic value for user-side energy storage.

[0007] To solve the above technical problems, the application adopts the following technical solutions:

[0008] In a first aspect, the application provides a user-side energy storage participation auxiliary service method, which comprises:

[0009] A load prediction clustering method is used to predict the load of the prediction day to obtain the load prediction value of each adjustable resource;

[0010] A user-side energy storage physical-behavior model is used to calculate the adjustment cost of each adjustable resource on the prediction day, and then the willingness of the user to participate in the power grid aggregation is obtained;

[0011] According to the adjustment cost of each adjustable resource, the willingness of user participation in power grid aggregation, the load prediction value of each adjustable resource and the pre-acquired auxiliary service transaction announcement information, a quantity and price declaration of the adjustable resource participating in peak shaving auxiliary service is declared with the user benefit maximization as the target.

[0012] As a further improvement of the application, the construction process of the pre-constructed load prediction clustering method comprises:

[0013] Based on the user side historical power load data, the user side adjustable resources are clustered by using a clustering algorithm to obtain a load prediction clustering method based on the grouping load power consumption law.

[0014] As a further improvement of the application, the clustering of the user side adjustable resources based on the user side historical power load data by using the clustering algorithm comprises:

[0015] The user side historical power load data is analyzed according to seasons;

[0016] Based on the seasonal analysis result, the load data of each day throughout the year is collected and recorded to obtain the user side historical power load data;

[0017] The abnormal data and bad data of the user side historical power load data are identified and deleted to obtain the historical power load data of all the adjustable resources of the user side after correction;

[0018] Based on the historical power load data of all the adjustable resources of the user side after correction, a SOINN clustering algorithm is used to extract the typical curves and clustering centers of various types of adjustable resources, the clustering centers are taken as the load prediction values of the adjustable resources, and a load prediction clustering method based on the grouping load power consumption law is obtained.

[0019] As a further improvement of the application, the extraction of the typical curves and clustering centers of various types of adjustable resources based on the historical power load data of all the adjustable resources of the user side after correction by using the SOINN clustering algorithm comprises:

[0020] The network parameters and neuron set A={C1, C2} are initialized, and the initial weights and biases are initialized in a random initialization manner;

[0021] The actual load data is input, the first layer network of the SOINN generates neurons in an online adaptive manner to represent the input data, and the calculation method of the neurons is:

[0022] S1=argmin||ξ-W T ||,T∈A (1)

[0023] Wherein, ξ represents a new input sample; W Tdenote the weights of neurons; A denotes a set of neurons; T denotes any neuron in the set of neurons;

[0024] The method for winning neurons is:

[0025] S2=argmin||ξ-W T ||,T∈A\{S1} (2)

[0026] Wherein, S1 denotes a new neuron value;

[0027] The output of the first layer neural network is calculated using the current network parameters; the second layer network calculates the distance between neurons according to the results of the first layer network, and uses the distance between neurons as a parameter to run the SOINN algorithm again with the neurons of the first layer as input, thereby performing dynamic adjustment;

[0028] If there is new input data, repeat the loop until the stopping condition is reached, and finally output the winning neuron set as the load prediction value of the adjustable resource.

[0029] As a further improvement of the application, the user-side energy storage physical-behavior model comprises: the charge and discharge power of the battery energy storage system side, the upper and lower boundaries of the rated power of the energy storage system, the energy state of the battery energy storage system, the income obtained by the user-side energy storage for self-use, the adjustment cost of the adjustable resource of the user, and the willingness of the user to participate in the grid aggregation.

[0030] As a further improvement of the application, the adjustment cost of each adjustable resource on the predicted day is calculated, and then the willingness of the user to participate in the grid aggregation is obtained, comprising:

[0031] According to the charge and discharge power of the battery energy storage system side, the upper and lower boundaries of the rated power of the energy storage system, and the energy state of the battery energy storage system, the income obtained by the user-side energy storage for self-use is calculated;

[0032] The adjustment cost of the adjustable resource is calculated;

[0033] According to the income obtained by the user-side energy storage for self-use and the adjustment cost of the adjustable resource, the willingness of the user to participate in the grid aggregation is obtained.

[0034] As a further improvement of the application, the process of obtaining the pre-obtained auxiliary service transaction announcement information comprises:

[0035] Obtain auxiliary service transaction announcement information, send information to the user to which the adjustable resource belongs, and obtain the adjustable resource power information and adjustable information fed back by the user to which the adjustable resource belongs;

[0036] As a further improvement of the application, the price of the adjustable resource participating in the peak shaving auxiliary service is declared with the maximum user benefit as the target, and the specific pricing method is:

[0037] P apply =ηP allow_max

[0038] Wherein, P apply is the adjustable resource bidding price of a single user, P allow max is the maximum bidding price specified by the transaction rules, and η is an adjustment coefficient.

[0039] As a further improvement of the application, the bidding of the quantity and price of the adjustable resource participating in the peak shaving auxiliary service with the maximum user benefit as the target further comprises the following steps:

[0040] According to the reported idle energy storage resource information, the adjustable resource bidding price and the actual operation day power grid demand, the actual transaction volume is obtained by matching;

[0041] After the end of the actual operation day, the settlement result of the user side energy storage charging and discharging fee is obtained according to the actual transaction volume.

[0042] In a second aspect, the application provides a user side energy storage participating in auxiliary service device, comprising:

[0043] A load prediction module is used to perform load prediction on the prediction day by using a pre-constructed load prediction clustering method, and obtain the load prediction value of each adjustable resource;

[0044] A cost willingness calculation module is used to calculate the adjustment cost of each adjustable resource of the prediction day by using a pre-constructed user side energy storage physical-behavior model, and further obtain the willingness of the user to participate in the power grid aggregation;

[0045] An auxiliary service participating module is used to bid the quantity and price of the adjustable resource participating in the peak shaving auxiliary service with the maximum user benefit as the target according to the adjustment cost of each adjustable resource, the willingness of the user to participate in the power grid aggregation, the load prediction value of each adjustable resource and the pre-acquired auxiliary service transaction announcement information.

[0046] Optionally, in the load prediction module, the construction process of the pre-constructed load prediction clustering method comprises:

[0047] Based on the user side historical power load data, the user side adjustable resource is clustered by using a clustering algorithm, and a load prediction clustering method based on the grouping load power consumption law is obtained.

[0048] Optionally, in the load prediction module, the clustering of the user side adjustable resource based on the user side historical power load data by using the clustering algorithm comprises:

[0049] The user side historical power load data is analyzed according to seasons;

[0050] Based on the seasonal analysis results, all-weather load data of each day in a year is collected, and user-side historical power consumption load data is recorded;

[0051] Abnormal data and bad data of the user-side historical power consumption load data are identified and deleted, and historical power consumption load data of all adjustable resources of the user-side after correction is obtained;

[0052] Based on the historical power consumption load data of all adjustable resources of the user-side after correction, SOINN clustering algorithm is used to extract typical curves and clustering centers of each type of adjustable resource, and the clustering center is taken as the load prediction value of the adjustable resource, so as to obtain a load prediction clustering method based on grouping load power consumption law.

[0053] Optionally, based on the historical power consumption load data of all adjustable resources of the user-side after correction, SOINN clustering algorithm is used to extract typical curves and clustering centers of each type of adjustable resource, and the clustering center is taken as the load prediction value of the adjustable resource, including:

[0054] Initialize network parameters and neuron set A={C1,C2}, and initialize initial weights and biases in a random initialization manner;

[0055] Input actual load data, and the first layer network of SOINN generates neurons in an online adaptive manner to represent input data, and the calculation method of the neurons is:

[0056] S1=argmin|||ξ-W T ||,T∈A (1)

[0057] Wherein, ξ represents a new input sample; W T represents the weight of the neuron; A represents the neuron set; T represents any neuron in the neuron set;

[0058] The method of the winning neuron is:

[0059] S2=argmin|||ξ-W T ||,T∈A\{S1} (2)

[0060] Wherein, S1 represents a new neuron value;

[0061] The output of the first layer neural network is calculated using the current network parameters; the second layer network calculates the distance between neurons according to the result of the first layer network, and takes the distance between neurons as a parameter, and takes the neurons of the first layer as input to run the SOINN algorithm again, so as to perform dynamic adjustment;

[0062] If there is new input data, repeat the loop until the stopping condition is reached, and finally output the winning neuron set as the load prediction value of the adjustable resource.

[0063] Optionally, the user-side energy storage physical-behavioral model includes: the charging and discharging power of the battery energy storage system, the upper and lower bounds of the rated power of the energy storage system, the energy state of the battery energy storage system, the revenue obtained from user-side energy storage for self-use, the adjustment cost of the user's adjustable resources, and the user's willingness to participate in grid aggregation.

[0064] Optionally, the adjustment cost of each adjustable resource on the forecast date is calculated, thereby obtaining the user's willingness to participate in grid aggregation, including:

[0065] The revenue obtained by users from self-use of energy storage is calculated based on the charging and discharging power of the battery energy storage system, the upper and lower limits of the rated power of the energy storage system, and the energy state of the battery energy storage system.

[0066] Calculate the adjustment costs of adjustable resources;

[0067] The willingness of users to participate in grid aggregation is determined by the revenue obtained from user-side energy storage for self-use and the adjustment costs of adjustable resources.

[0068] Optionally, the process of obtaining the pre-acquired ancillary service transaction announcement information includes:

[0069] Obtain ancillary service transaction announcement information, send information to users of adjustable resources, and obtain adjustable resource electricity consumption information and adjustable information fed back by users of adjustable resources;

[0070] Optionally, the bidding process for the quantity and price of adjustable resources participating in peak-shaving auxiliary services, with the goal of maximizing user benefits, is as follows:

[0071] P apply =ηP allow_max

[0072] Among them, P apply For a single user, the adjustable resource bid price, P allow max η is the maximum bid price stipulated by the trading rules, and η is the adjustment coefficient.

[0073] Thirdly, the present invention provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the user-side energy storage participation ancillary service method.

[0074] Fourthly, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the user-side energy storage participation ancillary service method.

[0075] In a fifth aspect, the present application provides a computer program product comprising computer instructions, wherein the computer instructions instruct a computer to execute the user-side energy storage participating in auxiliary service method.

[0076] The present application has the following beneficial effects compared with the prior art:

[0077] In the present application, when establishing the user-side historical power consumption load data, the clustering algorithm is used to predict the user-side power consumption load data with large data volume and great uncertainty, so as to obtain the load prediction value; the accuracy of clustering can be improved, and the time for processing the user-side power consumption load data can be greatly reduced. The load prediction value of each adjustable resource of the prediction day can be more accurately generated, and the reasonable matching and auxiliary pricing mechanism of the energy storage can be provided with the help of the adjustment cost of each adjustable resource, the willingness of the user to participate in the power grid aggregation, the load prediction value of each adjustable resource and the pre-acquired auxiliary service transaction announcement information. The present application finally provides a certain degree of reference for the reasonable distribution of the user-side energy storage participating in the auxiliary service, reduces the transaction risk and improves the power grid utilization rate. BRIEF DESCRIPTION OF DRAWINGS

[0078] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following introduces the drawings of the related technical solutions in the embodiments of the present application or the prior art. It should be understood that the drawings in the following introduction are only for the convenience of clearly describing part of the embodiments in the technical solutions of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0079] Figure 1 A user-side energy storage participating in auxiliary service method flow chart is provided in the present application.

[0080] Figure 2 A flowchart of the embodiment of the present application is provided.

[0081] Figure 3 The user-side energy storage charging and discharging power; (a) is the electricity price, (b) is the power;

[0082] Figure 4 A SOINN algorithm flow chart is provided.

[0083] Figure 5 A user-side energy storage participating in auxiliary service device is provided in the present application.

[0084] Figure 6 An electronic device schematic diagram is provided in the present application. DETAILED DESCRIPTION

[0085] Embodiments of the present application are described below in detail with reference to the accompanying drawings, wherein the same or similar components or components having the same or similar functions are denoted by the same or similar reference numerals throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present application and cannot be understood as a limitation of the present application. For the step numbers in the following embodiments, they are only set for the convenience of explanation and description, and the order between the steps is not limited in any way, and the execution order of each step in the embodiments can be adaptively adjusted according to the understanding of those skilled in the art.

[0086] In the description of the present application, the words such as setting, installing, connecting and the like should be understood in a broad sense unless otherwise explicitly limited, and those skilled in the art can reasonably determine the specific meaning of the above words in the present application in combination with the specific content of the technical solution.

[0087] The applicant found that when obtaining user-side adjustable resource power consumption information and cost information, the traditional method is to let the user side report the energy storage capacity that can participate in grid regulation by itself, but this method is not accurate, causing the demand power of the grid and the reported power to be mismatched, and the demand of the grid cannot be quickly responded. Establishing a user-side energy storage physical-behavior model can analyze the market behavior of energy storage resource surplus users and understand the willingness of user-side energy storage to participate in grid regulation. More accurately, the demand power of the grid and the surplus power of the user side are matched.

[0088] When determining the load prediction value of the adjustable resource, the user-side power consumption load data has the characteristics of time sequence, large data volume, infinity, etc., and will change with the change of seasons. The traditional method is to predict the user-side energy storage load based on big data. However, the prediction accuracy is insufficient, and the real-time of the data and the influence of the season on the load are not considered. Moreover, if new load data arrives in the database, it will consume a lot of time and space to re-predict.

[0089] Therefore, the applicant believes that establishing a reasonable incremental learning neural network can greatly reduce the training data size and reduce the computational overhead of the training model, thereby constructing a more reasonable user-side power consumption load prediction.

[0090] As shown in Figure 1 The present application provides a user-side energy storage participating auxiliary service method, comprising the following steps:

[0091] S100, using a pre-constructed load prediction clustering method to perform load prediction on the prediction day to obtain a load prediction value of each adjustable resource;

[0092] It should be pointed out that this step finds the mode and rule of different load types by analyzing and clustering historical load data, so as to realize accurate prediction of the predicted day load. Load prediction is the basis of power system operation and dispatch, and accurate load prediction helps to arrange and adjust power grid resources in advance to meet the needs of users.

[0093] S200, the pre-constructed user side energy storage physical-behavior model is used to calculate the adjustment cost of each adjustable resource of the predicted day, and then the willingness of the user to participate in the power grid aggregation is obtained;

[0094] It should be pointed out that the pre-constructed user side energy storage physical-behavior model in this step comprehensively considers the physical characteristics (such as charging and discharging power, rated power upper and lower limit, energy state) of the battery energy storage system and the benefits obtained by the user side energy storage for self-use and the adjustment cost of the adjustable resource. By calculating the adjustment cost of each adjustable resource, the economic efficiency and feasibility of its participation in the power grid aggregation can be evaluated.

[0095] Most importantly, the willingness of the user to participate in the power grid aggregation is also considered, which helps to better reflect the actual needs and preferences of the user in the optimization process.

[0096] S300, according to the adjustment cost of each adjustable resource, the willingness of the user to participate in the power grid aggregation, the load prediction value of each adjustable resource and the pre-obtained auxiliary service transaction announcement information, the quantity and price declaration of the adjustable resource participating in the peak shaving auxiliary service is made with the target of maximizing the benefits of the user.

[0097] It should be pointed out that the quantity and price declaration strategy of the adjustable resource participating in the peak shaving auxiliary service in this step realizes the effective utilization and maximum benefit of the user side energy storage and other adjustable resources while meeting the demand of the power grid.

[0098] The user side energy storage participating auxiliary service method provided by the present application determines the load prediction value of each adjustable resource of the predicted day based on the clustering algorithm and the historical electricity load data of the user side; obtains the adjustment cost of each adjustable resource based on the user side energy storage physical-behavior model; and makes quantity and price declaration with the target of maximizing the benefits of the user according to the adjustment cost of each adjustable resource, the load prediction value of each adjustable resource and the pre-obtained auxiliary service transaction announcement information, which provides support for the reasonable allocation of energy storage resources. The present application improves the response speed and accuracy of the calculation of the adjustable resource capacity and provides support for the reasonable allocation of energy storage resources through the energy storage matching mechanism, which can effectively reduce the transaction risk and improve the utilization rate of the power grid.

[0099] As an optional solution, the construction process of the pre-constructed load prediction clustering method includes: based on the historical electricity load data of the user side, the user side adjustable resource is clustered by using the clustering algorithm, and a load prediction clustering method based on the grouping load electricity rule is obtained.

[0100] As an example, in this construction process, first, the historical electricity load data of the user side is collected. These data may include the electricity load amount in different time periods as the basis for subsequent clustering and prediction. Then, the user-side adjustable resources are clustered using a clustering algorithm. The clustering algorithm can divide data points with similar characteristics into the same category, so that the data points in the same category are as similar as possible, while the data points between different categories are quite different. These methods can automatically divide the user-side adjustable resources into several categories according to the distribution and characteristics of the data. Then, the load electricity law of each category is analyzed. By analyzing the data of each category, the load electricity law unique to the category can be extracted, such as the load peak period, the average load level, the load fluctuation, etc. These laws provide an important basis for subsequent load prediction. Finally, load prediction is performed based on the grouped load electricity law. According to the load electricity law of each category, the corresponding load prediction clustering method can be established.

[0101] These load prediction clustering methods can predict the load change trend in the future period of time according to the historical data and the current situation. Since the clustering algorithm has divided data points with similar characteristics into the same category, the load prediction method based on the grouped load electricity law can more accurately reflect the electricity consumption behavior of users in different categories, thereby improving the accuracy of prediction.

[0102] Among them, in the power load clustering, the commonly used clustering methods include distance-based clustering methods (such as K-means clustering, hierarchical clustering, etc.) and density-based clustering methods (such as DBSCAN clustering). The clustering algorithm of the present application can adopt the SOINN algorithm, of course, other clustering algorithms can also be selected, for example, KMeans clustering, K-nearest neighbor classification model (KNN), etc.

[0103] The core content of the present application is the processing method of data. The clustering algorithm such as SOINN algorithm can adaptively perform online clustering and topology on dynamic data without prior knowledge. For the data that has been learned, SOINN can discover new patterns in the data and learn without affecting the previous learning results.

[0104] Illustratively, the user-side historical electricity load data is used to cluster the user-side adjustable resources using a clustering algorithm, including the following steps:

[0105] The user-side historical electricity load data is analyzed according to the season;

[0106] Based on the seasonal analysis results, the load data of each day throughout the year is collected and recorded to obtain the user-side historical electricity load data;

[0107] Identify and delete abnormal data and bad data of user side historical power load data, and obtain the historical power load data of all adjustable resources of the user side after correction;

[0108] Based on the historical power load data of all adjustable resources of the user side after correction, the SOINN clustering algorithm is used to extract typical curves and clustering centers of various categories of adjustable resources, the clustering center is used as the load prediction value of the adjustable resource, and a load prediction clustering method based on grouping load power consumption law is obtained.

[0109] For example, when collecting user side historical load data, 3 seconds is usually used as a time interval to collect data, so the amount of data collected is very large, and some noise data will have a certain influence on obtaining user side historical load data later. SOINN adopts a double-layer model structure, which can remove noise data and improve the accuracy of clustering. And the user side historical power load data will change with the change of seasons, in order to obtain more accurate user side historical power load data prediction, the real-time performance of data processing has a higher requirement.

[0110] Therefore, the present application distinguishes seasons and uses SOINN algorithm, which can learn new data information on the basis of original data learning, greatly reduces the user side historical power load data prediction time, and improves the user side historical power load data prediction accuracy.

[0111] As an optional solution, the user side energy storage physical-behavior model includes: including the charge and discharge power of the battery energy storage system side, the upper and lower boundaries of the rated power of the energy storage system, the energy state of the battery energy storage system, the income obtained by the user side energy storage for self use, the adjustment cost of the user's adjustable resources, and the willingness of the user to participate in the aggregation of the power grid.

[0112] Among them, the adjustment cost of each adjustable resource of the prediction day is calculated, and then the willingness of the user to participate in the aggregation of the power grid is obtained, including:

[0113] According to the charge and discharge power of the battery energy storage system side, the upper and lower boundaries of the rated power of the energy storage system, and the energy state of the battery energy storage system, the income obtained by the user side energy storage for self use is calculated;Calculate the adjustment cost of the adjustable resource;According to the income obtained by the user side energy storage for self use and the adjustment cost of the adjustable resource, the willingness of the user to participate in the aggregation of the power grid is obtained.

[0114] Specifically, the model first calculates the income obtained by the user side energy storage for self use according to the charge and discharge power of the battery energy storage system, the upper and lower boundaries of the rated power and the energy state. These parameters reflect the operation efficiency and capacity of the energy storage system, which directly affect the energy cost saved by the user through the energy storage system and the additional income that can be obtained.

[0115] The adjustment cost of adjustable resources includes the cost of charging and discharging the energy storage system, and the cost of adjusting other adjustable resources according to the demand of the power grid. These cost factors reflect the economic pressure of users participating in power grid aggregation, and are important factors in determining the willingness of users to participate.

[0116] Finally, the willingness of users to participate in power grid aggregation is evaluated. If the self-use benefit of the energy storage is high and the adjustment cost is low, the willingness of users to participate in power grid aggregation will be enhanced. On the contrary, if the benefit is low and the cost is high, the willingness of users to participate will be reduced.

[0117] In practical applications, this model can help users better understand and utilize the advantages of energy storage systems, achieve energy cost reduction and maximize benefits. At the same time, it helps power grid operators more accurately evaluate user demand and response capacity, optimize power grid operation and dispatching strategies, and improve the stability and economy of the power system.

[0118] The application will be further described in detail below in combination with the drawings and specific embodiments.

[0119] Figure 2 The user side energy storage participates in the auxiliary service process diagram provided by the application; a user side energy storage participates in the auxiliary service method, including the following steps:

[0120] S1, constructing a user side energy storage physical-behavior model, including the charging and discharging power of the battery energy storage system side, the upper and lower boundaries of the rated power of the energy storage system, the energy state of the battery energy storage system, the self-use benefit obtained by the user side energy storage, the adjustment cost of the user's adjustable resources, and the willingness of the user to participate in the power grid aggregation;

[0121] S2, determining the load prediction value of each adjustable resource on the prediction day based on clustering algorithm and user side historical electricity load data;

[0122] S3, according to the adjustment cost of each adjustable resource, the load prediction value of each adjustable resource, and the pre-obtained auxiliary service transaction announcement information, the quantity and price are declared with the goal of maximizing user benefits;

[0123] Among them, Figure 3 is a user side energy storage charging and discharging power curve diagram, used to construct a user side energy storage physical-behavior model, including the following processes:

[0124] S1-1, the charging and discharging power expression of the battery energy storage system side is as follows:

[0125]

[0126] Among them, is the charging and discharging power demand of the battery energy storage system side at time t; is the grid-connected power output by the battery energy storage system at time t; γ ch is the charging efficiency of the battery energy storage system; γ disch is the discharging efficiency of the battery energy storage system.

[0127] S1-2, the upper and lower limit expressions of the rated power of the energy storage system are as follows:

[0128]

[0129] wherein, is the rated power of the battery energy storage system.

[0130] S1-3, the energy state expression of the battery energy storage system is as follows:

[0131]

[0132] SOC min ≤ SOC t ≤ SOC max

[0133] wherein, ΔT is the energy storage power instruction interval; E rate is the rated capacity of the energy storage system.

[0134] S1-4, the expression of the income obtained by the user-side energy storage for self-use is as follows:

[0135]

[0136] wherein, Ω i is the income obtained by the i-th user-side energy storage through the user self-use mode per unit time, is the cumulative discharging electric quantity of the i-th user-side energy storage system, is the cumulative charging electric quantity of the i-th user-side energy storage system, Pr1 is the electricity price when the energy storage is discharged, and Pr2 is the electricity price when the energy storage is charged.

[0137] S1-5, the adjustment cost of each adjustable resource

[0138]

[0139] wherein, C i is the adjustment cost of the i-th user's adjustable resource, i ∈ {m1,...,md}, β i is the adjustment cost weight of the i-th user's adjustable resource, α is the aggregation merchant subsidy ratio, C si is the cost of the user's investment in hardware equipment, C oi is the cost of the user's investment in software system, C yi is the annual operation and management cost of the system, T is the number of years of project implementation, C xioThe annual project management fee paid by the user to the aggregator in the oth year.

[0140] The willingness expression of the S1-6 user to participate in the grid aggregation is as follows:

[0141]

[0142] η usr The willingness of the user to participate in the grid aggregation.

[0143] It can be seen that the model can consider the physical characteristics of the energy storage system, and can also combine the economic benefits of the user and the demand of the grid to provide decision support for the optimization configuration of the user-side energy storage and the grid aggregation. In addition, the model can be flexibly adjusted according to different grid policies, electricity price mechanisms and user demands to adapt to different application scenarios and market demands.

[0144] In the S2, the load prediction value of each adjustable resource on the prediction day is determined based on the clustering algorithm and the historical user-side load data, and includes the following steps:

[0145] S21) The historical user-side load data is analyzed according to four different seasons of spring, summer, autumn and winter;

[0146] S22) The sampling points are collected all day long every day of the year, and the historical user-side load data of 96 points of each sampling point are recorded;

[0147] S23) The abnormal data and bad data of the historical user-side load data are identified and deleted;

[0148] S24) Based on the corrected historical load data of all adjustable resources of the user side, the SOINN clustering algorithm is used to extract the clustering centers of each category;

[0149] S25) Based on the power consumption law of the grouped load, the corresponding load prediction algorithm is determined, and the load of each adjustable resource on the prediction day is predicted to obtain the load prediction value of each adjustable resource on the prediction day.

[0150] In establishing user side historical power consumption load data, user side power consumption load data is different with the change of seasons, unlike previous research, the application makes a distinction between the four seasons, which can more accurately reflect the user side historical power consumption load. Unlike the widely used supervised learning clustering algorithms such as K-means algorithm, DBSCAN algorithm, SOINN algorithm can adaptively cluster and topology online dynamic data without prior knowledge. For the user side power consumption load data with large data volume and large uncertainty, the accuracy of clustering can be improved, and the time for processing user side power consumption load data can be greatly reduced.

[0151] In the step S2, the SOINN clustering algorithm is used to extract the typical curve and clustering center of each category of adjustable resources based on the corrected historical power consumption load data of all user side adjustable resources, and the clustering center is taken as the load prediction value of the adjustable resources, including the following Figure 4 The process is shown as follows:

[0152] S2-1, the network parameters and neuron set A={C1, C2} are initialized, and the initial weight and bias are initialized in a random initialization manner.

[0153] S2-2, the actual load data is input, and the first layer network of SOINN adaptively generates neurons to represent the input data in an online manner; the calculation formula of the neurons is as formula

[0154] S1=argmin||ξ-W T ||,T∈A (1)

[0155] In formula (1), ξ represents a new input sample; W T represents the weight of the neuron; A represents the neuron set; T represents any neuron in the neuron set. The winning neuron can be calculated by formula (2).

[0156] S2=argmin|||ξ-W- T ||,T∈A\{S1} (2)

[0157] Wherein, S1 represents the new neuron value.

[0158] S2-3, the output of the first layer neural network is calculated using the current network parameters;

[0159] S2-4, the second layer network calculates the distance between neurons according to the result of the first layer network, and takes the first layer neurons as input to run the SOINN algorithm again, so as to dynamically adjust, eliminate noise, and further stabilize the learning result.

[0160] If new input data is received, repeat steps S2-2 to S2-5 until the stopping condition is met, and finally output the set of winning neurons.

[0161] In S3, an energy storage matching and auxiliary pricing mechanism is established, specifically including:

[0162] 1) Aggregators send information to users of adjustable resources based on the transaction announcement information for peak shaving ancillary services, and users report the electricity consumption information and adjustable information of adjustable resources;

[0163] 2) Aggregate and comprehensively report information, and cluster adjustable resources based on load forecasts, user participation intentions and cost analysis of adjustable resources in S1 and S2.

[0164] 3) Based on the clustering results and considering the maximization of overall user benefits, submit bids for the quantity and price of adjustable resources for ancillary services. The specific bidding method is as follows:

[0165] P apply =ηP allow_max

[0166] Among them, P apply For a single user, the adjustable resource declaration price, P allow max η is the maximum bid price stipulated by the trading rules, and η is the adjustment coefficient.

[0167] 4) Match the reported idle energy storage resources, the declared price of adjustable resources with the actual electricity demand of the power grid on the actual operating day, and publish the actual transaction volume.

[0168] 5) After the actual operation day ends, the user-side energy storage receives the settlement of charging and discharging fees.

[0169] This invention addresses several issues in the current market, including inaccurate quantitative assessments of user-side energy storage participation in ancillary services, unclear user willingness to participate, and low market efficiency. It establishes a physical-behavioral model for user-side energy storage, which can effectively analyze user willingness to participate and provide maximum support for the rational allocation of surplus energy storage resources. Secondly, historical load data for adjustable user-side resources is characterized by seasonality, large data volume, and high volatility. The SOINN clustering algorithm can more accurately generate load forecasts for each adjustable resource on the forecast day, providing assistance for the rational matching of energy storage and ancillary pricing mechanisms. Ultimately, this invention provides a certain degree of reference for the rational allocation of user-side energy storage participation in ancillary services, reducing transaction risks, and improving grid utilization.

[0170] like Figure 5 As shown, a second objective of this invention is to provide a user-side energy storage participation ancillary service system, comprising:

[0171] a load prediction module configured to perform load prediction on the prediction day using a pre-constructed load prediction clustering method to obtain a load prediction value of each adjustable resource;

[0172] a cost-willingness calculation module configured to calculate the adjustment cost of each adjustable resource on the prediction day using a pre-constructed user-side energy storage physical-behavior model, and further obtain the willingness of the user to participate in grid aggregation;

[0173] an auxiliary service participation module configured to report the quantity and price of the adjustable resource participating in the peak shaving auxiliary service based on the adjustment cost of each adjustable resource, the willingness of the user to participate in grid aggregation, the load prediction value of each adjustable resource, and pre-acquired auxiliary service transaction announcement information, with the goal of maximizing the benefits of the user.

[0174] The system is based on the above-mentioned user-side energy storage participation auxiliary service method. Through accurate load prediction and optimized management of adjustable resources, the present application can better match the demand of the power grid and the supply capacity of the user-side energy storage, thereby improving the utilization efficiency of resources. By calculating the adjustment cost of the adjustable resource and considering the willingness of the user to participate in grid aggregation, the operating cost of the user can be reduced while meeting the demand of the power grid. The participation of the user-side energy storage helps to alleviate the peak-valley difference of the power grid and reduce the peak shaving pressure of the power grid, thereby improving the stability and reliability of the power grid. The method helps to better integrate distributed energy resources, promote the consumption and utilization of new energy, and promote the clean and low-carbon of the power system.

[0175] As shown in Figure 6 the third object of the present application is to provide an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the user-side energy storage participation auxiliary service method when executing the computer program.

[0176] The fourth object of the present application is to provide a computer-readable storage medium storing a computer program, wherein the computer program is executed by a processor to implement the user-side energy storage participation auxiliary service method.

[0177] The fifth object of the present application is to provide a computer program product comprising computer instructions, characterized in that the computer instructions instruct a computer to execute the user-side energy storage participation auxiliary service method.

[0178] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific way, so that the instructions stored in the computer-readable memory produce a product comprising instruction devices, which implement the flowFigure 1 one or more processes and / or functions described in the one or more blocks. Figure 1 one or more blocks or any combination thereof.

[0179] These computer program instructions can also be loaded into a computer or other programmable data processing devices, so that a series of operation steps are performed on the computer or other programmable devices to generate computer-implemented processes, thus the instructions executed on the computer or other programmable devices provide a process for implementing the functions described in the flow Figure 1 one or more processes and / or functions described in the one or more blocks. Figure 1 one or more blocks or any combination thereof.

[0180] The present application can take the form of a completely hardware embodiment, a completely software embodiment or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage etc.) containing computer usable program code.

[0181] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of the flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing devices to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a means for implementing the functions described in the flow Figure 1 one or more processes and / or functions described in the one or more blocks. Figure 1 one or more blocks or any combination thereof.

[0182] Obviously, the described embodiments are only some embodiments but not all embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work should fall within the protection scope of the present application.

[0183] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application but not to limit it, although the present application has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: the specific embodiments of the present application can be modified or replaced by the same, without departing from the spirit and scope of the present application, any modification or equivalent replacement, which should be covered within the protection scope of the claims of the present application.

Claims

1. A method for user-side energy storage to participate in ancillary services, characterized in that, include: A pre-built load forecasting clustering method is used to forecast the load for the forecast date, and the load forecast value for each adjustable resource is obtained. By using a pre-built user-side energy storage physical-behavior model, the adjustment cost of each adjustable resource on the predicted day is calculated, thereby obtaining the user's willingness to participate in grid aggregation; Based on the adjustment cost of each adjustable resource, the user's willingness to participate in grid aggregation, the load forecast value of each adjustable resource, and the pre-obtained ancillary service transaction announcement information, the user's interests are maximized when the adjustable resources participate in peak shaving ancillary services and the quantity and price are declared. The construction process of the pre-built load prediction clustering method includes: Based on historical electricity load data on the user side, a clustering algorithm is used to cluster adjustable resources on the user side, resulting in a load prediction clustering method based on the grouped load electricity consumption pattern. The method of clustering adjustable resources on the user side based on historical electricity load data includes: Historical electricity load data on the user side is analyzed according to season; Based on the seasonal analysis results, load data for each day of the year is collected and recorded to obtain historical electricity load data on the user side. Abnormal and undesirable data in the historical electricity load data of the user side are identified and deleted to obtain corrected historical electricity load data of all adjustable resources on the user side; Based on the historical electricity load data of all adjustable resources on the user side after correction, the SOINN clustering algorithm is used to extract the typical curves and cluster centers of each category of adjustable resources. The cluster centers are used as the load prediction values ​​of adjustable resources, thus obtaining a load prediction clustering method based on the grouped load electricity consumption pattern. Based on the historical electricity load data of all adjustable resources on the user side after correction, the SOINN clustering algorithm is used to extract typical curves and cluster centers for each category of adjustable resources. The cluster centers are used as the load prediction values ​​for adjustable resources, including: Initialize the network parameters and neuron set A = {C1,C2} by randomly initializing the initial weights and biases. Given actual load data, the first layer of the SOINN network adaptively generates neurons online to represent the input data. The neuron computation method is as follows: (1) in, This represents a new input sample; The weights of neurons are represented by A; the set of neurons is represented by A; and any neuron in the set of neurons is represented by T. The method for winning neurons is as follows: (2) in, This represents the new neuron value; The output of the first layer of the neural network is calculated using the current network parameters. The second layer calculates the distance between neurons based on the result of the first layer and uses this distance as a parameter to run the SOINN algorithm again with the neurons of the first layer as input, thereby making dynamic adjustments. If new input data is received, the loop repeats until the stopping condition is met, and finally the set of winning neurons is output as the load prediction value of the adjustable resources.

2. The method for user-side energy storage to participate in ancillary services according to claim 1, characterized in that, The user-side energy storage physical-behavioral model includes: the charging and discharging power of the battery energy storage system, the upper and lower bounds of the rated power of the energy storage system, the energy state of the battery energy storage system, the revenue obtained from user-side energy storage for self-use, the adjustment cost of the user's adjustable resources, and the user's willingness to participate in grid aggregation.

3. The method for user-side energy storage to participate in ancillary services according to claim 2, characterized in that, The adjustment cost of each adjustable resource on the forecast date is calculated, thereby obtaining the user's willingness to participate in grid aggregation, including: The revenue obtained by users from self-use of energy storage is calculated based on the charging and discharging power of the battery energy storage system, the upper and lower limits of the rated power of the energy storage system, and the energy state of the battery energy storage system. Calculate the adjustment costs of adjustable resources; The willingness of users to participate in grid aggregation is determined by the revenue obtained from user-side energy storage for self-use and the adjustment costs of adjustable resources.

4. The method for user-side energy storage to participate in ancillary services according to claim 1, characterized in that, The process of obtaining the pre-acquired ancillary service transaction announcement information includes: Obtain ancillary service transaction announcements, send information to users of adjustable resources, and obtain power consumption and adjustable information of adjustable resources from users of adjustable resources.

5. The method for user-side energy storage to participate in ancillary services according to claim 1, characterized in that, The application for pricing of adjustable resources participating in peak-shaving auxiliary services, with the goal of maximizing user benefits, is conducted using the following pricing method: in, Adjustable resource pricing for individual users. η is the maximum bid price stipulated by the trading rules, and η is the adjustment coefficient.

6. The method for user-side energy storage to participate in ancillary services according to claim 1, characterized in that, Following the declaration of quantity and price for adjustable resources participating in peak-shaving auxiliary services with the goal of maximizing user benefits, the following is also included: The actual transaction volume is obtained by matching the reported information on idle energy storage resources, the declared price of adjustable resources, and the electricity demand of the power grid on the actual operating day. After the actual operation day ends, the settlement result of the user-side energy storage charging and discharging fee is obtained based on the actual transaction volume.

7. A user-side energy storage participating in ancillary services system, characterized in that, include: The load forecasting module is used to perform load forecasting on the forecast date using a pre-built load forecasting clustering method to obtain the load forecast value for each adjustable resource. The cost willingness calculation module is used to calculate the adjustment cost of each adjustable resource on the forecast date using a pre-built user-side energy storage physical-behavior model, thereby obtaining the user's willingness to participate in grid aggregation; The ancillary service participation module is used to submit applications for the quantity and price of adjustable resources participating in peak shaving ancillary services, based on the adjustment cost of each adjustable resource, the user's willingness to participate in grid aggregation, the load forecast value of each adjustable resource, and the pre-acquired ancillary service transaction announcement information, with the goal of maximizing user benefits. In the load forecasting module, the construction process of the pre-built load forecasting clustering method includes: Based on historical electricity load data on the user side, a clustering algorithm is used to cluster adjustable resources on the user side, resulting in a load prediction clustering method based on the grouped load electricity consumption pattern. In the load forecasting module, the step of clustering adjustable resources on the user side based on historical electricity load data on the user side using a clustering algorithm includes: Historical electricity load data on the user side is analyzed according to season; Based on the seasonal analysis results, load data for each day of the year is collected and recorded to obtain historical electricity load data on the user side. Abnormal and undesirable data in the historical electricity load data of the user side are identified and deleted to obtain corrected historical electricity load data of all adjustable resources on the user side; Based on the historical electricity load data of all adjustable resources on the user side after correction, the SOINN clustering algorithm is used to extract the typical curves and cluster centers of each category of adjustable resources. The cluster centers are used as the load prediction values ​​of adjustable resources, thus obtaining a load prediction clustering method based on the grouped load electricity consumption pattern. Based on the historical electricity load data of all adjustable resources on the user side after correction, the SOINN clustering algorithm is used to extract typical curves and cluster centers for each category of adjustable resources. The cluster centers are used as the load prediction values ​​for adjustable resources, including: Initialize the network parameters and neuron set A = {C1,C2} by randomly initializing the initial weights and biases. Given actual load data, the first layer of the SOINN network adaptively generates neurons online to represent the input data. The neuron computation method is as follows: (1) in, This represents a new input sample; The weights of neurons are represented by A; the set of neurons is represented by A; and any neuron in the set of neurons is represented by T. The method for winning neurons is as follows: (2) in, This represents the new neuron value; The output of the first layer of the neural network is calculated using the current network parameters. The second layer calculates the distance between neurons based on the result of the first layer and uses this distance as a parameter to run the SOINN algorithm again with the neurons of the first layer as input, thereby making dynamic adjustments. If new input data is received, the loop repeats until the stopping condition is met, and finally the set of winning neurons is output as the load prediction value of the adjustable resources.

8. The user-side energy storage participating in ancillary services system according to claim 7, characterized in that, In the cost willingness calculation module, the user-side energy storage physical-behavioral model includes: the charging and discharging power of the battery energy storage system, the upper and lower bounds of the rated power of the energy storage system, the energy state of the battery energy storage system, the revenue obtained by the user-side energy storage for self-use, the adjustment cost of the user's adjustable resources, and the user's willingness to participate in grid aggregation.

9. The user-side energy storage participating in ancillary services system according to claim 7, characterized in that, The adjustment cost of each adjustable resource on the forecast date is calculated, thereby obtaining the user's willingness to participate in grid aggregation, including: The revenue obtained by users from self-use of energy storage is calculated based on the charging and discharging power of the battery energy storage system, the upper and lower limits of the rated power of the energy storage system, and the energy state of the battery energy storage system. Calculate the adjustment costs of adjustable resources; The willingness of users to participate in grid aggregation is determined by the revenue obtained from user-side energy storage for self-use and the adjustment costs of adjustable resources.

10. The user-side energy storage participating in ancillary services system according to claim 7, characterized in that, In the auxiliary service participation module, the process of obtaining the pre-acquired auxiliary service transaction announcement information includes: Obtain ancillary service transaction announcements, send information to users of adjustable resources, and obtain power consumption and adjustable information of adjustable resources from users of adjustable resources.

11. The user-side energy storage participating in ancillary services system according to claim 7, characterized in that, In the auxiliary service participation module, the declaration of quantity and price for adjustable resources to participate in peak shaving auxiliary services with the goal of maximizing user benefits is carried out, and the specific quotation method is as follows: in, Adjustable resource pricing for individual users. η is the maximum bid price stipulated by the trading rules, and η is the adjustment coefficient.

12. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, implements the user-side energy storage participation ancillary service method according to any one of claims 1-6.

13. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the user-side energy storage participation ancillary service method according to any one of claims 1-6.

14. A computer program product, the computer program product comprising computer instructions, characterized in that, The computer instructions instruct the computer to execute the user-side energy storage participation ancillary service method according to any one of claims 1-6.

Images