CN116579884A - Power user behavior analysis method and system - Google Patents

Abstract

The invention belongs to the technical field of electric power control, and relates to a method and system for analyzing power user behavior. The method includes the following steps: calculating the related user set corresponding to the target user according to the related degree; Electricity consumption under the target time period; distribute electric power to the target user according to the electricity consumption of the target user under the target time period. The present invention can more accurately predict the power consumption of the target user in a specific time period by analyzing the power consumption behavior of the user set with a high degree of correlation with the target user; Distribution, so as to avoid the waste of power resources in the park, through accurate prediction and distribution of power, can effectively reduce the operating pressure of the power system in the park and improve its operating efficiency.

Description

A method and system for analyzing power user behavior

technical field

The present invention relates to the technical field of electric power control, in particular to an electric power user behavior analysis method and system.

Background technique

The park power system mainly refers to the power supply services provided for enterprises or institutions in the park, such as industrial parks, science and technology parks, university towns, etc.; the power distribution of the park usually needs to consider the power demand and consumption of each user in the park behavior to achieve efficient use of power resources.

In the traditional power distribution system, the prediction and distribution of users' power demand is usually based on historical data and empirical rules. It often requires a lot of computing resources and time, and it is difficult to deal with the mutual influence and dependence between users, and it is impossible to respond to the operation status of the power system and the power demand of users in real time to cope with sudden changes in power demand, which not only reduces the The operating efficiency of the power system may also lead to unfair distribution of power resources.

Therefore, how to accurately predict the power demand of users in the park and realize accurate power distribution according to the prediction results is an urgent problem to be solved in the power control of the park.

Contents of the invention

In view of the fact that in the prior art: it is impossible to accurately predict the power demand of users in the park, and realize accurate power distribution according to the prediction results, the present invention provides a power user behavior analysis method and system, which can effectively solve the problems in the background technology.

In order to solve the problems of the technologies described above, the technical solutions provided by the present invention are specifically as follows:

In the first aspect, a power user behavior analysis method includes the following steps:

Obtain the power consumption data corresponding to the users in the park, and construct a three-dimensional matrix corresponding to the park according to the power consumption data;

According to the three-dimensional matrix, respectively calculate the correlation degree corresponding to the target user and the user in the park;

Calculate the associated user set corresponding to the target user according to the degree of association;

Calculate the power consumption of the target user in the target time period according to the associated user set;

Allocating power to the target users according to the power consumption of the target users in the target time period.

In any of the above schemes, it is preferable to obtain the electricity consumption data corresponding to the users in the park respectively, and construct a three-dimensional matrix corresponding to the park according to the electricity consumption data, including:

Setting a time separation period, the time separation period includes at least one time separation interval;

Obtain the corresponding user power consumption data in at least one time interval interval under the current time interval period;

Sorting and grouping the user electricity consumption data according to the unique identifiers corresponding to the users and time separation intervals respectively, to obtain a three-dimensional matrix corresponding to the park under the current time separation period.

In any of the above schemes, it is preferable to calculate the corresponding degree of association between the target user and the user in the park according to the three-dimensional matrix, including:

Calculate the first data link corresponding to the target user and the user in the park respectively according to the three-dimensional matrix corresponding to the park under the current time separation period;

Obtain at least one first sub-chain of the first data link according to at least one time-separated interval under the current time-separated period;

Set the matching threshold, calculate the matching degree between the first sub-chain corresponding to the target user and the user in the park, and if the matching degree is less than the matching threshold, then judge that the first sub-chain corresponding to the target user and the in-park The first sub-chain corresponding to the user matches;

Extract the matching first sub-chains from the first data chains corresponding to the target user and the users in the park respectively, and generate a second data chain;

Traversing the time division intervals corresponding to the first sub-chains in the second data chain, and connecting the first sub-chains corresponding to the adjacent time division intervals to obtain the second sub-chains;

Extracting the second sub-chain corresponding to the maximum time length to obtain at least one third data chain.

In any of the above schemes, preferably, the first data link corresponding to the target user and the user in the park is calculated according to the three-dimensional matrix corresponding to the park under the current time separation period, including:

According to the three-dimensional matrix corresponding to the park in the previous time separation period, the feature points are obtained , where /> for the category of the user, /> for user /> in the time-separated interval /> power consumption, /> is the i-th time interval, /> for user /> in the time-separated interval /> Timestamp of electricity consumption;

The feature points are sorted according to the time stamp, and the first data links corresponding to the target user and the user in the park are respectively obtained.

In any of the above schemes, it is preferable to calculate the matching degree between the first sub-chain corresponding to the target user and the user in the park, including:

By formula: , calculate the matching degree between the first sub-chain corresponding to the target user and the users in the park respectively/> , where /> is the power consumption of target user u at time stamp t, /> is the power consumption of user w in the park at time stamp t, /> and /> Respectively, the target user u corresponds to the minimum and maximum power consumption of the first sub-chain, /> and /> are the minimum and maximum values of power consumption corresponding to the first sub-chain for user w in the park, respectively.

In any of the above schemes, preferably, calculating the associated user set corresponding to the target user according to the degree of association includes:

According to the length of the first sub-chain of the second data link and the length of the third data link, the degree of assimilation between the target user and different users in the park is calculated sequentially;

According to the degree of assimilation, different users in the park are distributed in a layered manner to obtain a layered distribution map;

Calculate the selected area, extract the park users corresponding to the selected area in the layered distribution map, and obtain the associated user set.

In any of the above schemes, preferably, according to the length of the first sub-chain of the second data link and the length of the third data link, the degree of assimilation between the target user and different users in the park is sequentially calculated, including:

By formula: , calculate the degree of assimilation between the target user and different users in the park /> , where /> is the Pearson coefficient of target user u and campus user w, /> is the dynamic weight.

In any of the above schemes, it is preferable to calculate the power consumption of the target user in the target time period according to the associated user set, including:

Corresponding the target time period to the time interval interval under the previous time interval period to obtain the target historical time interval interval;

Obtain the power consumption of each element in the associated user set in the target historical time interval;

Calculate the average power consumption of the target user and each element in the associated user set for the current segmentation period;

According to the power consumption of each element in the associated user set in the target historical time interval, and the average power consumption of the target user and each element in the associated user set for the current segmentation period, determine the target user at the target time power consumption under the segment.

In any of the above schemes, preferably, according to the power consumption of each element in the associated user set in the target historical time interval, and the average value of the target user and each element in the associated user set for the current division period Power consumption, to determine the power consumption of target users in the target time period, including:

By formula: , to calculate the power consumption of the target user in the target time period, where, /> is the power consumption of the target user u in the target time period corresponding to the time division interval under the current time division period, /> is the average power consumption of the target user in the corresponding time division interval of the current time division period, /> The average power consumption of user v in the corresponding time division interval of the current time division period is set as the associated user, /> Collect the power consumption of user v in the target historical time interval for the associated user, /> is the degree of assimilation between user v between the target user and the associated user.

In a second aspect, a power user behavior analysis system, the system includes:

A building module, used to obtain power consumption data corresponding to users in the park, and construct a three-dimensional matrix corresponding to the park according to the power consumption data;

A calculation module, configured to calculate the degree of relevance between the target user and the user in the park respectively according to the three-dimensional matrix;

An association module, configured to calculate an associated user set corresponding to the target user according to the degree of association;

A prediction module, configured to calculate the power consumption of the target user in the target time period according to the associated user set;

An allocating module, configured to allocate power to target users according to the target user's power consumption in a target time period.

Compared with prior art, the beneficial effect of the present invention:

The present invention can more accurately predict the power consumption of the target user in a specific time period by analyzing the power consumption behavior of the user set with a high degree of correlation with the target user; Distribution, so as to avoid the waste of power resources in the park, through accurate prediction and distribution of power, can effectively reduce the operating pressure of the power system in the park and improve its operating efficiency.

Description of drawings

The accompanying drawings are used for further understanding of the present invention, are used together with the embodiments of the present invention to explain the present invention, and do not constitute a limitation to the present invention.

Fig. 1 is a schematic flow chart of the power user behavior analysis method of the present invention;

Fig. 2 is a block diagram of the power user behavior analysis system of the present invention.

Explanation of symbols in the figure:

10. Construction module; 20. Calculation module; 30. Association module; 40. Prediction module; 50. Assignment module.

Detailed ways

In order to make the object, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described here are only used to explain the present invention, not to limit the present invention.

In addition, the terms "first" and "second" are used for descriptive purposes only, and cannot be interpreted as indicating or implying relative importance or implicitly specifying the quantity of indicated technical features. Thus, a feature defined as "first" and "second" may explicitly or implicitly include one or more of these features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

In order to better understand the above technical solution, the technical solution of the present invention will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

As shown in Figure 1, a power user behavior analysis method includes the following steps:

Step 1. Obtain the power consumption data corresponding to the users in the park, and construct a three-dimensional matrix corresponding to the park according to the power consumption data;

Step 2, according to the three-dimensional matrix, respectively calculate the correlation degree corresponding to the target user and the user in the park;

Step 3, calculating the associated user set corresponding to the target user according to the degree of association;

Step 4, calculating the power consumption of the target user in the target time period according to the associated user set;

Step 5, allocating power to the target user according to the power consumption of the target user in the target time period.

It should be pointed out that the above steps are only a preferred implementation sequence. During the specific implementation process, some steps can be replaced without affecting the overall implementation effect. In order to explain the technical solution of the application more clearly, the following content The solution is explained in a preferred manner.

In another optional embodiment of the present invention, the above step 1 also includes:

Step 11, setting a time separation cycle, the time separation cycle includes at least one time separation interval;

Step 12, obtaining the corresponding user electricity consumption data in at least one time interval interval under the current time interval period;

Step 13: Sorting and grouping the user electricity consumption data according to the unique identifiers corresponding to the users and time separation intervals respectively, to obtain a three-dimensional matrix corresponding to the park under the current time separation period.

In this embodiment, for the determination of the time separation period involved in the above step 11, the time separation period may be divided by hour, day, week or month. For example, according to one day as the time separation period, each hour of the day constitutes at least one time separation interval, so as to better analyze and predict the user's electricity consumption in different time periods; for the above step 13, After obtaining the user's electricity consumption data, it is necessary to sort and group the user's electricity consumption data. The basis for sorting and grouping is the unique identifier of the user and the time interval, such as user ID and date. After sorting, the park can be directly obtained The electricity consumption of users in each time interval in the park; after grouping, you can see the electricity consumption of each user in the park in the same time interval for easy comparison and analysis; finally, the sorted and grouped data is converted into a three-dimensional matrix, and in In this matrix, one dimension represents users, one dimension represents time intervals, and the other dimension represents electricity consumption.

In another optional embodiment of the present invention, the above step 2 also includes:

Step 21, according to the three-dimensional matrix corresponding to the park under the current time separation period, calculate the first data link corresponding to the target user and the users in the park respectively;

Step 22: Obtain at least one first sub-chain of the first data link according to at least one time-separated interval under the current time-separated cycle;

Step 23: Set a matching threshold, calculate the degree of matching between the first sub-chains corresponding to the target user and users in the park, and if the matching degree is less than the matching threshold, determine the first sub-chain corresponding to the target user Match the first sub-chain corresponding to users in the park;

Step 24, extracting the matching first sub-links from the first data links corresponding to the target user and the users in the park respectively, to generate a second data link;

Step 25, traversing the time division intervals corresponding to the first sub-chain in the second data chain, and connecting the first sub-chains corresponding to the adjacent time division intervals to obtain the second sub-chain;

Step 26, extracting the second sub-chain corresponding to the maximum time length to obtain at least one third data chain.

In this embodiment, a first data link can be calculated for each user (including the target user and other users in the park) according to the three-dimensional matrix corresponding to the park in the current time separation period; the first data link is Under a certain time separation period, the user's electricity consumption behavior sequence; for the first data link, multiple first sub-chains can be divided according to the time separation period. For example, if the data link is based on the electricity consumption recorded per hour, then 24 hours of each day can be regarded as a sub-chain; then a matching threshold can be set, and then the first data link of the target user and other users in the park can be calculated The matching degree of the first sub-chain, find the matching first sub-chain, extract them from their respective first data chains, and form a second data chain that only contains the same electricity consumption behavior of the target user and other users in the park , and then through the above processing to obtain the third data link with the longest timeline coherence, which can make it easier to judge the similarity of electricity consumption behavior between two users.

Optionally, in the above step 21, the first data link corresponding to the target user and the user in the park is calculated according to the three-dimensional matrix corresponding to the park under the current time separation period, including:

According to the three-dimensional matrix corresponding to the park in the previous time separation period, the feature points are obtained , where /> for the category of the user, /> for user /> in the time-separated interval /> power consumption, /> is the i-th time interval, /> for user /> in the time-separated interval /> Timestamp of electricity consumption;

The feature points are sorted according to the time stamp, and the first data links corresponding to the target user and the user in the park are respectively obtained.

In this embodiment, through the above content, the original three-dimensional matrix data can be converted into a data chain that is more convenient for processing and analysis, and the efficiency of data processing can be improved.

Optionally, in the above step 23, calculating the matching degree between the first sub-chains corresponding to the target user and the users in the park respectively includes:

By formula: , calculate the matching degree between the first sub-chain corresponding to the target user and the users in the park respectively/> , where /> is the power consumption of target user u at time stamp t, /> is the power consumption of user w in the park at time stamp t, /> and /> Respectively, the target user u corresponds to the minimum and maximum power consumption of the first sub-chain, /> and /> are the minimum and maximum values of power consumption corresponding to the first sub-chain for user w in the park, respectively.

In this embodiment, because the power consumption of different users may have large differences, direct comparison may not reflect the actual matching degree, and the power consumption of the target user and users in the park can be normalized by the above formula The power consumption of different users at the same time stamp can be compared under the same standard; and the similarity between the power consumption patterns of the target user and other users in the park in the same time period can be calculated, so that it can be identified Users with similar electricity usage patterns.

In another optional embodiment of the present invention, the above step 3 may include:

Step 31, according to the length of the first sub-chain of the second data link and the length of the third data link, the degree of assimilation between the target user and different users in the park is calculated sequentially;

Step 32, according to the degree of assimilation, different users in the park are distributed in a layered manner to obtain a layered distribution map;

Step 33, calculate the selected area, extract the park users corresponding to the selected area from the layered distribution map, and obtain the associated user set.

Optionally, in the above step 31, the degree of assimilation between the target user and different users in the park is sequentially calculated according to the length of the first sub-chain of the second data link and the length of the third data link, including:

By formula: , calculate the degree of assimilation between the target user and different users in the park /> , where /> is the Pearson coefficient of target user u and campus user w, /> is the dynamic weight.

In this example, by formula:

, calculate dynamic weights /> , where /> is a constant, /> is the number of the first sub-chain in the second data link between the target user u and the user w in the park, /> is the number of the first sub-chain in the first data chain of the target user u, /> is the number of first sub-chains in the first data chain of user w in the park, /> is the length of the third data link between the target user u and the user w in the park, /> is the length of the first data link of the target user u, /> is the length of the first data link of the target user w; in this embodiment, the more the number of the first sub-links in the second data link, the higher the similarity between the target user and the users in the park, so The longer the length of the third data link, the higher the similarity between the target user and the users in the park can be reflected. Therefore, the characteristics of the second data link and the third data link can be calculated according to the above formula Fusion, so as to obtain a more accurate calculation result compared with the general similarity algorithm, and then by integrating the fused similarity calculation result into the Pearson coefficient, the final assimilation calculation result can be obtained, which can fully and accurately calculate the target user The sameness with the electricity consumption behavior of users in the park.

Optionally, in the above step 32, different users in the park are distributed layered according to the degree of assimilation to obtain a layered distribution map, including:

Set the assimilation degree threshold and obtain the number k of users in the park that meet the assimilation degree threshold;

Take the square root of k and round the square root of k to the nearest integer m;

It will be judged whether k/m is an integer, and if it is not an integer, k/m will be rounded to the nearest integer Y, and m will be set as the number of layers distributed in layers, and Y will be the number of users distributed in each layer;

Sort the users according to the assimilation degree of users from high to low, and get the user assimilation degree range of each layer according to the number Y of users in each layer;

According to the range of user assimilation degree of each layer, users are assigned to the corresponding layers to obtain a layered distribution map.

Optionally, in the above step 33, the selected area is calculated, and the park users corresponding to the selected area in the layered distribution map are extracted to obtain the associated user set, including:

Obtain the maximum layer number U whose minimum user assimilation degree range meets the assimilation degree threshold in the layered distribution graph;

By formula: , to calculate the area ST of the selected area, where, /> is the layer width, is the layer spacing;

Extract the park users corresponding to the area ST in the layered distribution map to obtain the associated user set.

In this embodiment, through the above content, large-scale user data can be processed more efficiently when the number of users in the park is very large; by layered distribution according to the user assimilation degree, the user assimilation degree of each layer is There is a clear range, which can more accurately reflect the differences between users, and can find user groups more similar to target users more quickly, saving computing resources and time.

In another optional embodiment of the present invention, the above step 4 includes:

Step 41, corresponding the target time period to the time separation interval under the previous time separation cycle, to obtain the target historical time separation interval;

Step 42, obtaining the power consumption of each element in the associated user set in the target historical time interval;

Step 43, calculating the average power consumption of the target user and each element in the associated user set for the current segmentation period;

Step 44: Determine the target user according to the power consumption of each element in the associated user set in the target historical time interval, and the average power consumption between the target user and each element in the associated user set for the current split period Electricity consumption during the target time period.

Further, step 44, according to the power consumption of each element in the associated user set in the target historical time interval, and the average power consumption of the target user and each element in the associated user set for the current segmentation period, Determine the electricity consumption of the target user in the target time period, including:

By formula: , to calculate the power consumption of the target user in the target time period, where, /> is the power consumption of the target user u in the target time period corresponding to the time division interval under the current time division period, /> is the average power consumption of the target user in the corresponding time division interval of the current time division period, /> The average power consumption of user v in the corresponding time division interval of the current time division period is set as the associated user, /> Collect the power consumption of user v in the target historical time interval for the associated user, /> is the degree of assimilation between user v between the target user and the associated user.

In this embodiment, the historical power consumption of the target user and the assimilation degree and power consumption of the associated user set can be aggregated through the above formula, which can improve the prediction accuracy of the power consumption of the target user in the target time period, and further According to the above step 5, reasonable power is allocated to the target users in the target time period; it can be dynamically adjusted according to the actual application scenario, which expands the application scenarios of the overall algorithm.

As shown in Figure 2, the present invention also provides a power user behavior analysis system, the system comprising:

The construction module 10 is used to obtain the power consumption data corresponding to the users in the park, and construct a three-dimensional matrix corresponding to the park according to the power consumption data;

Calculation module 20, is used for respectively calculating the correlation degree corresponding to the target user and the user in the park according to the three-dimensional matrix;

An association module 30, configured to calculate an associated user set corresponding to the target user according to the degree of association;

A prediction module 40, configured to calculate the power consumption of the target user in the target time period according to the associated user set;

The allocation module 50 is configured to allocate power to the target user according to the power consumption of the target user within the target time period.

The above are only preferred embodiments of the present invention, and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, for those skilled in the art, it can still be described in the foregoing embodiments. modify the technical solution, or replace some of the technical features in an equivalent manner. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims (8)

1. A power user behavior analysis method is characterized in that: the method comprises the following steps:

the method for obtaining the electricity utilization data respectively corresponding to the users in the park and constructing the three-dimensional matrix corresponding to the park according to the electricity utilization data comprises the following steps: setting a time separation period, wherein the time separation period comprises at least one time separation interval; acquiring user electricity data corresponding to at least one time separation interval in the current time separation period; sorting and grouping the user electricity data according to unique identifiers respectively corresponding to the user and the time separation interval to obtain a three-dimensional matrix corresponding to the park in the current time separation period;

and respectively calculating the association degree of the target user and the user in the park according to the three-dimensional matrix, wherein the association degree comprises the following steps: calculating first data chains respectively corresponding to the target users and the users in the park according to the three-dimensional matrix corresponding to the park in the current time separation period; obtaining at least one first sub-chain of the first data chain according to at least one time separation interval in the current time separation period; setting a matching threshold, calculating the matching degree between the first sub-chains corresponding to the target user and the users in the park, and if the matching degree is smaller than the matching threshold, judging that the first sub-chain corresponding to the target user is matched with the first sub-chain corresponding to the users in the park; extracting matched first sub-chains in first data chains respectively corresponding to the target user and the users in the park to generate a second data chain; traversing a time separation interval corresponding to a first sub-chain in the second data chain, and connecting the first sub-chains corresponding to adjacent time separation intervals to obtain a second sub-chain; extracting the second sub-chain corresponding to the maximum value of the time length to obtain at least one third data chain;

calculating an associated user set corresponding to the target user according to the association degree;

calculating the electricity consumption of the target user in the target time period according to the associated user set;

and distributing power to the target user according to the power consumption of the target user in the target time period.

2. The power consumer behavior analysis method of claim 1, wherein: calculating a first data chain respectively corresponding to the target user and the user in the park according to the three-dimensional matrix corresponding to the park in the current time separation period, wherein the first data chain comprises the following components:

obtaining characteristic points according to the three-dimensional matrix corresponding to the park in the previous time separation periodWherein->For the category of the user, < >>For user->In the time interval->Is (are) the electricity consumption of the car>For the ith time interval, +.>For user->In the time interval->Is used up of the time stamp of electricity;

and sequencing the characteristic points according to the time stamps to respectively obtain first data chains corresponding to the target users and the users in the campus.

3. The power consumer behavior analysis method of claim 2, wherein: calculating the matching degree between the first sub-chains respectively corresponding to the target user and the users in the campus comprises the following steps:

by the formula:calculating the matching degree between the first sub-chains respectively corresponding to the target user and the users in the campus>Wherein->For the power consumption of the target user u at the time stamp t +.>For the power consumption of the users w in the campus at the time stamp t +.>And->Respectively corresponding to the minimum value and the maximum value of the electricity consumption of the first sub-chain for the target user u, +.>And->And respectively obtaining the minimum value and the maximum value of the electricity consumption of the first sub-chain corresponding to the user w in the park.

4. A method of power consumer behavior analysis according to claim 3, wherein: calculating an associated user set corresponding to the target user according to the association degree, wherein the method comprises the following steps:

sequentially calculating the degree of assimilation between the target user and different users in the park according to the lengths of the first sub-chain and the third data chain of the second data chain;

carrying out layered distribution on different users in a park according to the degree of assimilation to obtain a layered distribution diagram;

and calculating a selected area, and extracting park users corresponding to the selected area by the layered distribution map to obtain an associated user set.

5. The power consumer behavior analysis method of claim 4, wherein: and sequentially calculating the degree of identity between the target user and different users in the campus according to the lengths of the first sub-chain and the third data chain of the second data chain, wherein the method comprises the following steps:

by the formula:calculating the degree of identity between the target user and the different users in the campus +.>Wherein->Pearson coefficients for target user u and campus user w>Is a dynamic weight.

6. The power consumer behavior analysis method of claim 5, wherein: calculating the electricity consumption of the target user in the target time period according to the associated user set, wherein the electricity consumption comprises the following steps:

the target time period corresponds to the time separation interval in the last time separation period, and a target historical time separation interval is obtained;

acquiring the electricity consumption of each element in the associated user set in the target historical time separation interval;

calculating average power consumption of each element in the target user and the associated user set for the current segmentation period;

and determining the electricity consumption of the target user in the target time period according to the electricity consumption of each element in the associated user set in the target historical time separation interval and the average electricity consumption of each element in the target user and the associated user set for the current segmentation period.

7. The power consumer behavior analysis method of claim 6, wherein: according to the electricity consumption of each element in the associated user set in the target historical time separation interval and the average electricity consumption of each element in the associated user set for the current segmentation period, determining the electricity consumption of the target user in the target time period comprises the following steps:

by the formula:calculating the power consumption of the target user in the target time period, wherein +.>For the power consumption of the target user u in the time division section of the target time period corresponding to the current time division period,/>To average the power consumption of the target user for the time-divided interval corresponding to the current time-divided period,for the average power consumption of the user v in the time division section corresponding to the current time division period in the associated user set,/>For the associated user set the power consumption of user v in said target historical time interval, +.>The degree of assimilation between the user v is concentrated for the target user and the associated user.

8. An electric power user behavior analysis system, characterized in that: the system comprises:

the construction module (10) is used for acquiring the electricity utilization data corresponding to the users in the park and constructing a three-dimensional matrix corresponding to the park according to the electricity utilization data, and comprises the following steps: setting a time separation period, wherein the time separation period comprises at least one time separation interval; acquiring user electricity data corresponding to at least one time separation interval in the current time separation period; sorting and grouping the user electricity data according to unique identifiers respectively corresponding to the user and the time separation interval to obtain a three-dimensional matrix corresponding to the park in the current time separation period;

the calculating module (20) is used for calculating the association degree of the target user and the user in the park according to the three-dimensional matrix, and comprises the following steps: calculating first data chains respectively corresponding to the target users and the users in the park according to the three-dimensional matrix corresponding to the park in the current time separation period; obtaining at least one first sub-chain of the first data chain according to at least one time separation interval in the current time separation period; setting a matching threshold, calculating the matching degree between the first sub-chains corresponding to the target user and the users in the park, and if the matching degree is smaller than the matching threshold, judging that the first sub-chain corresponding to the target user is matched with the first sub-chain corresponding to the users in the park; extracting matched first sub-chains in first data chains respectively corresponding to the target user and the users in the park to generate a second data chain; traversing a time separation interval corresponding to a first sub-chain in the second data chain, and connecting the first sub-chains corresponding to adjacent time separation intervals to obtain a second sub-chain; extracting the second sub-chain corresponding to the maximum value of the time length to obtain at least one third data chain;

the association module (30) is used for calculating an association user set corresponding to the target user according to the association degree;

a prediction module (40) for calculating the electricity consumption of the target user in the target time period according to the associated user set;

and the distribution module (50) is used for distributing power to the target users according to the power consumption of the target users in the target time period.