阅读说明：本技术 基于多层复杂网络的四网融合用户用能数据模型构建方法 (Method for constructing energy consumption data model of four-network fusion user based on multilayer complex network ) 是由 朱华 赵现平 刘柱揆 杨政 潘侃 尹春林 于 2020-05-25 设计创作，主要内容包括：本发明涉及一种基于多层复杂网络的四网融合用户用能数据模型构建方法,包括：S01,采集多维样本使用数据；S02,基于所述多维样本数据对每个粒度网络进行建模；S03,构建四层网络；S04构建属性层；S05,将每一层均构建为图G(V,E,A)。本发明基于四网融合采集的用户数据,结合多层复杂网络方法,从多层次、多维度、多粒度3个层面构建成用户用能数据模型,具有结构清晰、关系分明的特点,再现用户数据在3个方面呈现特性的形成机理,实现用户数据的全方位研究,更能体现真实四网融合复杂数据系统,全面直观反映不同客户、产品、渠道在“四网融合”背景下的关联关系,同时有利于对用户用能行的有效分析及预测。(The invention relates to a method for constructing a four-network convergence user energy consumption data model based on a multilayer complex network, which comprises the following steps: s01, collecting multidimensional sample use data; s02, modeling each particle size network based on the multi-dimensional sample data; s03, constructing a four-layer network; s04 constructing an attribute layer; s05, constructing each layer into a graph G (V, E, A). The invention constructs a user energy data model from 3 levels of multiple levels, multiple dimensions and multiple granularities based on the user data collected by the four-network integration and combines a multi-level complex network method, has the characteristics of clear structure and clear relationship, reproduces the formation mechanism of the presenting characteristics of the user data in 3 aspects, realizes the omnibearing research of the user data, can embody a real four-network integration complex data system, comprehensively and intuitively reflects the incidence relationship of different customers, products and channels under the background of the four-network integration, and is simultaneously beneficial to the effective analysis and prediction of the user energy.)

1. A method for constructing a four-network convergence user energy consumption data model based on a multilayer complex network is characterized by comprising the following steps:

collecting multi-dimensional sample use data, wherein the multi-dimensional sample data comprises power grid use data, telecommunication network use data, broadcasting and television network use data and internet use data;

modeling each granularity network based on the multi-dimensional sample data, taking the sample data of each dimension used by each user in each granularity as a node, and connecting the use edges with an interactive relation among the nodes of the sample data of each dimension;

building a four-layer network, wherein the four-layer network comprises an electric network layer, a telecommunication network layer, a broadcasting and television network layer and an interconnection network layer, putting the electric network use data, the telecommunication network use data, the broadcasting and television network use data and the internet use data of a plurality of users with each granularity in the corresponding layers, projecting all the nodes into each layer, and connecting the nodes representing the same user interaction relationship on the adjacent layers by using coupling edges;

constructing an attribute layer, wherein the attribute layer comprises a plurality of user types, each user type is used as an attribute node, the attribute node represents an attribute value of a node in each layer of the network, and when one node in each layer of the network contains the attribute value, the attribute node is connected with the node by using an attribute edge;

and constructing each layer as a graph G (V, E, A), wherein V represents a node set, E represents an edge set, and A is information stored by each node, and the information stored by each node comprises a user name, a user type and corresponding energy data.

2. The method for constructing the energy data model for the four-network converged user based on the multi-layer complex network as claimed in claim 1, wherein: the granularity includes units by user, units by business/cell, and units by city/district/county.

3. The method for constructing the energy data model for the four-network converged user based on the multi-layer complex network as claimed in claim 1, wherein: the plurality of user types of the attribute layer comprise enterprise users and residential users.

4. The method for constructing the energy data model for the four-network converged user based on the multi-layer complex network as claimed in claim 1, wherein: the internet layer, the radio and television network layer, the telecommunication layer, the television network layer and the attribute layer are arranged in sequence.

5. The method for constructing the energy data model for the four-network converged user based on the multi-layer complex network as claimed in claim 4, wherein: each attribute node of the attribute layer is connected to a user of the electrical grid layer.

6. The method for constructing the energy data model for the four-network converged user based on the multi-layer complex network as claimed in claim 1, wherein the method is characterized in thatIn the following steps: the said electric network layer, telecommunication network layer, broadcasting network layer and Internet layer are respectively constructed into G_A、G_B、G_CAnd G_DThe attribute layer is constructed as G_EA(V_E,E_EA,A_E) In which V is_EIs a collection of attribute nodes, E_EAIs a set of connected edges of said attribute node and said node in the electrical grid layer, A_EIs attribute information of an attribute node, G_A、G_B、G_CAnd G_DRespectively denoted as E_AB、E_AC、E_AD、E_BC、E_BD、E_CD。

Technical Field

The application relates to the technical field of data processing, in particular to a method for constructing a four-network fusion user energy consumption data model based on a multi-layer complex network.

Background

The core value of the four-network integration is a large database of power energy, the value increment is extremely high, and the four-network integration has strategic significance at the national level. The power industry is one of the important fields of big data technology application, and the big data of power is generated in each link of power production and has abundant commercial and social values. The electric power big data aims at service trend prediction and data value mining, and mode innovation and application promotion facing typical service scenes are achieved by using core key technologies in the aspects of data integration management, data storage, data calculation, analysis mining and the like. The power consumer behavior analysis is a typical user scene of power big data analysis and mining, and the behavior analysis work is further carried out on the user by analyzing and mining information such as a user power load curve and the like. The user behavior analysis result can be used for specific scenes such as power load prediction, power marketing, power planning operation and the like.

The power grid with the four-network integration shows the remarkable characteristic of high integration of power flow, information flow and service flow, and by means of the fact that a large amount of data of users can be collected by enabling power fibers to enter the household, the data comprise power utilization data, broadcast and television use data, telecommunication use data, internet use data and the like, how to analyze and mine the value of the user data (find the characteristics of the users and the behavior habits of energy utilization) and use the value, differentiated high-cost-performance products or services are finally provided for the users, and the power grid integration becomes one of the serious challenges faced by power grid enterprises after the four-network integration.

Disclosure of Invention

The application provides a method for constructing a four-network convergence user energy consumption data model based on a multilayer complex network, so that the four-network convergence user behavior can be conveniently mined, analyzed and predicted.

The technical scheme adopted by the application is as follows:

the invention provides a method for constructing a four-network convergence user energy consumption data model based on a multilayer complex network, which comprises the following steps:

collecting multi-dimensional sample use data, wherein the multi-dimensional sample data comprises power grid use data, telecommunication network use data, broadcasting and television network use data and internet use data;

modeling each granularity network based on the multi-dimensional sample data, taking the sample data of each dimension used by each user in each granularity as a node, and connecting the use edges with an interactive relation among the nodes of the sample data of each dimension;

building a four-layer network, wherein the four-layer network comprises an electric network layer, a telecommunication network layer, a broadcasting and television network layer and an interconnection network layer, putting the electric network use data, the telecommunication network use data, the broadcasting and television network use data and the internet use data of a plurality of users with each granularity in the corresponding layers, projecting all the nodes into each layer, and connecting the nodes representing the same user interaction relationship on the adjacent layers by using coupling edges;

constructing an attribute layer, wherein the attribute layer comprises a plurality of user types, each user type is used as an attribute node, the attribute node represents an attribute value of a node in each layer of the network, and when one node in each layer of the network contains the attribute value, the attribute node is connected with the node by using an attribute edge;

and constructing each layer as a graph G (V, E, A), wherein V represents a node set, E represents an edge set, and A is information stored by each node, and the information stored by each node comprises a user name, a user type and corresponding energy data.

Further, the granularity includes units by users, units by businesses/cells, and units by cities/districts/counties.

Further, the plurality of user types of the attribute layer include enterprise users, residential users.

Further, the internet layer, the radio and television network layer, the telecommunication layer, the television network layer and the attribute layer are sequentially arranged.

Further, each attribute node of the attribute layer is connected to a user of the electrical grid layer.

Furthermore, the electric network layer, the telecommunication network layer, the broadcasting network layer and the internet layer are respectively constructed into G_A、G_B、G_CAnd G_DThe attribute layer is constructed as G_EA(V_E,E_EA,A_E) In which V is_EIs a collection of attribute nodes, E_EAIs a set of connected edges of said attribute node and said node in the electrical grid layer, A_EIs attribute information of an attribute node, G_A、G_B、G_CAnd G_DRespectively denoted as E_AB、E_AC、E_AD、E_BC、E_BD、E_CD。

The technical scheme of the application has the following beneficial effects:

the invention constructs a user energy data model from 3 levels of multiple levels, multiple dimensions and multiple granularities based on the user data collected by the four-network integration and combines a multi-level complex network method, has the characteristics of clear structure and clear relationship, reproduces the formation mechanism of the presenting characteristics of the user data in 3 aspects, realizes the omnibearing research of the user data, can embody a real four-network integration complex data system, comprehensively and intuitively reflects the incidence relationship of different customers, products and channels under the background of the four-network integration, and is simultaneously beneficial to the effective analysis and prediction of the user energy.

Drawings

In order to more clearly explain the technical solution of the present application, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious to those skilled in the art that other drawings can be obtained according to the drawings without creative efforts.

FIG. 1 is a flow chart of a method for constructing a four-network convergence user energy consumption data model based on a multi-layer complex network;

FIG. 2 is a schematic diagram of a multi-dimensional network node projected as a multi-layer network;

fig. 3 is a schematic diagram of a user energy data model structure for four-network fusion.

Detailed Description

Reference will now be made in detail to embodiments, examples of which are illustrated in the accompanying drawings. When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following examples do not represent all embodiments consistent with the present application. But merely as exemplifications of systems and methods consistent with certain aspects of the application, as recited in the claims.

Fig. 1 is a flow chart of a method for constructing a four-network convergence user energy consumption data model based on a multi-layer complex network;

FIG. 2 is a schematic diagram of a multi-dimensional network node projected as a multi-layer network; fig. 3 is a schematic diagram of a user energy data model structure for four-network fusion.

The application provides a method for constructing a four-network convergence user energy consumption data model based on a multilayer complex network, which specifically comprises the following steps:

s01: collecting multi-dimensional sample use data, wherein the multi-dimensional sample data comprises power grid use data, telecommunication network use data, broadcasting and television network use data and internet use data;

s02: modeling each granularity network based on the multi-dimensional sample data, taking the sample data of each dimension used by each user in each granularity as a node, and connecting the use continuous edges with an interactive relation among the nodes of the sample data of each dimension, wherein the granularity comprises the user as a unit, the enterprise/district as a unit and the city/district/county as a unit;

with the entrance of power optical fiber, as the support of "three networks integration", the power grid is the most critical network. Any user has data using telecommunication, broadcast television and internet, and the data of the user can be reflected in the power grid, the telecommunication network, the broadcast television network and the internet.

S03: building a four-layer network, wherein the four-layer network comprises an electric network layer, a telecommunication network layer, a broadcasting and television network layer and an interconnection network layer, putting the electric network use data, the telecommunication network use data, the broadcasting and television network use data and the internet use data of a plurality of users with each granularity in the corresponding layers, projecting all the nodes into each layer, and connecting the nodes representing the same user interaction relationship on the adjacent layers by using coupling edges;

the step follows the multi-layer and multi-dimension of the energy data model for the four-network fusion user, wherein the multi-layer is realized by an electric network layer and a social layer consisting of a telecommunication network layer, a broadcasting network layer and an internet layer; multidimensional rendering projects different interactions of the same user in "four nets" into different layers.

The basic principle of a multi-dimensional network represented as a multi-layer network is shown in fig. 2. In the multidimensional network, 4 sections a, b, c and d have two different connecting edges which respectively represent a relation 1 and a relation 2. Separating relation 1 from relation 2 to form two relation layers, projecting four nodes into two layers respectively, and projecting a in relation 1 layer₁、b₁、c₁、d₁In relation 2 layer is a₂、b₂、c₂、d₂And connecting the same node between different layers by interlayer coupling edges, e.g. a₁And a₂Connected, etc. And finally, only adding connecting edges of the corresponding relation among all nodes in each layer. After the process, two mixed relations in the original network are separated, and each connection relation between the nodes is clearly visible.

S04: constructing an attribute layer, wherein the attribute layer comprises a plurality of user types, each user type is used as an attribute node, the attribute node represents an attribute value of a node in each layer of the network, and when one node in each layer of the network contains the attribute value, the attribute node is connected with the node by using an attribute edge, wherein the plurality of user types comprise enterprise users and resident users;

in many real-world applications, the topology of the network and the attribute information of the vertices are very important. However, existing graph clustering methods only consider one aspect and ignore the other. Therefore, the distribution of the attribute values of the vertexes in the clusters generated by the existing method is random; or loose intra-cluster association. Therefore, to improve the accuracy of the energy-use behavior prediction, this step integrates the structural similarity and the attribute similarity of the network in a complex network model containing attribute information. The basic idea is as follows:

it is assumed that a complex network containing attribute information can be represented as a graph G (V, E, a), where V denotes a set of nodes, E denotes a set of edges, and a ═ a₁,a₂,...,a_mRepresents m attributes associated with the vertex to describe the characteristics of the vertex; and attribute a_iHas a value range of Dom (a)_i)＝{a_i1,a_i2,...,a_ini}. Each node v_i∈ V contains an attribute vector a₁(v_i),...,a_m(v_i)]. Then the representation of the attribute amplification map is G_a＝(V∪V_a,E∪E_a) Wherein

Representing attribute nodes. Attribute node v_ijThe value representing attribute i is the jth attribute value. If a is_j(v_i)＝a_jkThen at node v_iAnd v_jkGenerate a property edge therebetween, i.e., (v)_i,v_jk)∈E_a. Called E structural edge set, E_aIs an attribute edge set.

The step is favorable for analyzing the similarity of the user energy behaviors more quickly by constructing the attribute layer, and the attribute similarity calculation of the user is converted into the node adjacency calculation, so that the load of analyzing the similarity by a clustering method is reduced.

In particular, according to the description of S01, if any user has data using telecommunication, broadcast television and internet, which will have electricity data, the data of the user will be reflected in the power grid, telecommunication network, broadcast television network and internet. In order to simplify the model, the attribute nodes of the attribute layer are only connected with the user nodes of the power grid layer, and the attribute relationship can be transmitted to the same user node of other layers through multi-hierarchy and multi-dimensionality.

The network formed after the above steps is schematically shown in fig. 3.

S05: constructing each layer as a graph G (V, E, A), wherein V represents a node set, E represents an edge set, A is information stored by each node, the information stored by each node comprises a user name, a user type and corresponding energy data, and a power grid layer, a telecommunication network layer, a broadcasting and television network layer and an internet layer are respectively constructed as G_A、G_B、G_CAnd G_DThe attribute layer is constructed as G_EA(V_E,E_EA,A_E) In which V is_EIs a collection of attribute nodes, E_EAIs a set of connected edges of said attribute node and said node in the electrical grid layer, A_EIs attribute information of an attribute node, G_A、G_B、G_CAnd G_DRespectively denoted as E_AB、E_AC、E_AD、E_BC、E_BD、E_CD。

In particular, based on the description of S03, the attribute amplification is mainly applied to attributes having a limited number of value ranges, and the attribute such as the user type meets the condition. However, the various energy-use data is an attribute of an infinite number of value ranges, and is not suitable for being expanded into an attribute layer. Therefore, this step makes some improvements to a in graph G (V, E, a), so that a can directly store various data of user nodes without attribute vector correspondence, and extend a limited number of value range attributes thereof into the attribute layer.

As shown in fig. 3, in the present embodiment, an internet layer, a radio and television network layer, a telecommunication layer, a radio and television network layer, and an attribute layer are sequentially arranged. Each attribute node of the attribute layer is connected to a user of the electrical grid layer.

The embodiment is based on user data acquired by four-network fusion, combines a multi-layer complex network method, constructs a user energy data model from 3 levels of multiple layers, multiple dimensions and multiple granularities, has the characteristics of clear structure and clear relationship, reproduces a formation mechanism of the user data in 3 aspects of presenting characteristics, realizes the omnibearing research of the user data, can embody a real four-network fusion complex data system, comprehensively and intuitively reflects the incidence relationship of different customers, products and channels under the background of four-network fusion, and is simultaneously beneficial to the effective analysis and prediction of the user energy.

The embodiments provided in the present application are only a few examples of the general concept of the present application, and do not limit the scope of the present application. Any other embodiments extended according to the scheme of the present application without inventive efforts will be within the scope of protection of the present application for a person skilled in the art.