CN118586688B - Multi-level data processing method and system in smart power grid - Google Patents
Multi-level data processing method and system in smart power grid Download PDFInfo
- Publication number
- CN118586688B CN118586688B CN202411074783.8A CN202411074783A CN118586688B CN 118586688 B CN118586688 B CN 118586688B CN 202411074783 A CN202411074783 A CN 202411074783A CN 118586688 B CN118586688 B CN 118586688B
- Authority
- CN
- China
- Prior art keywords
- value
- electricity consumption
- user
- deviation
- interval
- 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.)
- Active
Links
- 238000003672 processing method Methods 0.000 title claims abstract description 22
- 230000005611 electricity Effects 0.000 claims abstract description 193
- 238000001514 detection method Methods 0.000 claims abstract description 41
- 238000000034 method Methods 0.000 claims abstract description 24
- 238000012545 processing Methods 0.000 claims description 45
- 238000004458 analytical method Methods 0.000 claims description 28
- 238000004364 calculation method Methods 0.000 claims description 26
- 238000012937 correction Methods 0.000 claims description 10
- 238000004088 simulation Methods 0.000 claims description 4
- 238000005516 engineering process Methods 0.000 abstract description 4
- 230000000694 effects Effects 0.000 abstract description 2
- 230000006399 behavior Effects 0.000 description 23
- 238000012216 screening Methods 0.000 description 7
- 230000008859 change Effects 0.000 description 6
- 230000008569 process Effects 0.000 description 4
- 238000012163 sequencing technique Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005587 bubbling Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000001932 seasonal effect Effects 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The application relates to a multi-level data processing method and a system in a smart grid, and relates to the field of power technology; establishing a history interval and a period detection interval on a time axis; generating a virtual comparison interval in the history interval; acquiring user electricity consumption at each time point and determining overall deviation electricity consumption according to the user electricity consumption of the period detection interval and the user electricity consumption of the virtual comparison interval; defining an effective point when the overall deviation electricity consumption is smaller than the reference electricity consumption, and establishing a target time interval according to the effective point; determining a similarity value according to weather comprehensive data of each time point in the opposite standard time interval; and determining a similarity degree value with the maximum value according to the ordering rule, and outputting the user power consumption at a time point corresponding to the similarity degree value as predicted power consumption for subsequent power dispatching. The method has the effect of improving the accuracy of electricity consumption data prediction.
Description
Technical Field
The application relates to the field of power technology, in particular to a multi-level data processing method and system in a smart grid.
Background
The intelligent power grid is a novel power supply system, which uses the technologies of modern communication, computers, automation and the like, and is organically integrated with the traditional power technology, so that the intelligent level of the power grid is greatly improved. There are multiple types of multi-level data in the smart grid, including equipment operation data, power transaction data, user power consumption data, environment monitoring data, fault diagnosis prediction data, economic operation data, etc., and smart operation of the smart grid can be enabled by performing smart analysis on each multi-level data.
In the related art, aiming at user electricity consumption data, the most needed analysis is user electricity consumption habit data, and the electricity consumption of a user in a specific period can be predicted through the analysis of the user electricity consumption habit data, so that the whole electricity consumption of the area where the user is located can be scheduled, and the situation that excessive electricity consumption resources are scheduled and wasted can be reduced while the area where the user is located can better consume electricity.
In the above related art, when analyzing the electricity consumption habit of the user, the current required electricity consumption is generally predicted by the electricity consumption data in the latest time period, but when the analyzed time point is exactly the time point with obvious seasonal change or temperature change, the use condition of the consumer will change obviously, and at this time, the accuracy of the prediction result obtained when the required electricity consumption is predicted by the favorite electricity consumption data in the latest time period is lower, which is not convenient for the effective scheduling processing of the subsequent electricity consumption.
Disclosure of Invention
In order to improve the accuracy of electricity consumption data prediction, the application provides a multi-level data processing method and system in a smart grid.
In a first aspect, the present application provides a multi-level data processing method in a smart grid, which adopts the following technical scheme:
a multi-level data processing method in a smart grid, comprising:
acquiring current weather comprehensive data and a demand prediction time point;
establishing a historical interval with a preset standard interval duration and a preset width being a historical duration between a rear endpoint and a demand prediction time point on a preset time axis, and establishing a period detection interval with the preset period duration by taking the demand prediction time point as the rear endpoint;
generating a virtual point with changeable positions in the history interval, and establishing a virtual comparison interval with the width consistent with the period detection interval by taking the virtual point as a rear endpoint;
Acquiring the user electricity consumption at each time point, and performing one-to-one comparison according to the user electricity consumption of the period detection interval and the user electricity consumption of the virtual comparison interval to determine the overall deviation electricity consumption;
When the total deviation electricity consumption is smaller than the preset reference electricity consumption, defining a corresponding virtual point as an effective point, and establishing a target matching time interval according to the effective point and a demand prediction time point;
comparing the weather comprehensive data according to each time point and the current weather comprehensive data in the target time interval to determine a similarity value;
And determining a similarity degree value with the largest numerical value according to a preset ordering rule, and outputting the user electricity consumption at a time point corresponding to the similarity degree value as predicted electricity consumption for subsequent electricity dispatching.
Optionally, the step of establishing the target time interval according to the effective point and the demand prediction time point includes:
Combining according to different effective points forward in sequence by using the last effective point on the time axis to determine effective combination;
Determining a combination duration span according to two effective points which are farthest apart in the effective combination, and counting according to the effective points in the effective combination to determine the effective number of the combination;
calculating according to the combined duration span and the combined effective quantity to determine effective distribution density;
defining the effective distribution density determined by the first effective combination as a reference density, and calculating according to the reference density and a preset allowable error proportion to determine a qualified density range;
defining valid points in the valid combinations as allowed points when the valid distribution density is within the qualified density range;
Determining a permission interval according to the continuous and uninterrupted permission points by taking the last permission point on the time axis as a starting point, and defining the permission point farthest from the last permission point in the permission interval as a boundary point;
And predicting time points according to the boundary points and the requirements to serve as two endpoints so as to establish a target time interval.
Optionally, the step of comparing the weather integrated data according to each time point and the current weather integrated data in the target time interval to determine the similarity value includes:
acquiring the weather data type of each weather comprehensive data;
Determining a data fitting value corresponding to weather comprehensive data under a weather data type according to a preset fitting matching relationship;
determining a calculated weight value corresponding to the weather data type according to a preset weight matching relation;
performing difference calculation according to the data fitting values of the same weather data type to determine a data fitting difference value;
and calculating according to all the data simulation difference values and the corresponding calculated weight values to determine the similarity degree value.
Optionally, after the similarity value is determined, the multi-level data processing method further includes:
Judging whether at least two time points with the same similarity value and the maximum similarity value exist or not;
If at least two time points with the same similarity value and the maximum similarity value do not exist, determining and outputting predicted power consumption according to the power consumption of the user corresponding to the time point;
If at least two time points with the same similarity value and the maximum similarity value exist, defining the corresponding time point as an alternative point, and determining the overall deviation electricity consumption according to each alternative point in the opposite standard time interval;
Determining the overall deviation electricity consumption with the smallest value in the alternative points according to the ordering rule, and defining the user electricity consumption corresponding to the alternative point corresponding to the overall deviation electricity consumption as the standard electricity consumption;
determining a standard electricity consumption range according to the standard electricity consumption and a preset permission deviation value, and defining the user electricity consumption in the standard electricity consumption range in the alternative point as effective electricity consumption;
and carrying out average value calculation according to all the effective electric quantity to determine and output the predicted electric quantity.
Optionally, after the definition of the valid point is completed, the multi-level data processing method in the smart grid further includes:
Acquiring the same users in the same area when no effective point exists;
in the period detection interval, performing one comparison according to the user electricity consumption of the same user in the area and the user electricity consumption of the current analysis user to determine a user similarity deviation value;
determining a user similarity deviation value with the smallest value according to the ordering rule, and defining the same user in the region corresponding to the user similarity deviation value as a behavior similar user;
Determining a corresponding similarity value according to the behavior similarity user, and judging whether the similarity value is larger than a preset reference requirement value or not;
if the similarity value is larger than the reference requirement value, the user electricity consumption at the corresponding time point of the behavior similar user is used as the predicted electricity consumption of the current analysis user and is output;
And if the similarity value is not greater than the reference demand value, re-determining the behavior similar users in the same users in the remaining areas according to the user similarity deviation value until the predicted power consumption is determined or the user similarity deviation value is greater than a preset invalid deviation value.
Optionally, after the user similarity deviation value is determined, the multi-level data processing method in the smart grid further includes:
Judging whether at least two users with the same similarity deviation value and the smallest area are the same;
If at least two users with the same user similarity deviation value and the smallest area are not present, determining the similar users according to the same users in the area;
If at least two users with the same user similarity deviation value and the same minimum area exist, selecting a comparison time point in a period detection interval according to the preset comparison quantity;
performing a comparison according to the user electricity consumption at the comparison time point to determine a comparison deviation value;
Determining a comparison deviation value with the smallest value according to the ordering rule, and taking the comparison deviation value as a representative deviation value of the users with the same area;
And determining the representing deviation value with the smallest value according to the ordering rule, and determining the same user in the region corresponding to the representing deviation value as the similar user.
Optionally, after predicting the power consumption output, the multi-level data processing method in the smart grid further includes:
Calculating a difference value in the period detection interval according to the user electricity consumption at each time point and the corresponding predicted electricity consumption to determine a predicted deviation value;
calculating the average value according to the absolute value of each predicted deviation value to determine an average deviation value, and calculating according to the absolute value of each predicted deviation value and the average deviation value to determine a predicted overall deviation;
judging whether the predicted integral deviation is smaller than a preset effective qualified deviation or not;
If the predicted integral deviation is smaller than the effective qualified deviation, carrying out mean value calculation according to all the predicted deviation values to determine a predicted correction value, and updating the predicted power consumption according to the predicted correction value;
if the predicted integral deviation is not smaller than the effective qualified deviation, determining a predicted signal corresponding to the average deviation value according to a preset predicted matching relation, and synchronously outputting the predicted signal and the predicted power consumption.
In a second aspect, the present application provides a multi-level data processing system in a smart grid, which adopts the following technical scheme:
A multi-level data processing system in a smart grid, comprising:
The acquisition module is used for acquiring current weather comprehensive data and a demand prediction time point;
the processing module is connected with the acquisition module and the judging module and is used for storing and processing information;
the judging module is connected with the acquisition module and the processing module and is used for judging information;
the processing module establishes a historical interval with a preset standard interval duration and a width of the historical duration between a rear endpoint and a demand prediction time point on a preset time axis, and establishes a period detection interval with the width of the preset period duration by taking the demand prediction time point as the rear endpoint;
The processing module generates a virtual point with changeable positions in the history interval, and establishes a virtual comparison interval with the width consistent with the period detection interval by taking the virtual point as a rear endpoint;
the acquisition module acquires the user electricity consumption at each time point and enables the processing module to conduct one-to-one comparison according to the user electricity consumption of the period detection interval and the user electricity consumption of the virtual comparison interval so as to determine the overall deviation electricity consumption;
When the judging module judges that the overall deviation electricity consumption is smaller than the preset reference electricity consumption, the processing module defines the corresponding virtual point as an effective point, and establishes a standard time interval according to the effective point and the demand prediction time point;
the processing module compares the weather comprehensive data according to each time point and the current weather comprehensive data in the target time interval to determine a similarity value;
the processing module determines a similarity value with the largest value according to a preset ordering rule, and outputs the user electricity consumption at a time point corresponding to the similarity value as predicted electricity consumption for subsequent electricity dispatching.
In summary, the present application includes at least one of the following beneficial technical effects:
when the electricity consumption is predicted under the condition of obvious weather change, the time period of an actual user can be determined, and then the electricity consumption is predicted by comparison according to specific weather conditions;
Aiming at the situation that the current electricity consumption cannot be predicted by the historical data of a single user, the electricity consumption of the current user can be predicted by combining the specific situations of other users in the area where the user is located, so that the effective prediction of the electricity consumption can be realized.
Drawings
Fig. 1 is a flowchart of a multi-level data processing method in a smart grid.
Fig. 2 is a flow chart of a method of establishing a time interval of a beacon.
Fig. 3 is a flowchart of a similarity value determination method.
Fig. 4 is a flow chart of a time point screening method.
Fig. 5 is a flow chart of a method of behavior similar to user determination.
Fig. 6 is a flow chart of a method of regional identical user screening.
Fig. 7 is a flowchart of a predicted power consumption correction method.
Fig. 8 is a block flow diagram of a multi-level data processing method in a smart grid.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application will be further described in detail with reference to fig. 1 to 8 and the embodiments. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the application.
Embodiments of the application are described in further detail below with reference to the drawings.
The embodiment of the application discloses a multi-level data processing method in a smart grid, which comprises the steps of determining an actual time period generated by a currently resident user according to historical data under a user name when predicting the power consumption of the user, searching power consumption data under the same weather condition in the time point, and predicting according to the power consumption data so as to better predict the power consumption of the user under the condition that the weather change is obvious.
Referring to fig. 1, a method flow of a multi-level data processing method in a smart grid includes the following steps:
step S100: and acquiring current weather comprehensive data and a demand prediction time point.
The weather comprehensive data comprise weather humidity, temperature, wind direction, air pressure, seasons and the like which can reflect specific weather conditions, and the weather comprehensive data can be obtained through prediction by a weather prediction system; the demand prediction time point is a time point when the user power consumption needs to be predicted, and generally, the time point is specific to a date, for example, the day of 11 months and 11 days is predicted.
Step S101: and establishing a historical interval with a preset standard interval duration and a preset width as a historical duration between the rear end point and the demand prediction time point on a preset time axis, and establishing a period detection interval with the preset period duration by taking the demand prediction time point as the rear end point.
The time axis is a coordinate axis formed by combining time points, the time axis points to a time point which is not reached yet from the time point which is already passed, the standard interval duration is the duration which is set by a worker and is used for the interval when the data acquisition is carried out on the historical behavior habit of the user, for example, 2 months, the historical duration is the duration which is set by the worker and can be used for detecting the historical behavior habit of the user, for example, 1 year, and the historical data can be effectively acquired and analyzed by establishing a historical interval; the period duration is a constant value duration set by a worker, for example, the period detection interval is established on seven days of a week, and electricity consumption data in the period duration can be acquired, so that subsequent analysis is facilitated.
Step S102: and generating a virtual point with changeable positions in the history interval, and establishing a virtual comparison interval with the width consistent with the period detection interval by taking the virtual point as a rear endpoint.
The virtual point is a time point in the history interval, and the establishment of the virtual comparison interval can facilitate the acquisition of data in the period duration in the history process and facilitate the subsequent analysis.
Step S103: and obtaining the user electricity consumption at each time point, and performing one-to-one comparison according to the user electricity consumption of the period detection interval and the user electricity consumption of the virtual comparison interval to determine the overall deviation electricity consumption.
The user power consumption is the power consumption value used by the current day below the corresponding time point, the overall deviation power consumption is the deviation value obtained by performing one comparison on the user power consumption in two intervals, and the calculation formula is thatWherein Y is the total deviation electricity consumption, I 1n is the electricity consumption of the user at the nth time point in the period detection interval, I 2n is the electricity consumption of the user at the nth time point in the virtual comparison interval, a comparison mode is that the first time point in the period detection interval is compared with the first time point in the virtual comparison interval, and so on.
Step S104: and when the total deviation electricity consumption is smaller than the preset reference electricity consumption, defining the corresponding virtual point as an effective point, and establishing a target matching time interval according to the effective point and the demand prediction time point.
The reference electricity consumption is the maximum overall deviation electricity consumption required to be met when the electricity consumption behaviors of the two intervals are determined to be similar, which is set by a worker, and when the overall deviation electricity consumption is smaller than the reference electricity consumption, the virtual comparison interval is indicated to be similar to the current period detection interval in electricity consumption, and corresponding virtual points are defined as effective points at the moment so as to realize the distinction of different virtual points, so that the follow-up analysis is convenient; the target time interval is an interval established by taking the effective point and the demand prediction time point as two endpoints, when a plurality of effective points exist, one effective point can be taken to determine the target time interval, and the target time interval can also be determined according to the methods in the steps S200-S206, which are not repeated herein.
Step S105: and comparing the weather comprehensive data at each time point with the current weather comprehensive data in the target time interval to determine the similarity value.
The similarity value is a value reflecting the similarity of the weather conditions at two time points, and the larger the value is, the more similar the weather conditions at two time points are, and the specific calculation method can be illustrated by step S300-step S304.
Step S106: and determining a similarity degree value with the largest numerical value according to a preset ordering rule, and outputting the user electricity consumption at a time point corresponding to the similarity degree value as predicted electricity consumption for subsequent electricity dispatching.
The sequencing rule is a method which is set by staff and can sequence the values, such as an bubbling method, and the similarity value with the largest value can be determined through the sequencing rule, namely, the time point corresponding to the similarity value is most similar to the electricity consumption behavior of the current demand prediction time point, and the electricity consumption of the user at the time point is output as the predicted electricity consumption so as to realize effective prediction of the electricity consumption condition.
Referring to fig. 2, the step of establishing a benchmarking time interval according to the effective point and the demand forecast time point includes:
step S200: and combining according to different effective points sequentially forward by using the last effective point on the time axis to determine effective combination.
The effective combination is a combination formed by combining all effective points, for example, three effective points are ABC from front to back on a time axis, and the corresponding effective combination comprises CB and CBA.
Step S201: the combination duration span is determined in the effective combination according to the two effective points which are furthest apart, and the effective number of the combination is determined by counting according to the effective points in the effective combination.
The combination duration span is the interval duration between two effective points which are furthest apart on a time axis in the effective combination, and taking the effective combination ABC as an example, the corresponding combination duration span is the duration interval between a time point A and a time point C; the combined effective amount is the total number of effective points in the effective combination.
Step S202: and calculating according to the combined duration span and the combined effective quantity to determine the effective distribution density.
The effective distribution density is a value obtained by dividing the combination duration span by the combination effective number.
Step S203: the effective distribution density determined by the first effective combination is defined as a reference density, and is calculated according to the reference density and a preset allowable error proportion to determine a qualified density range.
Defining a reference concentration degree to realize the distinction of different effective distribution concentration degrees, and facilitating the subsequent analysis, wherein a first effective combination is a first determined effective combination, the effective point is taken as ABC for example, and the corresponding effective combination is BC; the allowable error ratio is the ratio of the allowable density change when the rated density set by the staff is not changed greatly, generally 30% -50%, the corresponding value can be obtained by multiplying the allowable error ratio by the reference density, and the two end point values can be obtained by respectively adding the value to the reference density and subtracting the value from the reference density, wherein the range of the interval determined by the two end point values is the qualified density range.
Step S204: valid points within the valid combinations are defined as allowed points when the valid distribution densities are within the acceptable density range.
When the effective distribution density is in the qualified density range, the determined effective points are basically obtained by the same actual user under the same user name, and the corresponding effective points are defined as permission points for identification, so that the subsequent analysis is convenient.
Step S205: and determining a permission interval according to the continuous and uninterrupted permission points by taking the last permission point on the time axis as a starting point, and defining the permission point farthest from the last permission point in the permission interval as a boundary point.
The permission interval is the maximum interval when the effective points are taken as endpoints and all the effective points in the interval are permission points, and boundary points are defined to realize the distinction of different permission points, so that the subsequent analysis is convenient.
Step S206: and predicting time points according to the boundary points and the requirements to serve as two endpoints so as to establish a target time interval.
The effective points obtained by the same actual user can be determined by utilizing the boundary points to determine the target time interval, and the effective points determined before the boundary points are possibly not obtained by the current actual householder, so that the determined target time interval is more accurate.
Referring to fig. 3, the step of comparing weather integrated data according to each time point and current weather integrated data in the target time interval to determine the similarity value includes:
step S300: and acquiring the weather data type of each weather comprehensive data.
The weather data type is the data type of the weather integrated data, such as air temperature, season, and the like.
Step S301: and determining a data fitting value corresponding to the weather comprehensive data under the weather data type according to a preset fitting matching relationship.
The data fitting values are fitting values for calculating the similarity values corresponding to the weather comprehensive data under the weather data types, and different weather data types and different data fitting values corresponding to the weather comprehensive data, so that fitting matching relations are established in advance by staff, data which do not contain values such as seasons can be subjected to data fitting processing, and the calculation of the similarity values is facilitated.
Step S302: and determining a calculated weight value corresponding to the weather data type according to a preset weight matching relation.
The calculation weight value is a calculation parameter when the similarity value is calculated by various types of data, and because the influence conditions of the different types of data on the electricity consumption are different, the corresponding calculation weight values are also different, for example, the influence of the air temperature on the electricity consumption is far greater than the influence of haze on the electricity consumption, the calculation weight value corresponding to the air temperature is large, and the weight matching relationship between the two is established and recorded in advance by staff.
Step S303: and carrying out difference calculation according to the data fitting values of the same weather data type to determine the data fitting difference.
The data fitting difference is the difference of the data fitting values of the same weather data type, and the difference is an absolute value.
Step S304: and calculating according to all the data simulation difference values and the corresponding calculated weight values to determine the similarity degree value.
The calculation formula is thatWherein X is a similarity value, alpha is a fixed parameter used for calculation, tau n is a calculation weight value corresponding to nth class data, and M n is a data simulation difference value corresponding to nth class data.
Referring to fig. 4, after the similarity value is determined, the multi-level data processing method further includes:
step S400: and judging whether at least two time points with the same similarity value and the maximum similarity value exist.
The purpose of the judgment is to know whether a plurality of time points meeting the requirements exist or not so as to facilitate the subsequent determination of the predicted electricity consumption.
Step S4001: if at least two time points with the same similarity value and the maximum similarity value do not exist, the predicted power consumption is determined and output according to the power consumption of the user corresponding to the time point.
When at least two time points with the same similarity value and the maximum similarity value do not exist, only a unique time point meeting the requirements is indicated, and the predicted electricity consumption is determined according to the electricity consumption of the user at the corresponding time point.
Step S4002: if at least two time points with the same similarity value and the maximum similarity value exist, the corresponding time points are defined as alternative points, and the overall deviation electricity consumption is determined according to each alternative point in the opposite standard time interval.
When at least two time points with the same similarity value and the maximum similarity value exist, a plurality of time points meeting the requirements are indicated to exist, and further screening processing is needed for the time points; alternative points are defined to enable differentiation of different points in time.
Step S401: and determining the overall deviation electricity consumption with the smallest value in the alternative points according to the ordering rule, and defining the user electricity consumption corresponding to the alternative point corresponding to the overall deviation electricity consumption as the standard electricity consumption.
The overall deviation electricity consumption with the smallest value can be determined through the ordering rule, namely the electricity consumption habit of the alternative point is closest to that of the current demand prediction time point, and the corresponding electricity consumption of the user is defined as the standard electricity consumption at the moment so as to facilitate subsequent analysis.
Step S402: and determining a standard electricity consumption range according to the standard electricity consumption and a preset permission deviation value, and defining the electricity consumption of the user in the standard electricity consumption range in the alternative point as effective electricity consumption.
The allowable deviation value is the maximum allowable deviation value when the identification set by the staff is relatively similar to the integral electricity consumption condition of the standard electricity consumption, and two endpoints of the standard electricity consumption range are respectively the standard electricity consumption minus the allowable deviation value and the standard electricity consumption plus the allowable deviation value; the user electricity consumption in the standard electricity consumption range is defined as effective electricity consumption so as to distinguish different user electricity consumption, and subsequent analysis is facilitated.
Step S403: and carrying out average value calculation according to all the effective electric quantity to determine and output the predicted electric quantity.
And calculating the average value by using the effective electric quantity to determine the predicted electric quantity, so that the determined predicted electric quantity is accurate.
Referring to fig. 5, after the definition of the valid point is completed, the multi-level data processing method in the smart grid further includes:
step S500: and acquiring the same users in the same area when no effective point exists.
When no effective point exists, the fact that the specific situation of the actual resident under the user name cannot be determined is indicated, and at the moment, if weather similarity value judgment is carried out in the historical data of the user currently carrying out analysis, the situation that electricity consumption prediction is inaccurate is likely to occur, so that further analysis is needed; the users with the same area are other users in the same area with the users currently carrying out electricity consumption predictive analysis, wherein the specific range of the area is set by staff according to actual conditions, such as the same cell, the same street, the same city and the like.
Step S501: and in the period detection interval, performing a comparison according to the user power consumption of the users with the same area and the user power consumption of the current analysis user to determine the user similarity deviation value.
The user similarity deviation value is a deviation value of electricity consumption habits of two users in the same time period, the larger the value is, the more different the electricity consumption habits of the two users are, and a specific calculation method is consistent with the overall deviation electricity consumption, and is not repeated here.
Step S502: and determining a user similarity deviation value with the smallest value according to the ordering rule, and defining the same user of the region corresponding to the user similarity deviation value as a behavior similar user.
The user similarity deviation value with the smallest value can be determined through the ordering rule, namely the power consumption habit of the same user in the region corresponding to the user similarity deviation value is closest to the power consumption habit of the current analysis user, and the user similarity deviation value is defined as the behavior similar user at the moment so as to realize the distinction of the same user in different regions, and the subsequent analysis is convenient.
Step S503: and determining a corresponding similarity value according to the behavior similarity user, and judging whether the similarity value is larger than a preset reference requirement value or not.
The reference requirement value is a minimum similarity value which is set by staff and is required to be met when weather is compared and is considered to be similar when the weather is compared, and the purpose of judgment is to know whether the weather experienced by the behavior similar user comprises the weather of the current required predicted electricity consumption.
Step S504: and if the similarity value is larger than the reference requirement value, the user electricity consumption at the corresponding time point of the similar user is used as the predicted electricity consumption of the current analysis user and is output.
When the similarity value is larger than the reference demand value, the weather meeting the demand is indicated, and the user electricity consumption at the time point of the weather meeting the demand corresponding to the behavior similar user is output as the predicted electricity consumption at the time point, so that the reasonable prediction of the electricity consumption of the current analysis user at the demand prediction time point is realized.
Step S500: and if the similarity value is not greater than the reference demand value, re-determining the behavior similar users in the same users in the remaining areas according to the user similarity deviation value until the predicted power consumption is determined or the user similarity deviation value is greater than a preset invalid deviation value.
When the similarity value is not greater than the reference demand value, the behavior similar user is determined to be similar to the current weather condition in history, and the behavior similar user is determined again in the rest users until the predicted electricity consumption can be reasonably predicted and output or the user similarity deviation value is greater than the invalid deviation value, wherein the invalid deviation value is the user similarity deviation value which is set by a staff and is the smallest when the difference of electricity consumption habits of the two users is greater, and when the user similarity deviation value is greater than the invalid deviation value, the fact that the rest users do not have people with electricity consumption habits similar to that of the current user is determined, and at the moment, an unpredictable signal can be output to a management platform, so that the staff can perform manual intervention processing.
Referring to fig. 6, after the user similarity deviation value is determined, the multi-level data processing method in the smart grid further includes:
Step S600: and judging whether at least two users with the same similarity deviation value and the same smallest area exist.
The purpose of the judgment is to know whether a plurality of users meeting the requirement exist in the same area or not, so that when a plurality of users meeting the requirement exist in the same area, the users can be used as similar users in behavior, and further screening processing is performed on the users.
Step S6001: and if at least two users with the same user similarity deviation value and the smallest area are not present, determining the similar users according to the area.
When at least two users with the same user similarity deviation value and the smallest region are not present, only the only region with the same user meeting the requirements is indicated, and the user is determined to be the similar user.
Step S6002: if at least two users with the same user similarity deviation value and the same minimum area exist, selecting a comparison time point in the period detection interval according to the preset comparison quantity.
When at least two users with the same user similarity deviation value and the smallest region are the same, the users with the same region are required to be screened; the comparison number is the fixed value number set by the staff, generally the number of time points in the period detection interval minus one, for example, 3 time points exist in the period detection interval, and the comparison number is 2; the comparison time points, i.e. the time points selected in the period detection interval, i.e. the time points to be analyzed, e.g. the 3 time points in the period detection interval are J, K, L, can be selected by the colleague as the combination of comparison time points including JK, JL, KL.
Step S601: and performing a comparison according to the user electricity consumption at the comparison time point to determine a comparison deviation value.
The comparison deviation value is a result value obtained by comparing the same user in the area with the comparison time point in the current analysis user and the period detection interval one by one, the value reflects the electricity habit difference of the two users at the comparison time point, the larger the value is, the more different the electricity habits of the two users are, and the specific calculation method is consistent with the overall deviation electricity consumption, and is not described in detail herein.
Step S602: and determining a comparison deviation value with the smallest value according to the ordering rule, and taking the comparison deviation value as a representative deviation value of the users with the same area.
And determining a comparison deviation value with the smallest value through the ordering rule, namely, explaining that the comparison time point determined under the comparison deviation value is closest to the electricity habit of the current analysis user, and defining the comparison time point as a representative deviation value at the moment to represent the electricity habit deviation of the same user in the area and the current analysis user.
Step S603: and determining the representing deviation value with the smallest value according to the ordering rule, and determining the same user in the region corresponding to the representing deviation value as the similar user.
The representative deviation value with the smallest value can be determined through the sorting rule, namely, the user with the same area corresponding to the representative deviation value is closest to the current electricity habit of the analysis user, and the user is determined to be the similar user in behavior at the moment, wherein when a plurality of users with the same and smallest representative deviation value still exist, one subtracting process can be carried out on the comparison quantity to redetermine the representative deviation value until the most suitable similar user in behavior is determined.
Referring to fig. 7, after predicting the power consumption output, the multi-level data processing method in the smart grid further includes:
Step S700: and in the period detection interval, calculating a difference value according to the user electricity consumption at each time point and the corresponding predicted electricity consumption to determine a predicted deviation value.
The predicted deviation value is the difference between the predicted electricity consumption and the actually generated electricity consumption of the user at each time point in the period detection interval, and the difference is a relative value and is determined by subtracting the predicted electricity consumption from the electricity consumption of the user.
Step S701: and carrying out average value calculation according to the absolute value of each predicted deviation value to determine an average deviation value, and carrying out calculation according to the absolute value of each predicted deviation value and the average deviation value to determine the predicted overall deviation.
The average deviation value is the average value of the absolute values of the predicted deviation values, the predicted overall deviation is a parameter reflecting whether the deviation between the predicted amount and the actual amount is stable, the smaller the value is, the more stable the deviation is, and the calculation formula isWherein Q is the predicted global bias, R n is the nth predicted bias value,Is the average deviation value.
Step S702: and judging whether the predicted integral deviation is smaller than a preset effective qualified deviation or not.
The effective qualified deviation is the maximum predicted overall deviation which is required to be met when the deviation between the predicted quantity and the actual quantity is stable and is set by a worker, and the purpose of judgment is to know whether the deviation between the predicted quantity and the actual quantity is stable.
Step S7021: if the predicted integral deviation is smaller than the effective qualified deviation, carrying out mean value calculation according to all the predicted deviation values to determine a predicted correction value, and updating the predicted power consumption according to the predicted correction value.
When the predicted integral deviation is smaller than the effective qualified deviation, the description deviation is stable, at the moment, the predicted correction value can be obtained by means of average calculation of all the predicted deviation values, and the predicted power consumption can be updated by means of the predicted power consumption and the predicted correction value, so that the accuracy of the determined predicted power consumption is further improved.
Step S7022: if the predicted integral deviation is not smaller than the effective qualified deviation, determining a predicted signal corresponding to the average deviation value according to a preset predicted matching relation, and synchronously outputting the predicted signal and the predicted power consumption.
When the predicted integral deviation is not less than the effective qualified deviation, the predicted deviation is unstable, and a predicted signal can be output to identify the situation, wherein the predicted signal comprises a signal with smaller deviation, a signal with moderate deviation and a signal with larger deviation, the predicted signals corresponding to different average deviation values are different, the predicted matching relationship between the two is recorded in advance by a worker, and the manager can know the predicted integral situation of the electricity consumption by synchronously outputting the predicted signal and the predicted electricity consumption, so that the follow-up processing is convenient.
Referring to fig. 8, based on the same inventive concept, an embodiment of the present invention provides a multi-level data processing system in a smart grid, including:
The acquisition module is used for acquiring current weather comprehensive data and a demand prediction time point;
the processing module is connected with the acquisition module and the judging module and is used for storing and processing information;
the judging module is connected with the acquisition module and the processing module and is used for judging information;
the processing module establishes a historical interval with a preset standard interval duration and a width of the historical duration between a rear endpoint and a demand prediction time point on a preset time axis, and establishes a period detection interval with the width of the preset period duration by taking the demand prediction time point as the rear endpoint;
The processing module generates a virtual point with changeable positions in the history interval, and establishes a virtual comparison interval with the width consistent with the period detection interval by taking the virtual point as a rear endpoint;
the acquisition module acquires the user electricity consumption at each time point and enables the processing module to conduct one-to-one comparison according to the user electricity consumption of the period detection interval and the user electricity consumption of the virtual comparison interval so as to determine the overall deviation electricity consumption;
When the judging module judges that the overall deviation electricity consumption is smaller than the preset reference electricity consumption, the processing module defines the corresponding virtual point as an effective point, and establishes a standard time interval according to the effective point and the demand prediction time point;
the processing module compares the weather comprehensive data according to each time point and the current weather comprehensive data in the target time interval to determine a similarity value;
The processing module determines a similarity degree value with the largest numerical value according to a preset ordering rule, and outputs the user electricity consumption at a time point corresponding to the similarity degree value as predicted electricity consumption for subsequent electricity dispatching;
the target time interval establishing module is used for establishing a proper target time interval;
The similarity value determining module is used for calculating and determining a similarity value according to different weather comprehensive data;
the time point screening module is used for screening the unique time point under the condition that the similarity value is consistent and the maximum;
the behavior similar user determining module is used for introducing the determination of the behavior similar user aiming at the situation that the power consumption cannot be predicted by the data of the user, so that the power consumption can be predicted conveniently;
the regional identical user screening module screens a plurality of regional identical users meeting the requirements to determine unique behavior similar users to carry out subsequent calculation analysis;
and the prediction correction module is used for correcting the current predicted electricity consumption according to the historical prediction condition.
It will be apparent to those skilled in the art that, for convenience and brevity of description, only the above-described division of the functional modules is illustrated, and in practical application, the above-described functional allocation may be performed by different functional modules according to needs, i.e. the internal structure of the apparatus is divided into different functional modules to perform all or part of the functions described above. The specific working processes of the above-described systems, devices and units may refer to the corresponding processes in the foregoing method embodiments, which are not described herein.
Claims (8)
1. A multi-level data processing method in a smart grid, comprising:
acquiring current weather comprehensive data and a demand prediction time point;
establishing a historical interval with a preset standard interval duration and a preset width being a historical duration between a rear endpoint and a demand prediction time point on a preset time axis, and establishing a period detection interval with the preset period duration by taking the demand prediction time point as the rear endpoint;
generating a virtual point with changeable positions in the history interval, and establishing a virtual comparison interval with the width consistent with the period detection interval by taking the virtual point as a rear endpoint;
obtaining the user electricity consumption at each time point and performing one comparison according to the user electricity consumption of the period detection interval and the user electricity consumption of the virtual comparison interval to determine the overall deviation electricity consumption, wherein Y is defined as the overall deviation electricity consumption, I 1n is the user electricity consumption at the nth time point in the period detection interval, I 2n is the user electricity consumption at the nth time point in the virtual comparison interval, ；
When the total deviation electricity consumption is smaller than the preset reference electricity consumption, defining a corresponding virtual point as an effective point, and establishing a target matching time interval according to the effective point and a demand prediction time point;
comparing the weather comprehensive data according to each time point and the current weather comprehensive data in the target time interval to determine a similarity value;
And determining a similarity degree value with the largest numerical value according to a preset ordering rule, and outputting the user electricity consumption at a time point corresponding to the similarity degree value as predicted electricity consumption for subsequent electricity dispatching.
2. The method for multi-level data processing in a smart grid according to claim 1, wherein the step of establishing the benchmarking time interval according to the valid point and the demand forecast time point comprises:
Combining according to different effective points forward in sequence by using the last effective point on the time axis to determine effective combination;
Determining a combination duration span according to two effective points which are farthest apart in the effective combination, and counting according to the effective points in the effective combination to determine the effective number of the combination;
calculating according to the combined duration span and the combined effective quantity to determine effective distribution density;
defining the effective distribution density determined by the first effective combination as a reference density, and calculating according to the reference density and a preset allowable error proportion to determine a qualified density range;
defining valid points in the valid combinations as allowed points when the valid distribution density is within the qualified density range;
Determining a permission interval according to the continuous and uninterrupted permission points by taking the last permission point on the time axis as a starting point, and defining the permission point farthest from the last permission point in the permission interval as a boundary point;
And predicting time points according to the boundary points and the requirements to serve as two endpoints so as to establish a target time interval.
3. The method for multi-level data processing in a smart grid according to claim 1, wherein the step of comparing the weather integrated data according to each time point and the current weather integrated data in the standard time interval to determine the similarity value comprises:
acquiring the weather data type of each weather comprehensive data;
determining a data fitting value corresponding to the weather comprehensive data under the weather data type according to a preset fitting matching relationship, wherein the data fitting value is a fitting value corresponding to the weather comprehensive data under the weather data type and used for calculating a similarity value;
determining a calculated weight value corresponding to the weather data type according to a preset weight matching relation;
performing difference calculation according to the data fitting values of the same weather data type to determine a data fitting difference value;
and calculating according to all the data simulation difference values and the corresponding calculated weight values to determine the similarity degree value.
4. The multi-level data processing method in a smart grid according to claim 3, wherein after the similarity value is determined, the multi-level data processing method further comprises:
Judging whether at least two time points with the same similarity value and the maximum similarity value exist or not;
If at least two time points with the same similarity value and the maximum similarity value do not exist, determining and outputting predicted power consumption according to the power consumption of the user corresponding to the time point;
If at least two time points with the same similarity value and the maximum similarity value exist, defining the corresponding time point as an alternative point, and determining the overall deviation electricity consumption according to each alternative point in the opposite standard time interval;
Determining the overall deviation electricity consumption with the smallest value in the alternative points according to the ordering rule, and defining the user electricity consumption corresponding to the alternative point corresponding to the overall deviation electricity consumption as the standard electricity consumption;
determining a standard electricity consumption range according to the standard electricity consumption and a preset permission deviation value, and defining the user electricity consumption in the standard electricity consumption range in the alternative point as effective electricity consumption;
and carrying out average value calculation according to all the effective electric quantity to determine and output the predicted electric quantity.
5. The method for multi-level data processing in a smart grid according to claim 1, wherein after the definition of the valid point is completed, the method for multi-level data processing in a smart grid further comprises:
Acquiring the same users in the same area when no effective point exists;
in the period detection interval, performing one comparison according to the user electricity consumption of the same user in the area and the user electricity consumption of the current analysis user to determine a user similarity deviation value;
determining a user similarity deviation value with the smallest value according to the ordering rule, and defining the same user in the region corresponding to the user similarity deviation value as a behavior similar user;
Determining a corresponding similarity value according to the behavior similarity user, and judging whether the similarity value is larger than a preset reference requirement value or not;
if the similarity value is larger than the reference requirement value, the user electricity consumption at the corresponding time point of the behavior similar user is used as the predicted electricity consumption of the current analysis user and is output;
And if the similarity value is not greater than the reference demand value, re-determining the behavior similar users in the same users in the remaining areas according to the user similarity deviation value until the predicted power consumption is determined or the user similarity deviation value is greater than a preset invalid deviation value.
6. The method for multi-level data processing in a smart grid according to claim 5, wherein after determining the user similarity bias value, the method for multi-level data processing in a smart grid further comprises:
Judging whether at least two users with the same similarity deviation value and the smallest area are the same;
If at least two users with the same user similarity deviation value and the smallest area are not present, determining the similar users according to the same users in the area;
If at least two users with the same user similarity deviation value and the same minimum area exist, selecting a comparison time point in a period detection interval according to the preset comparison quantity;
performing a comparison according to the user electricity consumption at the comparison time point to determine a comparison deviation value;
Determining a comparison deviation value with the smallest value according to the ordering rule, and taking the comparison deviation value as a representative deviation value of the users with the same area;
And determining the representing deviation value with the smallest value according to the ordering rule, and determining the same user in the region corresponding to the representing deviation value as the similar user.
7. The method for multi-level data processing in a smart grid according to claim 1, wherein after predicting the power consumption output, the method for multi-level data processing in a smart grid further comprises:
Calculating a difference value in the period detection interval according to the user electricity consumption at each time point and the corresponding predicted electricity consumption to determine a predicted deviation value;
calculating the average value according to the absolute value of each predicted deviation value to determine an average deviation value, and calculating according to the absolute value of each predicted deviation value and the average deviation value to determine a predicted overall deviation;
judging whether the predicted integral deviation is smaller than a preset effective qualified deviation or not;
If the predicted integral deviation is smaller than the effective qualified deviation, carrying out mean value calculation according to all the predicted deviation values to determine a predicted correction value, and updating the predicted power consumption according to the predicted correction value;
if the predicted integral deviation is not smaller than the effective qualified deviation, determining a predicted signal corresponding to the average deviation value according to a preset predicted matching relation, and synchronously outputting the predicted signal and the predicted power consumption.
8. A multi-level data processing system in a smart grid, comprising:
The acquisition module is used for acquiring current weather comprehensive data and a demand prediction time point;
the processing module is connected with the acquisition module and the judging module and is used for storing and processing information;
the judging module is connected with the acquisition module and the processing module and is used for judging information;
the processing module establishes a historical interval with a preset standard interval duration and a width of the historical duration between a rear endpoint and a demand prediction time point on a preset time axis, and establishes a period detection interval with the width of the preset period duration by taking the demand prediction time point as the rear endpoint;
The processing module generates a virtual point with changeable positions in the history interval, and establishes a virtual comparison interval with the width consistent with the period detection interval by taking the virtual point as a rear endpoint;
the acquisition module acquires the user electricity consumption at each time point and enables the processing module to conduct one comparison according to the user electricity consumption in the period detection interval and the user electricity consumption in the virtual comparison interval to determine the overall deviation electricity consumption, wherein Y is defined as the overall deviation electricity consumption, I 1n is the user electricity consumption at the nth time point in the period detection interval, I 2n is the user electricity consumption at the nth time point in the virtual comparison interval, ；
When the judging module judges that the overall deviation electricity consumption is smaller than the preset reference electricity consumption, the processing module defines the corresponding virtual point as an effective point, and establishes a standard time interval according to the effective point and the demand prediction time point;
the processing module compares the weather comprehensive data according to each time point and the current weather comprehensive data in the target time interval to determine a similarity value;
the processing module determines a similarity value with the largest value according to a preset ordering rule, and outputs the user electricity consumption at a time point corresponding to the similarity value as predicted electricity consumption for subsequent electricity dispatching.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411074783.8A CN118586688B (en) | 2024-08-07 | 2024-08-07 | Multi-level data processing method and system in smart power grid |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411074783.8A CN118586688B (en) | 2024-08-07 | 2024-08-07 | Multi-level data processing method and system in smart power grid |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118586688A CN118586688A (en) | 2024-09-03 |
| CN118586688B true CN118586688B (en) | 2024-11-01 |
Family
ID=92535961
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202411074783.8A Active CN118586688B (en) | 2024-08-07 | 2024-08-07 | Multi-level data processing method and system in smart power grid |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118586688B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113222403A (en) * | 2021-05-10 | 2021-08-06 | 湖南省计量检测研究院 | Power adjusting method and device based on big data, storage medium and electronic equipment |
| CN114548567A (en) * | 2022-02-24 | 2022-05-27 | 广西电网有限责任公司 | Load prediction method, device, equipment and readable storage medium |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP6372684B2 (en) * | 2014-03-31 | 2018-08-15 | 清水建設株式会社 | Electric power demand prediction apparatus and electric power demand prediction method |
| KR101962539B1 (en) * | 2017-07-06 | 2019-03-26 | 두산중공업 주식회사 | Method and Apparatus for Predicting Power Demand and Apparatus and Method for Controlling Charge/Discharge of ESS based on It |
| KR102514996B1 (en) * | 2022-09-13 | 2023-03-29 | 성한 주식회사 | Power demand forecasting system and method using unsupervised learning and supervised learning in machine learning |
| CN117791555A (en) * | 2024-01-26 | 2024-03-29 | 国网福建省电力有限公司 | Intelligent analysis method and system for power grid distribution transformer load |
| CN118214703B (en) * | 2024-05-17 | 2024-07-19 | 福建中信网安信息科技有限公司 | Digital twin data stream tracking and tracing system |
-
2024
- 2024-08-07 CN CN202411074783.8A patent/CN118586688B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113222403A (en) * | 2021-05-10 | 2021-08-06 | 湖南省计量检测研究院 | Power adjusting method and device based on big data, storage medium and electronic equipment |
| CN114548567A (en) * | 2022-02-24 | 2022-05-27 | 广西电网有限责任公司 | Load prediction method, device, equipment and readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118586688A (en) | 2024-09-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113255973A (en) | Power load prediction method, power load prediction device, computer equipment and storage medium | |
| CN112330077B (en) | Power load prediction method, power load prediction device, computer equipment and storage medium | |
| WO2020098728A1 (en) | Cluster load prediction method and apparatus, and storage medium | |
| CN110751416A (en) | Method, device and equipment for predicting water consumption | |
| KR102478684B1 (en) | Method for predicting energy consumption for using ensemble learning, and computing apparatus for performing the method | |
| CN117469603B (en) | Multi-water-plant water supply system pressure optimal control method based on big data learning | |
| CN114648155A (en) | Source analysis method and emergency response system based on weather typing and weather forecast | |
| CN115907822A (en) | Load characteristic index relevance mining method considering region and economic influence | |
| CN113177857B (en) | Resource elastic allocation method based on typhoon disaster estimation | |
| CN114925874A (en) | Carbon emission pre-judging method and device based on BERT neural network model | |
| CN111144628A (en) | Distributed energy supply type cooling, heating and power load prediction model system and method | |
| CN115775053A (en) | Distributed photovoltaic power short-term prediction method based on improved similar time method | |
| CN111723085B (en) | Method and system for cleaning electric energy consumption data of public building air conditioner | |
| CN117863947A (en) | Integrated power adjusting method and system based on charging station | |
| CN108346009A (en) | A kind of power generation configuration method and device based on user model self study | |
| CN116070795A (en) | Intelligent energy management and control method and system based on Internet of things | |
| CN116454928A (en) | Multi-type energy storage cooperative scheduling method considering multiple time scales | |
| CN118586688B (en) | Multi-level data processing method and system in smart power grid | |
| CN117610972B (en) | Green building digital management system and method based on artificial intelligence | |
| CN112365082A (en) | Public energy consumption prediction method based on machine learning | |
| CN117495056A (en) | Power consumption data monitoring and optimizing method and system | |
| CN117332288A (en) | Building energy saving technology evaluation system and method | |
| CN110209260A (en) | Power consumption method for detecting abnormality, device, equipment and computer readable storage medium | |
| CN116579884A (en) | Power user behavior analysis method and system | |
| Niu et al. | The prediction of carbon emissions from construction land in central Yunnan urban agglomeration area based on multiple linear regression model |
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 |