Method, device and storage medium for supporting electricity users to initiate retail transaction invitations

Abstract

The present invention relates to a method, device and storage medium for supporting power users to initiate retail transaction invitations, and is applied to the field of power market-oriented transaction technology. The method comprises: calculating the core points of the currently listed retail packages as representatives of the currently listed retail packages to characterize the preference of this type of power sales company for the retail packages under the current market conditions, performing offset calculation on the core points of the currently listed retail packages and the core points of historical retail packages, that is, correcting the historical statistics of the preferred retail packages of this type of power sales company according to the current market conditions, generating a retail package sample that can simultaneously reflect the historical retail packages and current market conditions of this type of power sales company according to the joint distribution of the corrected retail parameters, screening the retail package samples and then introducing the currently listed retail packages to jointly constitute the retail packages to be recommended by this type of power sales company, and using a preset recommendation algorithm to push the listed retail packages and virtual retail packages to users.

Description

Technical Field

[0001] The present invention relates to the technical field of electricity market-based trading, and in particular to a method, device and storage medium for supporting electricity users to initiate retail transaction invitations. Background Art

[0002] Electricity users can participate in the electricity market through the power sales company and sign a retail package with the power sales company. Instead of using the catalog electricity price, they will settle electricity bills based on the signed retail package. In the following years, in order to better serve the development of the electricity retail market, the Power Exchange Center launched an electricity retail trading platform. The retail packages signed between power sales companies and electricity users must be signed and filed through the retail trading platform to be legal for transactions and settlements.

[0003] In addition to the traditional transaction model of "power sales companies put packages on the shelves and users purchase packages", the current power retail platform has been configured with the function of "users initiating transaction invitations to power sales companies". However, in the power market, there is a natural information asymmetry between users and power sales companies. Users do not have the professional knowledge to configure the parameters of retail packages, nor do they know which power sales companies they should strive for and are more likely to close deals with. In addition, the current rules of the power market change every year. The number of retail packages listed on the retail platform is relatively small, so the frequency of users initiating transactions is low. Summary of the Invention

[0004] In view of this, the purpose of the present invention is to provide a method, device and storage medium for supporting electricity users to initiate retail transaction invitations, so as to solve the problem in the prior art that there is natural information asymmetry between users and power sales companies, users do not have the professional knowledge to configure the parameters of retail packages, and are not clear about which power sales companies they should strive for and are more likely to close deals with. In addition, the current rules of the electricity market change every year, and the number of retail packages on the retail platform is relatively small, so the frequency of users initiating transactions is low.

[0005] According to a first aspect of an embodiment of the present invention, a method for supporting an electricity user to initiate a retail transaction invitation is provided, the method comprising:

[0006] Obtain historical transaction data of power sales companies, parameters of power retail packages signed by power sales companies in the past, and current retail packages on power retail platforms;

[0007] Constructing a feature vector for each power sales company based on the historical transaction behavior data of the power sales company and the parameters of the power retail packages historically signed by the power sales company;

[0008] According to the characteristic vector of each power sales company, a preset clustering algorithm is used to cluster the power sales companies to obtain N types of power sales companies;

[0009] Construct the core feature vector of the retail package of each type of power sales company based on all the parameters of the power retail package signed historically by each type of power sales company;

[0010] Use the vine-copula function to obtain the joint distribution fitting parameters of each type of power sales company;

[0011] Obtaining offsets of N retail package parameters for each type of electricity sales company based on the retail packages currently available on the electricity retail platform;

[0012] The mean of the N retail package parameters of each type of electricity sales company is corrected according to the offset of each retail package parameter to obtain the joint distribution of the corrected retail parameters of each type of electricity sales company;

[0013] Generate a retail package sample for each type of electricity sales company according to the joint distribution of the retail parameters modified for each type of electricity sales company;

[0014] Screening the retail package samples of each type of electricity sales company to obtain virtual retail packages of each type of electricity sales company after screening;

[0015] The virtual retail packages of each type of power sales company and the currently available power retail packages are combined to obtain the recommended retail packages of each type of power sales company. The recommended retail packages of N types of power sales companies are combined to obtain the recommended retail packages of the entire market.

[0016] The market-wide retail packages to be recommended are recommended to electricity users according to a preset recommendation method.

[0017] Preferably,

[0018] The joint distribution fitting parameters of each type of power sales company obtained by the vine-copula function include:

[0019] Set the marginal distribution model for each parameter, use the marginal distribution model with different parameters to describe the parameter distribution of N retail packages of each type of electricity retail company, use the vine-copula function to describe the correlation between the parameters, and obtain the joint distribution fitting parameters of each type of electricity retail company;

[0020] The fitting parameters of each marginal distribution model include the mean or any one parameter is obtained by calculating the mean.

[0021] Preferably,

[0022] The offsets for obtaining N retail package parameters for each type of electricity sales company include:

[0023] Obtain the currently available electricity retail packages of each type of electricity sales company based on the currently available retail packages of the electricity retail platform; obtain the current feature vector of each type of electricity sales company based on the currently available electricity retail packages of each type of electricity sales company;

[0024] Obtain offsets of N retail package parameters according to the core feature vector of the retail package of each type of electricity sales company and the current feature vector.

[0025] Preferably,

[0026] The retail package samples of each type of electricity sales company are screened to obtain the virtual retail packages of each type of electricity sales company after screening, including:

[0027] Obtain the core point feature vectors of the currently available electricity retail packages of each type of electricity sales company, calculate the Euclidean distance between the feature vectors of the currently available electricity retail packages and the core point feature vectors of the currently available electricity retail packages one by one, and obtain the maximum Euclidean distance between the currently available electricity retail packages of each type of electricity sales company and the core point feature vectors of the currently available electricity retail packages;

[0028] From the retail package samples of each type of power sales company, all retail packages whose Euclidean distance to the core point feature vector of the retail package is less than the maximum Euclidean distance are removed, and a virtual retail package of each type of power sales company is constructed using the remaining samples.

[0029] Preferably,

[0030] The method of recommending the retail packages to be recommended in the entire market to the electricity users according to the preset recommendation method includes:

[0031] Obtain historical behavior data of power users to be recommended;

[0032] Obtaining a feature vector of the power user to be recommended based on the historical behavior data of the power user to be recommended;

[0033] Obtaining a feature vector of each retail package among the retail packages to be recommended in the entire market;

[0034] Obtain similarities between the feature vector of the power user to be recommended and the feature vector of each retail package in the retail packages to be recommended in the entire market, sort the similarities by size, and sort the packages to be recommended to the power user to be recommended according to the similarity sorting.

[0035] Preferably,

[0036] The step of recommending the market-wide retail packages to be recommended to electricity users according to a preset recommendation method further includes:

[0037] Obtain all electricity retail packages that have been signed by the electricity user to be recommended in the past;

[0038] Obtaining the feature vector of each electricity retail package that the electricity user to be recommended has signed in the past;

[0039] The similarity between the feature vector of each electricity retail package historically signed by the electricity user to be recommended and the feature vector of each retail package in the entire market to be recommended is calculated pairwise, the similarities are sorted by size, and the recommended packages are sorted for the electricity user to be recommended based on the similarity sorting.

[0040] Preferably,

[0041] The step of recommending the market-wide retail packages to be recommended to electricity users according to a preset recommendation method further includes:

[0042] Calculate the similarity between the feature vector of the power user to be recommended and the feature vectors of other power users, sort the similarities, and obtain the power user with the highest similarity to the power user to be recommended;

[0043] Obtain all historically purchased retail packages of the electricity user with the highest similarity;

[0044] Obtain the feature vector of each historically purchased retail package of the electricity user with the highest similarity;

[0045] Calculate the similarity between the feature vector of each retail package purchased historically by the electricity user with the highest similarity and the feature vector of each retail package in the retail packages to be recommended in the entire market;

[0046] The similarities are sorted by size, and the packages to be recommended are sorted for the power users to be recommended according to the similarity sorting.

[0047] Preferably, it also includes:

[0048] When any electricity user selects any package to be recommended in the list of packages to be recommended;

[0049] Determine whether the selected package to be recommended is a currently available electricity retail package or a virtual retail package;

[0050] If a retail electricity package is currently available, the purchase page for the package will be displayed directly;

[0051] If it is a virtual retail package, the electricity user transaction invitation initiation page will be displayed, and the default parameters of the invitation initiation page are the parameters of the virtual retail package; when the electricity user confirms the invitation initiation, the transaction invitation will be sent to all electricity sales companies in the electricity sales company category to which the virtual retail package belongs.

[0052] According to a second aspect of an embodiment of the present invention, a device is provided for supporting an electricity user initiating a retail transaction invitation, the device comprising:

[0053] Data acquisition module: used to obtain historical transaction behavior data of power sales companies, parameters of power retail packages historically signed by power sales companies, and current retail packages on the power retail platform;

[0054] The power sales company feature acquisition module is used to construct a feature vector for each power sales company based on the historical transaction behavior data of the power sales company and the parameters of the power retail packages previously signed by the power sales company;

[0055] Power sales company classification module: used to cluster power sales companies according to the feature vector of each power sales company using a preset clustering algorithm to obtain N types of power sales companies;

[0056] Core point feature acquisition module: used to construct the core point feature vector of each type of power sales company's retail package based on all historically contracted power retail package parameters of each type of power sales company;

[0057] Joint distribution module: used to obtain the joint distribution fitting parameters of each type of power sales company through the vine-copula function;

[0058] An offset acquisition module is configured to acquire the offsets of N retail package parameters of each type of power sales company according to the retail packages currently available on the power retail platform;

[0059] Correction module: used to correct the mean of N retail package parameters of each type of power sales company according to the offset of each retail package parameter, and obtain the joint distribution of the corrected retail parameters of each type of power sales company;

[0060] Package sample generation module: used to generate retail package samples for each type of power sales company based on the joint distribution of the retail parameters revised for each type of power sales company;

[0061] Package sample screening module: used to screen the retail package samples of each type of power sales company and obtain the virtual retail packages of each type of power sales company after screening;

[0062] The module for obtaining the recommended packages is used to aggregate the virtual retail packages of each type of power sales company and the currently available power retail packages to obtain the recommended retail packages of each type of power sales company. The module also aggregates the recommended retail packages of N types of power sales companies to obtain the recommended retail packages for the entire market.

[0063] Recommendation module: used to recommend the retail packages to be recommended in the entire market to electricity users according to a preset recommendation method.

[0064] According to a third aspect of an embodiment of the present invention, a storage medium is provided, wherein the storage medium stores a computer program, and when the computer program is executed by a host controller, each step in the above method is implemented.

[0065] The technical solutions provided by the embodiments of the present invention may have the following beneficial effects:

[0066] This application calculates the core points of the currently listed retail packages as representatives of the currently listed retail packages to characterize the preferences of this type of power sales companies for retail packages under the current market conditions, and calculates the offset between the core points of the currently listed retail packages and the core points of the historical retail packages, that is, the historical statistics of the preferred retail packages of this type of power sales companies are revised according to the current market conditions, and a retail package sample that can simultaneously reflect the historical retail packages and current market conditions of this type of power sales companies is generated according to the joint distribution of the revised retail parameters. After screening the retail package samples, the currently listed retail packages are introduced to jointly constitute the retail packages to be recommended for this type of power sales companies, which is beneficial to Use a preset recommendation algorithm to push listed retail packages and virtual retail packages to users; the solution of this application provides technical support for electricity users to initiate transactions under the condition of information asymmetry in the electricity retail market, and makes full use of the virtual retail packages generated above to facilitate their transactions. It solves the problem in the existing technology that due to the natural information asymmetry between users and electricity sales companies, users do not have the professional knowledge to configure the parameters of the retail packages, and are not clear about which electricity sales companies they should strive for and are more likely to close deals with. In addition, the current rules of the electricity market change every year, the number of listed retail packages on the retail platform is small, and therefore the frequency of users initiating transactions is low.

[0067] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0069] Figure 1 is a flowchart illustrating a method for supporting an electricity user to initiate a retail transaction invitation according to an exemplary embodiment;

[0070] Figure 2 is a system diagram of an apparatus for supporting an electricity user to initiate a retail transaction invitation according to another exemplary embodiment;

[0071] In the attached figure: 1-data acquisition module, 2-power sales company feature acquisition module, 3-power sales company classification module, 4-core point feature acquisition module, 5-joint distribution module, 6-offset acquisition module, 7-correction module, 8-package sample generation module, 9-package sample screening module, 10-recommended package acquisition module, 11-recommendation module. DETAILED DESCRIPTION

[0072] Exemplary embodiments will be described in detail herein, examples of which are illustrated in the accompanying drawings. In the following description, when referring to the drawings, like numbers in different figures represent like or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all possible embodiments consistent with the present invention. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present invention, as detailed in the appended claims.

[0073] Example 1

[0074] Figure 1 FIG. 1 is a flow chart showing a method for supporting an electricity user to initiate a retail transaction invitation according to an exemplary embodiment. Figure 1 As shown, the method includes:

[0075] S1, obtain the historical transaction behavior data of the power sales company, the parameters of the power retail packages signed by the power sales company in the past, and the retail packages currently available on the power retail platform;

[0076] S2, constructing a feature vector for each power sales company based on the historical transaction behavior data of the power sales company and the parameters of the power retail packages previously signed by the power sales company;

[0077] S3, clustering the power sales companies using a preset clustering algorithm based on the characteristic vector of each power sales company to obtain N types of power sales companies;

[0078] S4, constructing the core feature vector of the retail package of each type of power sales company based on all the parameters of the power retail packages signed historically by each type of power sales company;

[0079] S5, obtain the joint distribution fitting parameters of each type of power sales company through the vine-copula function;

[0080] S6, obtaining offsets of N retail package parameters of each type of electricity sales company based on the retail packages currently available on the electricity retail platform;

[0081] S7, correcting the mean of the N retail package parameters of each type of electricity sales company according to the offset of each retail package parameter, to obtain a corrected joint distribution of the retail parameters of each type of electricity sales company;

[0082] S8, generating a retail package sample for each type of electricity sales company according to the joint distribution of the modified retail parameters of each type of electricity sales company;

[0083] S9, screening the retail package samples of each type of electricity sales company, and obtaining the virtual retail packages of each type of electricity sales company after screening;

[0084] S10, combining the virtual retail packages of each type of power sales company and the currently available power retail packages to obtain a recommended retail package for each type of power sales company, and combining the recommended retail packages of N types of power sales companies to obtain a recommended retail package for the entire market;

[0085] S11, recommending the market-wide retail package to be recommended to the electricity user according to a preset recommendation method;

[0086] It is understandable that obtaining the historical transaction behavior data of the power sales company, the parameters of the power retail packages historically signed by the power sales company, and the current retail packages on the power retail platform specifically includes:

[0087] Power sales company transaction behavior data:

[0088] Shareholding type, total assets, credit rating, historical contracted electricity users, annual electricity sales, electricity market revenue, electricity retail platform interaction data (including but not limited to historical retail packages launched, store maintenance status, etc.), electricity retail platform transaction behavior data (including but not limited to the number of transaction invitations initiated, the number of contract applications submitted, the number of contract terminations, the number of contract changes, etc.), etc.;

[0089] Parameters of electricity retail packages signed by power sales companies in the past:

[0090] Parameters in electricity retail packages that have a direct impact on electricity user settlement, such as package type, package price, package energy consumption, and profit sharing ratio;

[0091] The electricity retail platform currently offers retail packages;

[0092] Power sales company clustering:

[0093] Based on the power sales company data in the historical data of the power retail market (including the power sales company's transaction behavior data and the parameters of the power retail package signed by the power sales company in the past), the feature vector of the power sales company is constructed. , this eigenvector fully describes the transaction preferences of power sales companies in the electricity retail market;

[0094] During feature vector extraction, feature selection was performed based on the effectiveness of the feature vector in classifying electricity sales companies. Parameters whose data types were not numerical variables, such as the text variable "Package Type" in the electricity retail package parameters, were first converted to 0-1 variables before being used in feature vector extraction.

[0095] Use Euclidean distance to describe the characteristic vectors of different power sales companies The distance between them is used to cluster the power sales companies using density clustering, mean shift clustering and other clustering algorithms to obtain N types of power sales companies.

[0096] Get the core points of each type of electricity sales company's retail package:

[0097] Obtain all historically signed electricity retail packages for each type of electricity sales company and construct a retail package feature vector , the feature vector contains all the numerical parameters designed for all types of retail packages in the market. For parameters whose data types are not numerical variables, such as the "package type" in the electricity retail package parameters which is a text variable, they are first converted into 0-1 variables and then used for feature vector extraction;

[0098] The core feature vector of the retail package of this type of electricity sales company The calculation is as follows:

[0099]

[0100] Where, They represent the mean values ​​of the first, second, ... parameters in all historically contracted electricity retail packages of this type of electricity sales company;

[0101] Joint distribution statistics of retail packages of similar electricity sales companies:

[0102] Fit the joint distribution of N retail parameters in the retail package of the mth type of electricity sales company and calculate its distribution parameters, including:

[0103] For the retail packages of the mth type of power sales company, each parameter type in all historically contracted retail packages is considered as a statistical variable, such as the "package price" of all historically contracted retail packages as one statistical variable and the "sharing ratio" as another statistical variable; when the retail package types are different, the parameters of the retail packages are also different. For example, if there is a "sharing ratio" parameter in the "sharing package" but not in other packages, then the "sharing ratio" parameter in the other packages will be filled with the mean value of the "sharing ratio" parameter in the "sharing package" of this type of power sales company;

[0104] Using historical data, the vine-copula function is used to fit the joint distribution between the N parameters of the retail package, and the parameters of the joint distribution are calculated. The marginal distribution model of each parameter is selected according to the fitting effect of different models on historical data, but the selected model must include the mean , or one of the parameters can be directly obtained from the mean Calculated, optional models include but are not limited to:

[0105] Gaussian distribution: The fitting parameters are μ 、 ,in μ is the sample mean ;

[0106] Exponential distribution: The fitting parameters are λ ,in λ is the sample mean The reciprocal of ;

[0107] Poisson distribution: The fitting parameters are λ ,in λ is the sample mean ;

[0108] Chi-square distribution: The fitting parameters are k ,in k is the sample mean ;

[0109] Similar electricity sales companies generate virtual retail packages:

[0110] Obtain all the electricity retail packages currently on the shelves of the m-th type of electricity sales company, use the data of all the electricity retail packages currently on the shelves, and use the above-mentioned feature vector calculation method to recalculate the retail package feature vector at this time ;

[0111] For the mth type of electricity sales company, calculate the offset of N retail package parameters between the core point of the retail package under historical data and the core point of the currently available retail package;

[0112] The offset is calculated as follows:

[0113]

[0114] Where, 、 ,……and 、 ...represent the core feature vectors of the current retail package and core points of historical retail packages the 1st, 2nd, ... elements of ;

[0115] Calculate the offset of the mean of N retail parameters of the m-th type of power sales company to obtain the modified joint distribution of retail parameters;

[0116] Based on the offset calculated in the previous step, the joint distribution model of the N retail parameters of the m-th type of power sales company is modified. The mean of the n-th parameter is modified as follows:

[0117]

[0118] Generate retail package samples for the mth type of electricity sales company based on the modified joint distribution of retail parameters.

[0119] According to the corrected mean and other distribution parameters calculated above, a sample of retail packages of the mth type of electricity sales company is randomly generated. ;

[0120] Calculate the maximum Euclidean distance between all currently listed retail parameter packages and the core point:

[0121] For all retail packages currently on the shelves, calculate the feature vectors and compare them with the core feature vectors of the currently available retail packages one by one Calculate the Euclidean distance to obtain the maximum Euclidean distance between all currently listed retail parameter packages and the core point;

[0122] From the virtual retail package sample set Remove all virtual retail packages whose Euclidean distance to the core point of the current retail package is less than the maximum Euclidean distance calculated above, and the remaining samples are constructed into a virtual retail package set ; For virtual retail package collection In this step, the validity of the virtual retail package is corrected by removing parameters that do not conform to the "Package Type" in the sample retail package parameters and the current market definition of retail packages. For example, for sample packages that are not "sharing packages," the "sharing ratio" parameter is directly removed.

[0123] The mth type of electricity sales company has listed all retail packages With virtual retail package collection , together constitute the set of retail packages to be recommended by the mth type of electricity sales company The set of retail packages to be recommended by N types of power sales companies together constitutes the set of retail packages to be recommended in the entire market ;

[0124] To push retail packages and their ranking to electricity users, the steps are as follows:

[0125] Obtain historical behavior data of electricity users, including:

[0126] City of location, electricity consumption type or industry, voltage level, electricity consumption, load curve, historical electricity retail packages, electricity sales companies that have historically represented users, electricity bills, electricity retail platform interaction data (including but not limited to clicks / visits, favorites, price comparisons, chats initiated, etc.), electricity retail platform transaction behavior data (including but not limited to transaction invitations initiated, contract application submissions, contract terminations, contract changes, etc.), etc.

[0127] Calculate the power user feature vector based on the power user's historical behavior data;

[0128] Based on the electricity retail packages that electricity users have historically signed, calculate the feature vector of each historical retail package;

[0129] Calculate the set of retail packages to be recommended in the entire market The feature vector of each package to be recommended;

[0130] The feature vectors obtained by the above calculations are used to calculate the recommended retail packages and their rankings for each electricity user using a recommendation algorithm such as a content-based recommendation algorithm and a collaborative recommendation algorithm. Depending on the selected recommendation algorithm, the calculation of the feature vectors includes but is not limited to:

[0131] Content-based recommendation algorithm: For each electricity user, the similarity between the electricity user feature vector calculated above and the feature vectors of all the packages to be recommended is calculated pairwise, and the recommended packages are sorted by similarity;

[0132] The first typical collaborative recommendation algorithm: For each electricity user, based on the retail packages they have purchased in the past, the feature vectors of each historically purchased retail package calculated above are used to calculate the similarity between each pair and the feature vectors of the package to be recommended. The recommended packages are then sorted by similarity.

[0133] The second typical collaborative recommendation algorithm: For each electricity user, the electricity user feature vector calculated above is compared with the feature vectors of other electricity users in pairwise similarity calculation, and the electricity users are ranked by similarity. After selecting the electricity user with the highest similarity, the retail packages purchased by this electricity user with the highest similarity are selected. The feature vector of each historically purchased retail package is then compared with the feature vector of the package to be recommended in pairwise similarity calculation, and the packages to be recommended are ranked by similarity.

[0134] Push notifications to users on the electricity retail trading platform;

[0135] When any user clicks on the retail package recommended by the platform;

[0136] When the retail package clicked by the user is currently available, the user will be redirected to the purchase page of the retail package;

[0137] When the retail package clicked by the user is a virtual retail package, the user will be redirected to the electricity user transaction invitation initiation page, and the parameters of the virtual retail package will be used as the default parameters of the invitation initiation page; when the electricity user clicks to initiate the invitation, the transaction invitation will be sent to all electricity sales companies in the electricity sales company category to which the virtual retail package belongs.

[0138] Example 2

[0139] Figure 2 2 is a system diagram illustrating an apparatus for supporting an electricity user to initiate a retail transaction invitation according to another exemplary embodiment, the apparatus comprising:

[0140] Data acquisition module 1: used to obtain historical transaction behavior data of power sales companies, parameters of power retail packages historically signed by power sales companies, and current retail packages on the power retail platform;

[0141] Power sales company feature acquisition module 2: used to construct a feature vector for each power sales company based on the historical transaction behavior data of the power sales company and the parameters of the power retail packages previously signed by the power sales company;

[0142] Power sales company classification module 3: used to cluster the power sales companies according to the feature vector of each power sales company using a preset clustering algorithm to obtain N types of power sales companies;

[0143] Core point feature acquisition module 4: used to construct the core point feature vector of the retail package of each type of power sales company based on all the parameters of the power retail packages signed historically by each type of power sales company;

[0144] Joint distribution module 5: used to obtain the joint distribution fitting parameters of each type of power sales company through the vine-copula function;

[0145] Offset acquisition module 6: used to obtain the offsets of N retail package parameters of each type of power sales company according to the retail packages currently available on the power retail platform;

[0146] Correction module 7: used to correct the mean of N retail package parameters of each type of power sales company according to the offset of each retail package parameter, and obtain the corrected joint distribution of retail parameters of each type of power sales company;

[0147] Package sample generating module 8: used to generate retail package samples for each type of power sales company according to the joint distribution of the retail parameters corrected for each type of power sales company;

[0148] Package sample screening module 9: used to screen the retail package samples of each type of power sales company and obtain the virtual retail packages of each type of power sales company after screening;

[0149] The module 10 for obtaining the recommended packages is used to aggregate the virtual retail packages of each type of power sales company and the currently available power retail packages to obtain the recommended retail packages of each type of power sales company, and to aggregate the recommended retail packages of N types of power sales companies to obtain the recommended retail packages of the entire market;

[0150] Recommendation module 11: used for recommending the retail packages to be recommended in the entire market to electricity users according to a preset recommendation method.

[0151] Example 3:

[0152] This embodiment provides a storage medium, wherein the storage medium stores a computer program, and when the computer program is executed by a host controller, each step in the above method is implemented;

[0153] It is understandable that the storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.

[0154] It can be understood that the same or similar parts of the above embodiments can be referenced to each other, and the contents not described in detail in some embodiments can refer to the same or similar contents in other embodiments.

[0155] It should be noted that, in the description of the present invention, the terms "first", "second", etc. are used for descriptive purposes only and should not be understood as indicating or implying relative importance. In addition, in the description of the present invention, unless otherwise specified, the meaning of "plurality" is at least two.

[0156] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code comprising one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may be performed out of the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present invention pertain.

[0157] It should be understood that various components of the present invention may be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods may be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof may be used: a discrete logic circuit having logic gate circuits for implementing logic functions on data signals, an application-specific integrated circuit having suitable combinational logic gate circuits, a programmable gate array (PGA), a field-programmable gate array (FPGA), etc.

[0158] Those skilled in the art will understand that all or part of the steps in the method of the above embodiment can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium. When the program is executed, it includes one or a combination of the steps of the method embodiment.

[0159] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing module, or each unit may exist physically separately, or two or more units may be integrated into a single module. The aforementioned integrated modules may be implemented in the form of hardware or in the form of software functional modules. If the integrated modules are implemented in the form of software functional modules and sold or used as independent products, they may also be stored in a computer-readable storage medium.

[0160] The storage medium mentioned above can be a read-only memory, a magnetic disk or an optical disk, etc.

[0161] Throughout this specification, reference to terms such as "one embodiment," "some embodiments," "examples," "specific examples," or "some examples" means that a specific feature, structure, material, or characteristic described in conjunction with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0162] Although the embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and are not to be construed as limitations on the present invention. A person skilled in the art may change, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A method for supporting electricity users to initiate retail transaction invitations, characterized in that: The method comprises: Obtain historical transaction data of power sales companies, parameters of power retail packages signed by power sales companies in the past, and current retail packages on power retail platforms; Constructing a feature vector for each power sales company based on the power sales company's historical transaction behavior data and the parameters of the power retail packages it has historically signed, wherein the feature vector is used to describe the power sales company's transaction preferences in the power retail market; According to the characteristic vector of each power sales company, a preset clustering algorithm is used to cluster the power sales companies to obtain N types of power sales companies; Construct the core feature vector of the retail package of each type of power sales company based on all the parameters of the power retail package signed historically by each type of power sales company; Use the vine-copula function to obtain the joint distribution fitting parameters of each type of power sales company; Obtaining offsets of N retail package parameters for each type of electricity sales company based on the retail packages currently available on the electricity retail platform; The offsets for obtaining N retail package parameters for each type of electricity sales company include: Obtain the currently available electricity retail packages of each type of electricity sales company based on the currently available retail packages of the electricity retail platform; obtain the current feature vector of each type of electricity sales company based on the currently available electricity retail packages of each type of electricity sales company; Obtain offsets of N retail package parameters based on the core feature vector of the retail package of each type of electricity sales company and the current feature vector; The mean of the N retail package parameters of each type of electricity sales company is corrected according to the offset of each retail package parameter to obtain the joint distribution of the corrected retail parameters of each type of electricity sales company; Generate a retail package sample for each type of electricity sales company according to the joint distribution of the retail parameters modified for each type of electricity sales company; Screening the retail package samples of each type of electricity sales company to obtain virtual retail packages of each type of electricity sales company after screening; The retail package samples of each type of electricity sales company are screened to obtain the virtual retail packages of each type of electricity sales company after screening, including: Obtain the core point feature vectors of the currently available electricity retail packages of each type of electricity sales company, calculate the Euclidean distance between the feature vectors of the currently available electricity retail packages and the core point feature vectors of the currently available electricity retail packages one by one, and obtain the maximum Euclidean distance between the currently available electricity retail packages of each type of electricity sales company and the core point feature vectors of the currently available electricity retail packages; From the retail package samples of each type of electricity sales company, remove all retail packages whose Euclidean distance to the core point feature vector of the retail package is less than the maximum Euclidean distance, and construct a virtual retail package of each type of electricity sales company using the remaining samples; The virtual retail packages of each type of power sales company and the currently available power retail packages are combined to obtain the recommended retail packages of each type of power sales company. The recommended retail packages of N types of power sales companies are combined to obtain the recommended retail packages of the entire market. Recommending the market-wide retail packages to be recommended to electricity users according to a preset recommendation method; The method of recommending the retail packages to be recommended in the entire market to the electricity users according to the preset recommendation method includes: Obtain historical behavior data of the power user to be recommended, the historical behavior data including: voltage level, power consumption, load curve, historically signed power retail packages, power sales companies that have historically acted as agents for users, electricity charges, power retail platform interaction data, and power retail platform transaction behavior data; Obtaining a feature vector of the power user to be recommended based on the historical behavior data of the power user to be recommended; Obtaining a feature vector of each retail package among the retail packages to be recommended in the entire market; Obtaining similarities between a feature vector of the power user to be recommended and a feature vector of each retail package in the retail packages to be recommended in the entire market, sorting the similarities, and sorting the packages to be recommended to the power user to be recommended based on the similarity ranking; When any electricity user selects any package to be recommended in the list of packages to be recommended; Determine whether the selected package to be recommended is a currently available electricity retail package or a virtual retail package; If a retail electricity package is currently available, the purchase page for the package will be displayed directly; If it is a virtual retail package, the electricity user transaction invitation initiation page will be displayed, and the default parameters of the invitation initiation page are the parameters of the virtual retail package; when the electricity user confirms the invitation initiation, the transaction invitation will be sent to all electricity sales companies in the electricity sales company category to which the virtual retail package belongs.

2. The method according to claim 1, characterized in that The joint distribution fitting parameters of each type of power sales company obtained by the vine-copula function include: Set the marginal distribution model for each parameter, use the marginal distribution model with different parameters to describe the parameter distribution of N retail packages of each type of electricity retail company, use the vine-copula function to describe the correlation between the parameters, and obtain the joint distribution fitting parameters of each type of electricity retail company; The fitting parameters of each marginal distribution model include the mean or any one parameter is obtained by calculating the mean.

3. The method according to claim 2, characterized in that The step of recommending the market-wide retail packages to be recommended to electricity users according to a preset recommendation method further includes: Obtain all electricity retail packages that have been signed by the electricity user to be recommended in the past; Obtaining the feature vector of each electricity retail package that the electricity user to be recommended has signed in the past; The similarity between the feature vector of each electricity retail package historically signed by the electricity user to be recommended and the feature vector of each retail package in the entire market to be recommended is calculated pairwise, the similarities are sorted by size, and the recommended packages are sorted for the electricity user to be recommended based on the similarity sorting.

4. The method according to claim 3, characterized in that The step of recommending the market-wide retail packages to be recommended to electricity users according to a preset recommendation method further includes: Calculate the similarity between the feature vector of the power user to be recommended and the feature vectors of other power users, sort the similarities, and obtain the power user with the highest similarity to the power user to be recommended; Obtain all historically purchased retail packages of the electricity user with the highest similarity; Obtain the feature vector of each historically purchased retail package of the electricity user with the highest similarity; Calculate the similarity between the feature vector of each retail package purchased historically by the electricity user with the highest similarity and the feature vector of each retail package in the retail packages to be recommended in the entire market; The similarities are sorted by size, and the packages to be recommended are sorted for the power users to be recommended according to the similarity sorting.

5. A device for supporting electricity users to initiate retail transaction invitations, characterized in that: The device comprises: Data acquisition module: used to obtain historical transaction behavior data of power sales companies, parameters of power retail packages historically signed by power sales companies, and current retail packages on the power retail platform; The power sales company feature acquisition module is used to construct a feature vector for each power sales company based on the power sales company's historical transaction behavior data and the parameters of the power retail packages previously signed by the power sales company. The feature vector is used to describe the power sales company's transaction preferences in the power retail market. Power sales company classification module: used to cluster power sales companies according to the feature vector of each power sales company using a preset clustering algorithm to obtain N types of power sales companies; Core point feature acquisition module: used to construct the core point feature vector of each type of power sales company's retail package based on all historically contracted power retail package parameters of each type of power sales company; Joint distribution module: used to obtain the joint distribution fitting parameters of each type of power sales company through the vine-copula function; An offset acquisition module is configured to acquire the offsets of N retail package parameters of each type of power sales company according to the retail packages currently available on the power retail platform; The offsets for obtaining N retail package parameters for each type of electricity sales company include: Obtain the currently available electricity retail packages of each type of electricity sales company based on the currently available retail packages of the electricity retail platform; obtain the current feature vector of each type of electricity sales company based on the currently available electricity retail packages of each type of electricity sales company; Obtain offsets of N retail package parameters based on the core feature vector of the retail package of each type of electricity sales company and the current feature vector; Correction module: used to correct the mean of N retail package parameters of each type of power sales company according to the offset of each retail package parameter, and obtain the joint distribution of the corrected retail parameters of each type of power sales company; Package sample generation module: used to generate retail package samples for each type of power sales company based on the joint distribution of the retail parameters revised for each type of power sales company; Package sample screening module: used to screen the retail package samples of each type of power sales company and obtain the virtual retail packages of each type of power sales company after screening; The retail package samples of each type of electricity sales company are screened to obtain the virtual retail packages of each type of electricity sales company after screening, including: Obtain the core point feature vectors of the currently available electricity retail packages of each type of electricity sales company, calculate the Euclidean distance between the feature vectors of the currently available electricity retail packages and the core point feature vectors of the currently available electricity retail packages one by one, and obtain the maximum Euclidean distance between the currently available electricity retail packages of each type of electricity sales company and the core point feature vectors of the currently available electricity retail packages; From the retail package samples of each type of electricity sales company, remove all retail packages whose Euclidean distance to the core point feature vector of the retail package is less than the maximum Euclidean distance, and construct a virtual retail package of each type of electricity sales company using the remaining samples; The module for obtaining the recommended packages is used to aggregate the virtual retail packages of each type of power sales company and the currently available power retail packages to obtain the recommended retail packages of each type of power sales company. The module also aggregates the recommended retail packages of N types of power sales companies to obtain the recommended retail packages for the entire market. Recommendation module: used for recommending the retail packages to be recommended in the entire market to electricity users according to a preset recommendation method; The method of recommending the retail packages to be recommended in the entire market to the electricity users according to the preset recommendation method includes: Obtain historical behavior data of the power user to be recommended, the historical behavior data including: voltage level, power consumption, load curve, historically signed power retail packages, power sales companies that have historically acted as agents for users, electricity charges, power retail platform interaction data, and power retail platform transaction behavior data; Obtaining a feature vector of the power user to be recommended based on the historical behavior data of the power user to be recommended; Obtaining a feature vector of each retail package among the retail packages to be recommended in the entire market; Obtaining similarities between a feature vector of the power user to be recommended and a feature vector of each retail package in the retail packages to be recommended in the entire market, sorting the similarities, and sorting the packages to be recommended to the power user to be recommended based on the similarity ranking; When any electricity user selects any package to be recommended in the list of packages to be recommended; Determine whether the selected package to be recommended is a currently available electricity retail package or a virtual retail package; If a retail electricity package is currently available, the purchase page for the package will be displayed directly; If it is a virtual retail package, the electricity user transaction invitation initiation page will be displayed, and the default parameters of the invitation initiation page are the parameters of the virtual retail package; when the electricity user confirms the invitation initiation, the transaction invitation will be sent to all electricity sales companies in the electricity sales company category to which the virtual retail package belongs.

6. A storage medium, characterized in that The storage medium stores a computer program, and when the computer program is executed by the main controller, the computer program implements the steps of the method for supporting electricity users to initiate retail transaction invitations according to any one of claims 1 to 4.

Images