CN115829117A - Method and system for short-term prediction of electricity consumption in any area based on terminal electricity consumption information - Google Patents
Method and system for short-term prediction of electricity consumption in any area based on terminal electricity consumption information Download PDFInfo
- Publication number
- CN115829117A CN115829117A CN202211504108.5A CN202211504108A CN115829117A CN 115829117 A CN115829117 A CN 115829117A CN 202211504108 A CN202211504108 A CN 202211504108A CN 115829117 A CN115829117 A CN 115829117A
- Authority
- CN
- China
- Prior art keywords
- information
- user
- electricity consumption
- prediction
- grid
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Abstract
The invention provides a method and a system for predicting the electricity consumption in any area in a short time based on terminal electricity consumption information, wherein the method comprises the following steps: collecting power consumption information of a terminal: acquiring personal information and electricity utilization related information of a terminal user; collecting auxiliary weather information: collecting weather information related to the power utilization information of the user, wherein the weather information comprises daily average temperature information and time-period temperature information; preprocessing the collected terminal electricity consumption information and auxiliary weather information, constructing a prediction model, and predicting the electricity consumption of the user on the future date; associating the power consumption information of each user with the position information of the user acquired in the first step; predicting the electricity consumption of any selected area in a short time: and randomly selecting certain map areas in the map to realize the short-time electricity consumption prediction of the selected areas. According to the method, the power consumption prediction model is independently established for all the power consumption users, and the position information of each user is associated with the map, so that the accurate scheduling of the resources in any region defined on the map is realized.
Description
Technical Field
The invention relates to research on a power consumption prediction method, in particular to a short-time prediction method and a short-time prediction system for power consumption in any area based on terminal power consumption information.
Background
The prediction of the power consumption is important in the aspects of power guarantee and power marketing, and the accurate prediction of the power consumption can provide important data support for reasonable scheduling of power resources and regional marketing of power customers. In a traditional power consumption prediction model and system, for a fixed area, supported data are regional power consumption information, power consumption behaviors, weather information and the like, such as power supply prediction of administrative areas of provincial level, city level, county level and the like.
On the other hand, in many cases, it is necessary to predict the amount of electricity used for an arbitrary geographical area. The traditional fixed area prediction model is based on the power consumption information of the district users, and is not independently modeled for each independent user, so that the power consumption prediction of any area in the geographical position cannot be realized, and the application range and depth of the power consumption prediction model are restricted.
Disclosure of Invention
The invention mainly solves the technical problems in the prior art, and provides a method and a system for predicting the short-term electricity consumption of any area based on terminal electricity consumption information. According to the method, the power consumption prediction model is independently established for all power consumption users, and the position information (longitude and latitude coordinates) of each user is associated with the map, so that accurate scheduling of area resources arbitrarily defined on the map is finally realized, and more comprehensive and accurate data support is provided.
The technical problem of the invention is mainly solved by the following technical scheme:
a short-time prediction method for electricity consumption of any area based on terminal electricity consumption information comprises the following steps:
step one, collecting terminal power consumption information: acquiring personal information and electricity utilization related information of a terminal user, wherein the personal information and the electricity utilization related information comprise user types, personal user ages, electricity utilization periods, daily electricity consumption and position information of users;
step two, collecting auxiliary weather information: collecting weather information related to the power utilization information of the user, wherein the weather information comprises daily average temperature information and time-period temperature information;
step three, predicting the electricity consumption of a single user in a short time: preprocessing the terminal electricity consumption information and the auxiliary weather information acquired in the first step and the second step, constructing a prediction model, and predicting the electricity consumption of the user on the future date;
step four, associating the power consumption of the user with a map: associating the power consumption information of each user with the position information of the user acquired in the first step;
step five, predicting the electricity consumption of any selected area in short time: and randomly selecting certain map areas in the map to realize the short-time electricity consumption prediction of the selected areas.
Further, the first step specifically includes:
step 1.1, collecting category information of a user, wherein the category information is divided into 2 types: individual users and enterprise users, and records the data;
step 1.2, collecting and recording age, electricity consumption time period and daily electricity consumption information of an individual user;
step 1.3, collecting the position information of the user, wherein the position information adopts a relative position as a record, and the realization method comprises the following steps:
a) Gridding all areas of a local city where a user is located, and numbering each grid in sequence by using Arabic numerals;
b) For the grid size division, a dynamic division mode is adopted: the method is characterized in that grids are divided into town areas with large population density according to the length of 100 meters and the width of 100 meters; the suburban area with low population density is divided into grids with the length of 1000 meters and the width of 1000 meters;
c) And binding the number of the grid where the user is located with the user in the grid area.
Further, the second step specifically comprises:
step 2.1, acquiring daily average temperature information of a prediction area required in 2 years;
step 2.2, acquiring 18-00-8 days of a regional working day needing prediction in 2 years;
and 2.3, acquiring temperature information of the non-working day whole day time period of the prediction area required in 2 years.
Further, the third step specifically comprises:
step 3.1, preprocessing part of data in the step one, wherein the preprocessing method comprises the following steps:
a) The user category adopts Boolean type records, wherein 0 represents an enterprise user, and 1 represents an individual user;
b) The age information of the individual user is recorded in the form of age groups, and is divided into 3 age groups: under 20 years old, 20-60 years old, and over 60 years old;
c) The daily electric quantity information is directly recorded by the daily electric quantity;
d) The electricity utilization period information directly records the electricity utilization behaviors of the total hours in 24 hours in a day in a digital mode;
and 3.2, preprocessing the data in the second step, wherein the preprocessing method comprises the following steps:
a) Recording the daily average temperature information of the area where the user is located in last two years, and directly recording the daily average temperature information by using a temperature value;
b) Recording 18-00-8-00-minute time interval temperature information of a working day of an area where the user is located in the last two years, and recording only time intervals with the temperature higher than 30 degrees and lower than 10 degrees, wherein if the temperature of 2 time intervals of 18;
c) Recording the whole day time interval temperature information of the non-working day of the area where the user is located in last two years, and only recording the time interval number of the temperature higher than 30 ℃ and lower than 10 ℃;
3.3, adopting a recurrent neural network framework to realize the prediction of the power consumption of each user on the future day, training samples to be various quantitative behaviors of a certain user and weather information of a time period corresponding to the behaviors, and labeling the samples to be the power consumption degree of the user on the behavior occurrence day;
and 3.4, aiming at all the power users, independently training a short-time power consumption prediction model to realize the prediction of the power consumption of the user on the future date.
Further, the fourth step specifically includes:
step 4.1, for each grid, calculating the users in the grid according to the position information of the users;
step 4.2, associating the grids with the power consumption prediction models of all the users of the grid class, wherein the association method is to record the prediction models of the corresponding users;
and 4.3, for each grid, linearly superposing the prediction results of all the prediction models associated with the grid, namely the prediction result of the daily power consumption of the grid, and recording the prediction result as the predicted power consumption of the grid.
Further, the fifth step specifically includes:
step 5.1, when a regular or irregular area is drawn in the map, calculating the grids completely contained in the closed area, namely all parts of the grids are in the closed area;
step 5.2, examining boundary grids of the selected area, and solving all boundary grids, wherein the area proportion of each grid is in the selected area;
step 5.4, the grid prediction power consumption which is obtained in the step 5.1 and completely comprises the grid is subjected to linear superposition summation to obtain Sum 1 ；
Step 5.5, the predicted power consumption of the boundary grid in the step 5.2 is multiplied by the percentage of the corresponding grid obtained in the step 5.2, and all the boundary grid data are added to obtain Sum 2 ；
And 5.6, obtaining Sum = Sum1+ Sum2 which is a short-term prediction result of the electricity consumption of any selected area.
A short-term prediction system for electricity consumption of any area based on terminal electricity consumption information comprises:
the terminal electricity consumption information acquisition module is used for acquiring terminal electricity consumption information, wherein the terminal electricity consumption information comprises user types, the ages of individual users, electricity consumption periods, daily electricity consumption and position information of the users;
the auxiliary weather information acquisition module is used for acquiring auxiliary weather information, wherein the auxiliary weather information comprises daily average temperature information and time-interval temperature information;
the short-time prediction module for the electricity consumption of the single user preprocesses the terminal electricity consumption information and the auxiliary weather information acquired by the terminal electricity consumption information acquisition module and the auxiliary weather information acquisition module, constructs a prediction model and predicts the electricity consumption of the user on the selected future date;
the user electricity consumption and map association module is used for associating each piece of user electricity consumption information with the position information of the user collected in the step one;
and the short-time electricity consumption prediction module for randomly selecting some map areas in the map to realize the short-time electricity consumption prediction of the selected areas.
Further, the single-user power consumption short-time prediction module preprocesses the terminal power consumption information and the auxiliary weather information acquired by the terminal power consumption information acquisition module and the auxiliary weather information acquisition module, constructs a prediction model, and predicts the future date power consumption selected by the user, and specifically comprises the following steps:
preprocessing partial data in the terminal electricity utilization information, wherein the preprocessing method comprises the following steps:
a) The user category adopts Boolean type records, wherein 0 represents an enterprise user, and 1 represents an individual user;
b) The age information of the individual user is recorded in the form of age groups, and is divided into 3 age groups: under 20 years old, 20-60 years old, and over 60 years old;
c) The daily electricity consumption information is directly recorded by daily electricity measurement;
d) The electricity utilization period information directly records the electricity utilization behaviors of the total hours in 24 hours in a day in a digital mode;
the auxiliary weather information is preprocessed, and the preprocessing method comprises the following steps:
a) Recording the daily average temperature information of the area where the user is located in last two years, and directly recording the daily average temperature information by using a temperature value;
b) Recording 18-00-8-00-minute time interval temperature information of a working day of an area where the user is located in the last two years, and recording only time intervals with the temperature higher than 30 degrees and lower than 10 degrees, wherein if the temperature of 2 time intervals of 18;
c) Recording the whole day time interval temperature information of the non-working day of the area where the user is located in last two years, and only recording the time interval number of the temperature higher than 30 ℃ and lower than 10 ℃;
adopting a cyclic neural network framework to realize the prediction of the power consumption of each user on the future day, wherein a training sample is various quantitative behaviors of a certain user and weather information of a time period corresponding to the behaviors, and a sample label is the power consumption degree of the user on a behavior occurrence day;
aiming at all power consumers, a short-time power consumption prediction model is trained independently, and the power consumption of the user on the selected future date is predicted.
Further, the associating module of the power consumption of the user and the map associates the information of the power consumption of each user with the position information of the user collected in the step one, and specifically includes:
for each grid, calculating the users in the grid through the position information of the users;
associating the grids with the power consumption prediction models of all the users of the grid class by recording the prediction models of the corresponding users;
and for each grid, linearly superposing the prediction results of all prediction models associated with the grid, namely the prediction result of the daily power consumption of the grid is recorded as the predicted power consumption of the grid.
Further, the short-term electricity consumption prediction module for any selected area randomly selects some map areas in the map to realize the short-term electricity consumption prediction of the selected area, and specifically includes:
when a regular or irregular area is drawn in a map, calculating a grid completely contained in the closed area, namely all parts of the grid are in the closed area;
inspecting boundary grids of the selected area, and solving all boundary grids, wherein the area proportion of each grid is in the selected area;
linear superposition summation is carried out on the grid prediction power consumption which is obtained by calculation and completely contains the grid, and Sum is obtained 1 ；
And multiplying the predicted power consumption of the boundary grids by the obtained percentage of the corresponding grids, and adding all the boundary grid data to obtain Sum 2 ；
And obtaining Sum = Sum1+ Sum2 as a short-term prediction result of the electricity consumption of any selected area.
The invention has the following beneficial effects: a lightweight electricity consumption prediction model is established for each user independently; the power consumption prediction for all the individual users is realized; dividing a geographic space by adopting a two-level grid mode, and realizing the calculation and prediction of the power supply amount of a grid area based on the sum of the power supply amount prediction of each independent user; and the power consumption prediction of any selected area is realized by calculating the number of grids in any area selected on the map in linkage with the map.
Drawings
FIG. 1 is a flow chart diagram of a short-term prediction method for electricity consumption in any area based on terminal electricity consumption information according to the present invention;
FIG. 2 is a block diagram of a short-term prediction system for electricity consumption in any area based on terminal electricity consumption information according to the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be obtained by a person skilled in the art without any inventive step based on the embodiments of the present invention, are within the scope of the present invention.
In the embodiment of the invention, a certain-level city in Hubei province is taken as an example, a method and a system for predicting the short-time electricity consumption of any area in the city based on terminal electricity consumption information are constructed, and the whole system is realized in a software mode.
As shown in fig. 1, an embodiment of the present invention provides a method for predicting electricity consumption in any area in a short time based on terminal electricity consumption information, including the following steps:
step one, collecting terminal power consumption information: the method comprises the steps of collecting personal information and electricity utilization related information of end users, wherein the personal information and the electricity utilization related information comprise user categories (personal users or enterprise users), the ages of the personal users, electricity utilization periods, daily electricity consumption and position information of the users. The concrete implementation steps are as follows:
step 1.1, collecting category information of a user, wherein the category information is divided into 2 types: individual users and enterprise users, and records the data;
step 1.2, collecting and recording age, electricity consumption time period and daily electricity consumption information of an individual user;
step 1.3, collecting the position information of the user, wherein the position information adopts a relative position as a record, and the realization method comprises the following steps:
a) Gridding all areas of a local city where a user is located, and numbering each grid in sequence by using Arabic numerals;
b) For the grid size division, a dynamic division mode is adopted: the method is characterized in that grids are divided into town areas with large population density according to the length of 100 meters and the width of 100 meters; the suburban area with low population density is divided into grids with the length of 1000 meters and the width of 1000 meters;
c) And binding the number of the grid where the user is located with the user in the grid area.
Step two, collecting auxiliary weather information: the method comprises the following steps of collecting weather information related to user power utilization information, wherein the weather information comprises daily average temperature information and time-interval temperature information, and the specific implementation steps are as follows:
step 2.1, acquiring daily average temperature information of a prediction area required in 2 years;
step 2.2, acquiring 18 parts by time period to 8 parts by time period of a working day of a prediction area in 2 years;
and 2.3, acquiring temperature information of the non-working day whole day time period of the prediction area required in 2 years.
Step three, predicting the electricity consumption of the single user in a short time: preprocessing all data acquired in the first step and the second step, and constructing a prediction model, wherein the specific implementation steps are as follows:
step 3.1, preprocessing partial data (other data except grid data) in the step one, wherein the preprocessing method comprises the following steps:
a) The user category adopts Boolean type records, wherein 0 represents an enterprise user, and 1 represents an individual user;
b) The age information of the individual user is recorded in the form of age groups, and is divided into 3 age groups: under 20 years old, 20-60 years old, and over 60 years old;
c) The daily electricity consumption information is directly recorded by daily electricity measurement;
d) The electricity utilization period information directly records the electricity utilization behaviors of the total hours in 24 hours in a day in a digital mode;
step 3.2, preprocessing the data in the step two, wherein the preprocessing method comprises the following steps:
a) Recording the daily average temperature information of the area where the user is located in last two years, and directly recording the daily average temperature information by using a temperature value;
b) Recording 18-00-8-00-minute time interval temperature information of a working day of an area where the user is located in the last two years, and recording only time intervals with the temperature higher than 30 degrees and lower than 10 degrees, wherein if the temperature of 2 time intervals of 18;
c) Recording the whole day time interval temperature information of the non-working day of the area where the user is located in last two years, and only recording the time interval number of the temperature higher than 30 ℃ and lower than 10 ℃;
and 3.3, adopting a recurrent neural network framework to realize the prediction of the power consumption of each user in the future (in a short time). Training samples are various quantitative behaviors of a certain user and weather information of a period corresponding to the behaviors, and a sample label is the electricity consumption degree of the user on a behavior occurrence day;
and 3.4, aiming at all the power users, independently training a short-time power consumption prediction model to realize the prediction of the power consumption of the user on the future date.
Step four, associating the power consumption of the user with a map: associating the power consumption information of each user with the position information of the user collected in the first step, and specifically realizing the following steps:
step 4.1, calculating the number of users in each grid according to the position information of the users;
step 4.2, associating the grids with the power consumption prediction models of all the users of the grid class, wherein the association method is to record the prediction models corresponding to the users;
step 4.3, for each grid, linearly superposing the prediction results of all prediction models associated with the grid, namely the prediction result of the daily power consumption of the grid is recorded as the predicted power consumption of the grid;
step five, predicting the electricity consumption of any selected area in short time: randomly selecting (delineating) certain map areas in a map to realize the short-time electricity consumption prediction of the selected areas, and the specific implementation steps are as follows:
step 5.1, when a regular or irregular area is drawn in the map, calculating the grids completely contained in the closed area (namely all parts of the grids are in the closed area);
and 5.2, considering the boundary grids (partially in the range of the selected area) of the selected area, and solving all the boundary grids, wherein the area proportion of each grid is in the selected area.
Step 5.4, the grid prediction power consumption which is obtained in the step 5.1 and completely comprises the grid is linearly superposed (summed) to obtain Sum 1 ；
Step 5.5, the predicted power consumption of the boundary grid in the step 5.2 is multiplied by the percentage of the corresponding grid obtained in the step 5.2, and all the boundary grid data are added to obtain Sum 2 ；
And 5.6, obtaining Sum = Sum1+ Sum2 which is a short-term prediction result of the electricity consumption of any selected area.
As shown in fig. 2, an embodiment of the present invention provides a system for predicting short-term power consumption of any area based on terminal power consumption information, including:
the terminal power consumption information acquisition module 10 is configured to acquire terminal power consumption information, specifically acquire personal information and power consumption related information of a terminal user, including information such as a user category (a personal user or an enterprise user), an age of the personal user, a power consumption period, a daily power consumption amount, and location information of the user. Firstly, collecting category information of users, including personal users and enterprise users, and recording the category information by Boolean types; secondly, acquiring and recording the age, electricity consumption time period and daily electricity consumption information of the individual user; and finally, dividing the geographic space of the user by adopting the position information of the user in a plane grid mode, dividing grids in the town area with higher population density by 100 meters long and 100 meters wide, dividing grids in the suburban area with lower population density by 1000 meters long and 1000 meters wide, and recording the grid area where each user is located.
In particular, if a region where a major event or a major incident occurs has a large influence on the electricity consumption data of the user, the electricity consumption data in the region or the period cannot be used as basic data, for example, a large-scale long-time power failure, a large-scale event gathering activity, a sports event, a flood disaster, and the like. The method adopted by the invention is that if the events such as the above occur in a certain area in a certain time period, the user data in the time period is removed and is not used as the basic data of the subsequent training prediction model.
And the auxiliary weather information acquisition module 20 is used for acquiring auxiliary weather information, and the main function of the auxiliary weather information acquisition module is to acquire weather information related to the power utilization information of the user, including daily average temperature information and time-interval temperature information. Daily average temperature information requires data acquisition for nearly 2 years for different regions. Time-interval temperature information needs to be considered in a situation-divided manner according to two conditions of working days and non-working days, wherein 18-00-8 days of working days; collecting temperature information of all day time intervals on non-working days.
Further, for the situation that temperature information of a part of areas in different time periods is difficult to obtain, temperature information of adjacent areas can be adopted for replacing; for the condition that partial region time-interval information is missing, a linear interpolation mode of adjacent time intervals is adopted to simulate missing data; and for the abnormal condition of the temperature data, if the outdoor temperature data in a certain period is too high or too low, deleting the data, and simulating the data in a difference value simulation mode under the condition of information missing.
And the single-user power consumption short-time prediction module 30 has the main functions of preprocessing all data acquired by the two modules and constructing a power consumption prediction model. For different types of data, different preprocessing modes are adopted: the electricity user category divides the treatment into individual users and enterprise users; the user age data is processed into three age groups of 20 years old or below, 20-60 years old and 60 years old or above; the daily electricity quantity information is processed into a daily electricity measurement numerical value; the electricity utilization period information is the electricity utilization behaviors recorded by numerical values of the total number of hours in 24 hours a day; the information of the average daily temperature recorded in Fahrenheit degree of the area where the user is located in the last two years is processed into a numerical value (degree); the time-interval temperature information is processed into how many time intervals of each day the temperature is more than 30 degrees or less than 10 degrees. A prediction model construction part in the module is realized by adopting a recurrent neural network framework, a training sample is each quantitative behavior of a certain user and weather information of a period corresponding to the behavior, and a sample label is the power consumption degree of the user on a behavior occurrence day; aiming at all power consumers, a short-time power consumption prediction model is trained independently, and the power consumption of the user on the selected future date is predicted.
Specifically, in the aspect of building the power consumption prediction model, a recurrent neural network module in a deep learning platform Tensorflow 2.0 is directly adopted to realize the training process of the power consumption prediction model.
Furthermore, because a power consumption prediction model needs to be trained independently for all users, and the sample features have the same probability distribution features, a basic model can be trained first, and then independent power consumption prediction models for each user can be trained respectively according to the power consumption behavior data of each user. And (3) training the basic model uniformly by adopting sample data of all users, stopping training after training to a certain step number (the model is obviously converged), and obtaining the basic model. And on the basis of the basic model, selecting the same recurrent neural network framework, replacing the sample set with the power consumption behavior data of each independent user, and continuously training the basic model to obtain a power consumption prediction model converging to the data samples of each independent user.
And a user power consumption and map association module 40, which has the main function of associating each user power consumption information with the position information of the user collected in the step one. The method is mainly realized by calculating which users exist in each grid according to the position information of the users and then adding the power consumption information of all the users.
Specifically, the grid is superimposed with predicted power consumption data of all users in the grid area, and the output of the module is the predicted power consumption data of the grid area with the grid as a unit.
The short-time power consumption prediction module 50 for the arbitrarily selected area has the main function of randomly selecting (delineating) certain map areas in a map to realize short-time power consumption prediction of the selected area. The method comprises two implementation modes: the first method is that when the selected graduation area contains complete grids, the calculated grid prediction power consumption completely containing the grids is linearly superposed (summed) to obtain Sum 1 (ii) a Secondly, if the selected area comprises a boundary grid (part of the boundary grid is in the selected area range), the area of the boundary grid is required to be solved, the area ratio of the grid area is compared with the predicted power consumption ratio of the grid, all the predicted power consumption of the boundary grid are collected and summed to obtain Sum2; finally, sum = Sum is obtained 1 +Sum 2 Namely the short-time prediction result of the electricity consumption of any selected area.
Those of skill would further appreciate that the exemplary elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various exemplary components and steps have been described above generally in terms of their functionality in order to clearly illustrate this interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or in a combination of the two. A software module may reside in random access memory, read only memory, electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.
The above description is only an embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (10)
1. A short-time prediction method for electricity consumption of any area based on terminal electricity consumption information is characterized in that: the method comprises the following steps:
step one, collecting terminal power consumption information: acquiring personal information and electricity utilization related information of a terminal user, wherein the personal information and the electricity utilization related information comprise user types, personal user ages, electricity utilization periods, daily electricity consumption and position information of users;
step two, collecting auxiliary weather information: collecting weather information related to the power utilization information of the user, wherein the weather information comprises daily average temperature information and time-period temperature information;
step three, predicting the electricity consumption of a single user in a short time: preprocessing the terminal power consumption information and the auxiliary weather information acquired in the first step and the second step, constructing a prediction model, and predicting the power consumption of the user on the selected future date;
step four, associating the power consumption of the user with a map: associating the power consumption information of each user with the position information of the user collected in the step one;
step five, predicting the electricity consumption of any selected area in short time: and randomly selecting certain map areas in the map to realize the short-time electricity consumption prediction of the selected areas.
2. The method for short-term prediction of electricity consumption in any area based on terminal electricity consumption information as claimed in claim 1, wherein: the first step specifically comprises the following steps:
step 1.1, collecting category information of a user, wherein the category information is divided into 2 types: individual users and enterprise users, and records the data;
step 1.2, collecting and recording age, electricity consumption time period and daily electricity consumption information of an individual user;
step 1.3, collecting the position information of the user, wherein the position information adopts a relative position as a record, and the realization method comprises the following steps:
a) Gridding all areas of a local city where a user is located, and numbering each grid in sequence by using Arabic numerals;
b) For mesh size division, a dynamic division mode is adopted: the method is characterized in that grids are divided into town areas with large population density according to the length of 100 meters and the width of 100 meters; the suburban area with low population density is divided into grids with the length of 1000 meters and the width of 1000 meters;
c) And binding the number of the grid where the user is located with the user in the grid area.
3. The method for short-term prediction of electricity consumption in any area based on terminal electricity consumption information as claimed in claim 1, wherein: the second step specifically comprises:
step 2.1, acquiring daily average temperature information of a prediction area required in 2 years;
step 2.2, acquiring 18-00-8 days of a regional working day needing prediction in 2 years;
and 2.3, acquiring temperature information of the non-working day whole day time period of the prediction area required in 2 years.
4. The method for short-term prediction of electricity consumption in any area based on terminal electricity consumption information as claimed in claim 1, wherein: the third step specifically comprises:
step 3.1, preprocessing part of data in the step one, wherein the preprocessing method comprises the following steps:
a) The user category adopts Boolean type records, wherein 0 represents an enterprise user, and 1 represents an individual user;
b) The age information of the individual user is recorded in the form of age groups, and is divided into 3 age groups: under 20 years old, 20-60 years old, and over 60 years old;
c) The daily electricity consumption information is directly recorded by daily electricity measurement;
d) The electricity utilization period information directly records the electricity utilization behaviors of the total hours in 24 hours in a day in a digital mode;
and 3.2, preprocessing the data in the second step, wherein the preprocessing method comprises the following steps:
a) Recording the daily average temperature information of the area where the user is located in last two years, and directly recording the daily average temperature information by using a temperature value;
b) Recording 18-00-8-00-minute time period temperature information of a working day of an area where the user is located in the last two years, and recording only time periods with the temperature higher than 30 ℃ and lower than 10 ℃, if the temperature is higher than 30 ℃ in 2 time periods of 18;
c) Recording the whole day time interval temperature information of the non-working day of the area where the user is located in last two years, and only recording the time interval number of the temperature higher than 30 ℃ and lower than 10 ℃;
3.3, adopting a recurrent neural network frame to realize the prediction of the power consumption of each user in the future, training samples as various quantitative behaviors of a certain user and weather information of a time period corresponding to the behaviors, and labeling the samples as the power consumption degrees of the user in the behavior occurrence day;
and 3.4, aiming at all the power users, independently training a short-time power consumption prediction model to realize the prediction of the power consumption of the user on the future date.
5. The method for short-term prediction of electricity consumption in any area based on terminal electricity consumption information as claimed in claim 1, wherein: the fourth step specifically comprises:
step 4.1, for each grid, calculating the users in the grid according to the position information of the users;
step 4.2, associating the grids with the power consumption prediction models of all the users of the grid class, wherein the association method is to record the prediction models of the corresponding users;
and 4.3, for each grid, linearly superposing the prediction results of all the prediction models associated with the grid, namely the prediction result of the daily power consumption of the grid, and recording the prediction result as the predicted power consumption of the grid.
6. The method for short-term prediction of electricity consumption in any area based on terminal electricity consumption information as claimed in claim 1, wherein: the fifth step specifically comprises:
step 5.1, when a regular or irregular area is drawn in the map, calculating the grids completely contained in the closed area, namely all parts of the grids are in the closed area;
step 5.2, inspecting the boundary grids of the selected area, and solving all the boundary grids, wherein the area proportion of each grid is in the selected area;
step 5.4, the grid prediction power consumption which is obtained in the step 5.1 and completely comprises the grid is subjected to linear superposition summation to obtain Sum 1 ；
Step 5.5, the predicted power consumption of the boundary grid in the step 5.2 is multiplied by the percentage of the corresponding grid obtained in the step 5.2, and all the boundary grid data are added to obtain Sum 2 ；
And 5.6, obtaining Sum = Sum1+ Sum2 which is a short-term prediction result of the electricity consumption of any selected area.
7. A short-term prediction system for electricity consumption of any area based on terminal electricity consumption information is characterized by comprising:
the terminal electricity consumption information acquisition module is used for acquiring terminal electricity consumption information, wherein the terminal electricity consumption information comprises user types, the ages of individual users, electricity consumption periods, daily electricity consumption and position information of the users;
the auxiliary weather information acquisition module is used for acquiring auxiliary weather information, wherein the auxiliary weather information comprises daily average temperature information and time-interval temperature information;
the short-time prediction module for the electricity consumption of the single user preprocesses the terminal electricity consumption information and the auxiliary weather information acquired by the terminal electricity consumption information acquisition module and the auxiliary weather information acquisition module, constructs a prediction model and predicts the electricity consumption of the user on the selected future date;
the user electricity consumption and map association module is used for associating each piece of user electricity consumption information with the position information of the user collected in the step one;
and the short-time electricity consumption prediction module for randomly selecting some map areas in the map to realize the short-time electricity consumption prediction of the selected areas.
8. The system for short-term prediction of electricity usage in any area based on terminal electricity usage information as claimed in claim 7, wherein: the single-user electricity consumption short-time prediction module preprocesses the terminal electricity consumption information and the auxiliary weather information acquired by the terminal electricity consumption information acquisition module and the auxiliary weather information acquisition module, constructs a prediction model, and predicts the electricity consumption of the user on the selected future date, and specifically comprises the following steps:
preprocessing partial data in the terminal electricity utilization information, wherein the preprocessing method comprises the following steps:
a) The user category adopts Boolean type records, wherein 0 represents an enterprise user, and 1 represents an individual user;
b) The age information of the individual user is recorded in the form of age groups, and is divided into 3 age groups: under 20 years old, 20-60 years old, and over 60 years old;
c) The daily electricity consumption information is directly recorded by daily electricity measurement;
d) The electricity utilization period information directly records the total electricity utilization behaviors in 24 hours in a day in a digital mode;
the auxiliary weather information is preprocessed, and the preprocessing method comprises the following steps:
a) Recording the daily average temperature information of the area where the user is located in last two years, and directly recording the daily average temperature information by using a temperature value;
b) Recording 18-00-8-00-minute time interval temperature information of a working day of an area where the user is located in the last two years, and recording only time intervals with the temperature higher than 30 degrees and lower than 10 degrees, wherein if the temperature of 2 time intervals of 18;
c) Recording the whole day time interval temperature information of the non-working day of the area where the user is located in last two years, and only recording the time interval number of the temperature higher than 30 ℃ and lower than 10 ℃;
adopting a cyclic neural network framework to realize the prediction of the power consumption of each user on the future day, wherein a training sample is various quantitative behaviors of a certain user and weather information of a time period corresponding to the behaviors, and a sample label is the power consumption degree of the user on a behavior occurrence day;
aiming at all power consumers, a short-time power consumption prediction model is trained independently, and the power consumption of the future date selected by the user is predicted.
9. The system for short-term prediction of electricity usage in any area based on terminal electricity usage information as claimed in claim 7, wherein: the user power consumption and map associating module associates each piece of user power consumption information with the position information of the user collected in the first step, and specifically includes:
for each grid, calculating the users in the grid through the position information of the users;
associating the grids with the power consumption prediction models of all the users of the grid class by recording the prediction models of the corresponding users;
and for each grid, linearly superposing the prediction results of all prediction models associated with the grid, namely the prediction result of the daily power consumption of the grid is recorded as the predicted power consumption of the grid.
10. The system for short-term prediction of electricity usage in any area based on terminal electricity usage information as claimed in claim 7, wherein: the short-time electricity consumption prediction module for the arbitrarily selected area randomly selects certain map areas in the map to realize the short-time electricity consumption prediction of the selected area, and specifically comprises the following steps:
when a regular or irregular area is drawn in a map, calculating a grid completely contained in the closed area, namely all parts of the grid are in the closed area;
examining boundary grids of the selected region, and solving all boundary grids, wherein the area proportion of each grid is in the selected region;
linear superposition summation is carried out on the grid prediction power consumption which is obtained by calculation and completely contains the grid, and Sum is obtained 1 ；
And multiplying the predicted power consumption of the boundary grids by the obtained percentage of the corresponding grids, and adding all the boundary grid data to obtain Sum 2 ；
And obtaining Sum = Sum1+ Sum2 as a short-term prediction result of the electricity consumption of any selected area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211504108.5A CN115829117B (en) | 2022-11-29 | 2022-11-29 | Method and system for predicting electricity consumption in any area in short time based on terminal electricity consumption |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211504108.5A CN115829117B (en) | 2022-11-29 | 2022-11-29 | Method and system for predicting electricity consumption in any area in short time based on terminal electricity consumption |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115829117A true CN115829117A (en) | 2023-03-21 |
| CN115829117B CN115829117B (en) | 2023-09-19 |
Family
ID=85532283
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211504108.5A Active CN115829117B (en) | 2022-11-29 | 2022-11-29 | Method and system for predicting electricity consumption in any area in short time based on terminal electricity consumption |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115829117B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103971296A (en) * | 2014-05-16 | 2014-08-06 | 国家电网公司 | Power purchase method for mathematic model based on electrical loads and temperature |
| CN105260803A (en) * | 2015-11-06 | 2016-01-20 | 国家电网公司 | Power consumption prediction method for system |
| CN106096774A (en) * | 2016-06-07 | 2016-11-09 | 国网山东省电力公司菏泽供电公司 | A kind of large area region Analyzing Total Electricity Consumption Forecasting Methodology based on gridding method |
| CN111210058A (en) * | 2019-12-26 | 2020-05-29 | 深圳供电局有限公司 | Grid-based power distribution network top-down load prediction information method |
| US20200403406A1 (en) * | 2019-06-19 | 2020-12-24 | King Fahd University Of Petroleum And Minerals | Distribution grid fault analysis under load and renewable energy uncertainties |
| CN113505534A (en) * | 2021-07-07 | 2021-10-15 | 南京工程学院 | Load prediction method considering demand response |
-
2022
- 2022-11-29 CN CN202211504108.5A patent/CN115829117B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103971296A (en) * | 2014-05-16 | 2014-08-06 | 国家电网公司 | Power purchase method for mathematic model based on electrical loads and temperature |
| CN105260803A (en) * | 2015-11-06 | 2016-01-20 | 国家电网公司 | Power consumption prediction method for system |
| CN106096774A (en) * | 2016-06-07 | 2016-11-09 | 国网山东省电力公司菏泽供电公司 | A kind of large area region Analyzing Total Electricity Consumption Forecasting Methodology based on gridding method |
| US20200403406A1 (en) * | 2019-06-19 | 2020-12-24 | King Fahd University Of Petroleum And Minerals | Distribution grid fault analysis under load and renewable energy uncertainties |
| CN111210058A (en) * | 2019-12-26 | 2020-05-29 | 深圳供电局有限公司 | Grid-based power distribution network top-down load prediction information method |
| CN113505534A (en) * | 2021-07-07 | 2021-10-15 | 南京工程学院 | Load prediction method considering demand response |
Non-Patent Citations (3)
| Title |
|---|
| T. AHMAD, ET AL.: ""A review on renewable energy and electricity requirement forecasting models for smart grid and buildings"", 《SUSTAINABLE CITIES AND SOCIETY》, vol. 55, pages 1 - 31 * |
| 刘瑾 等: ""基于梯度提升决策树的电力物联网用电负荷预测"", 《智慧电力》, vol. 50, no. 8, pages 46 - 53 * |
| 孙书凯: ""基于资源特征挖掘的风电中长期电量预测方法"", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》, no. 5, pages 042 - 703 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115829117B (en) | 2023-09-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Fordham et al. | PaleoView: a tool for generating continuous climate projections spanning the last 21 000 years at regional and global scales | |
| Zheng et al. | Climate-change impacts on electricity demands at a metropolitan scale: A case study of Guangzhou, China | |
| Goosse et al. | Reconstructing surface temperature changes over the past 600 years using climate model simulations with data assimilation | |
| Prudhomme et al. | Assessing uncertainties in climate change impact analyses on the river flow regimes in the UK. Part 2: future climate | |
| Jansa et al. | MEDEX: a general overview | |
| Melbourne-Thomas et al. | Regional‐scale scenario modeling for coral reefs: a decision support tool to inform management of a complex system | |
| Carter et al. | SILMUSCEN and CLIGEN Users Guide | |
| Hashimoto et al. | Scenario analysis of land-use and ecosystem services of social-ecological landscapes: implications of alternative development pathways under declining population in the Noto Peninsula, Japan | |
| Iizumi et al. | ELPIS-JP: a dataset of local-scale daily climate change scenarios for Japan | |
| Kundzewicz et al. | Climatic change impact on water resources in a systems perspective | |
| Bradshaw et al. | Marsh bird occupancy of wetlands managed for waterfowl in the Midwestern USA | |
| Schuh et al. | Far-field biogenic and anthropogenic emissions as a dominant source of variability in local urban carbon budgets: A global high-resolution model study with implications for satellite remote sensing | |
| Segatto et al. | Modeling the coupled dynamics of stream metabolism and microbial biomass | |
| Seaborn et al. | Drivers of distributions and niches of North American cold‐adapted amphibians: evaluating both climate and land use | |
| Samadi et al. | Comparison of general circulation models: methodology for selecting the best GCM in Kermanshah Synoptic Station, Iran | |
| Shinbrot et al. | Quiahua, the first citizen science rainfall monitoring network in Mexico: Filling critical gaps in rainfall data for evaluating a payment for hydrologic services program | |
| English et al. | A review of the Forest Service's national visitor use monitoring (NVUM) program | |
| CN115829117B (en) | Method and system for predicting electricity consumption in any area in short time based on terminal electricity consumption | |
| Pavan et al. | Downscaling of DEMETER winter seasonal hindcasts over Northern Italy | |
| Singhee et al. | OPRO: Precise emergency preparedness for electric utilities | |
| Thomas | Quaternary climate variation in southern Africa | |
| Putri et al. | Building a Community Adaptive Capacity Model: A Case Study of the Tanjungmas Sub-District in Semarang Municipality | |
| Jones et al. | Generating climate change scenarios at high resolution for impact studies and adaptation: Focus on developing countries | |
| Dhammapala et al. | A Monte Carlo method for summing modeled and background pollutant concentrations | |
| Wu et al. | Area Frame Household Surveys |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |