CN116757874A - User energy situation deduction method based on group intelligent analysis - Google Patents
User energy situation deduction method based on group intelligent analysis Download PDFInfo
- Publication number
- CN116757874A CN116757874A CN202311033746.8A CN202311033746A CN116757874A CN 116757874 A CN116757874 A CN 116757874A CN 202311033746 A CN202311033746 A CN 202311033746A CN 116757874 A CN116757874 A CN 116757874A
- Authority
- CN
- China
- Prior art keywords
- situation
- energy
- energy consumption
- learning machine
- user
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 67
- 238000004458 analytical method Methods 0.000 title claims abstract description 26
- 238000005265 energy consumption Methods 0.000 claims abstract description 51
- 230000008878 coupling Effects 0.000 claims abstract description 10
- 238000010168 coupling process Methods 0.000 claims abstract description 10
- 238000005859 coupling reaction Methods 0.000 claims abstract description 10
- 238000012549 training Methods 0.000 claims description 44
- 238000012544 monitoring process Methods 0.000 claims description 19
- 230000006870 function Effects 0.000 claims description 15
- 239000003795 chemical substances by application Substances 0.000 claims description 12
- 238000012360 testing method Methods 0.000 claims description 12
- 230000005611 electricity Effects 0.000 claims description 8
- 230000004044 response Effects 0.000 claims description 8
- 239000013598 vector Substances 0.000 claims description 8
- 238000004519 manufacturing process Methods 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 4
- 238000007405 data analysis Methods 0.000 claims description 4
- 238000011156 evaluation Methods 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 4
- 238000005259 measurement Methods 0.000 claims description 4
- 238000010606 normalization Methods 0.000 claims description 4
- 238000011056 performance test Methods 0.000 claims description 4
- 238000005516 engineering process Methods 0.000 claims 1
- 230000007704 transition Effects 0.000 abstract description 3
- 238000012423 maintenance Methods 0.000 abstract 1
- 238000007726 management method Methods 0.000 description 8
- 230000008859 change Effects 0.000 description 6
- 230000008447 perception Effects 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000003860 storage Methods 0.000 description 2
- 238000012800 visualization Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/06—Energy or water supply
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/04—Inference or reasoning models
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02J—CIRCUIT ARRANGEMENTS OR SYSTEMS FOR SUPPLYING OR DISTRIBUTING ELECTRIC POWER; SYSTEMS FOR STORING ELECTRIC ENERGY
- H02J3/00—Circuit arrangements for ac mains or ac distribution networks
- H02J3/003—Load forecast, e.g. methods or systems for forecasting future load demand
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Software Systems (AREA)
- Business, Economics & Management (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- Economics (AREA)
- Computing Systems (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Strategic Management (AREA)
- Water Supply & Treatment (AREA)
- Computational Linguistics (AREA)
- General Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Tourism & Hospitality (AREA)
- Public Health (AREA)
- Primary Health Care (AREA)
- Marketing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Power Engineering (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application discloses a user energy situation deduction method based on group intelligent analysis, which relates to the technical field of energy situation deduction equipment and comprises the following specific steps: in the first stage, setting an energy consumption situation index system, performing transition treatment by setting and adopting three stages, setting the energy consumption situation index system in actual use, researching the coupling relation between an energy consumption situation data sequence and a time window for different time scales, constructing a time sequence sample set of the energy consumption situation, conforming to the requirement of constructing a park energy consumption accurate regulation model, managing, visualizing and storing energy consumption situation predicted value data and corresponding analysis results, facilitating the realization of energy consumption situation prediction management through Internet cloud management, improving the intelligent level of energy consumption situation prediction management, deducing the overall operation situation of a user from multi-dimensional energy consumption situation, and providing a new theoretical basis for maintenance and later examination of the user energy consumption situation deduction.
Description
Technical Field
The application relates to the technical field of energy situation deduction equipment, in particular to a group intelligent analysis-based user energy situation deduction method.
Background
The load clusters covered by the park are various, and the energy utilization mechanism and the characteristics are also different, so that the energy utilization situation value of the park user is difficult to obtain in a summation mode;
the prior art adopts a common method for describing the energy consumption state of the current park user by using the weighted sum energy consumption state values, wherein the energy consumption state values refer to the energy consumption state summary and the change trend of each load cluster in the park at a certain time monitoring point, and if the change trend of the energy consumption state values can be accurately predicted, the energy consumption rule of the park user can be effectively known and is used for constructing an accurate park energy consumption regulation model, so the application needs to design a user energy consumption state deduction method based on group intelligent analysis to solve the problems.
Disclosure of Invention
The application aims to provide a group intelligent analysis-based user energy situation deduction method, which solves the problems that the energy situation of the current park user is described by using the weighted sum energy situation value in the background art, the energy situation value refers to the energy situation summary of each load cluster in the park and the change trend thereof at a certain time monitoring point, and if the change trend of the energy situation value can be accurately predicted, the energy rule of the park user can be effectively known and is used for constructing a park energy accurate regulation model.
In order to solve the problems, the application provides a technical scheme that:
a user energy situation deduction method based on group intelligent analysis comprises the following specific steps:
s1, in a first stage, setting an energy consumption situation index system;
s2, in a second stage, researching the coupling relation between the power consumption situation data sequences and the time windows facing different time scales, constructing a time sequence sample set of power consumption situations, and completing power consumption situation prediction of a key load cluster of a future time node by fitting a regression function;
and S3, in a third stage, researching a key load cluster based on a multi-agent mode by adopting a mode of combining an electricity consumption situation prediction method and a park user electricity consumption situation combination prediction method, wherein each key load cluster corresponds to one agent, the energy consumption situations of the key load clusters are respectively predicted, and then the energy consumption situation predicted values of the clusters are integrated to obtain the energy consumption situation predicted value of the park user.
Preferably, the energy consumption situation index system in S1 includes an energy consumption sensing data, a demand response characteristic, a park production process, a load control characteristic, a response wish, a production working condition, and a perception system evaluation index system, in which different load clusters have corresponding weights, which is known asEach key load cluster is corresponding to the weight of +.>，/>A certain time monitoring point->The energy consumption measurement values of the key load clusters are respectively as follows。
Preferably, the time monitoring point in the S1The calculation formula of the energy situation value V of the park user is as follows:。
preferably, the step S2 further includes the specific steps of:
s201, predicting the power consumption situation of a key load cluster, wherein input quantity data are the power consumption situations of a plurality of continuous discrete time points, and constructing a sample set by utilizing a data analysis means in the time sequence field when in use;
s202, utilizing a sliding time window method to monitor each discrete time pointThe key load clusters are constructed in time series with the power state values Vi>If the sliding time window size is set to be 4 and the sliding step length is set to be 1, then，/>，/>。
Preferably, in the step S3, the time sequence S is constructed into a data sample set when the predicted value of the energy consumption situation is obtained, and the training is performed by using Boosting algorithm to obtain the weak learning machine sequenceAnd each weak learning machine corresponds to a key load cluster, T is the maximum iteration number of the Boosting algorithm, and the strong learning machine H, namely the energy situation predicted value of the park user, is obtained by using a method of weighting and summing the weak learning machine sequences H.
Preferably, the method further comprises the following steps of:
s301, setting the maximum iteration number T of a Boosting algorithm, and setting a weak learning machine as CVM;
s302, constructing a data sample set, if the time sequence isV1, V2, V3 and V4 are input vectors of a weak learning machine CVM, the energy situation value V5 for the load cluster predicted according to S301 is used as an output vector of the CVM, and the data sample set is divided into a training sample set and a test sample set according to the open set test requirement;
s303, setting a training sample set as an original data set D of a Boosting algorithm;
，/>for the original dataset DInitial weight of each sample, +.>，/>M is the total number of samples;
s304, extracting the first 20% from the original data set DThe samples with the largest values form a training sample subset Df, and the training sample subset Df is trained by the CVM to obtain a weak learning machine hf, ">，/>The selection condition of (c) is that,(4-2) if this condition is satisfied, +.>Calculating a result for the same; if the condition is not satisfied, thenWherein->Energy situation value for load cluster +.>A predicted value;
s305, calculating training errors of weak learning machine hf，/>Wherein->In order for the training error to be a function of the user,，/>is the extraction probability;
s306, calculating the weight of the weak learning machine hf，
；
S307, updating the weight of the training sample,
；
s308, executing the next link when one of the following two conditions is met, otherwise returning to S303;
condition 1: the Boosting algorithm reaches the maximum iteration times T;
condition 2: training sample weights in sample subset DfThe value no longer changes;
s309, outputting a strong learning machine H,sign is a sign function;
s3010, performing performance test on the strong learning machine H, and inputting a test sample set into the strong learning machine H to detect whether the performance of the strong learning machine H meets the initial setting requirement.
Preferably, it should be determined whether to execute the step S3010 according to the actual situation when executing the step S3010.
Preferably, the third stage is entered after the step S3010 is completed, and the method is used for predicting the energy situation value of the load cluster, and the strong learning machine H is used for completing the energy situation value prediction of the load cluster at the future time monitoring point.
Preferably, after the three stages, a Boosting-based energy situation value for the load cluster is deduced, and the energy situation value for the load cluster of the future time monitoring point is accurately predicted.
Preferably, in S307: wherein the method comprises the steps ofFor updated training sample weights, +.>For training the weights of the samples, +.>For the calculated value obtained by the weak learning machine according to the input quantity xl, vl is the actual energy situation value, mf is the normalization coefficient, ensuring +.>。
The beneficial effects of the application are as follows: according to the method, a transition processing is carried out by setting three stages, an energy situation index system is set in actual use, the coupling relation between an energy situation data sequence and a time window is researched, a time sequence sample set of energy situation is constructed, the energy situation prediction of a key load cluster of a future time node is completed by fitting a regression function, the key load clusters based on a multi-agent mode are researched by adopting a mode of combining an energy situation prediction method and an energy situation combination prediction method of a park user, each key load cluster corresponds to one agent, the energy situation of the key load cluster is respectively predicted, the energy situation predicted values of a plurality of clusters are merged to obtain the energy situation predicted value of the park user, the change trend of the energy situation value is accurately predicted, the energy rule of the park user is effectively known, the requirements of constructing a accurate energy regulation model of the park are met, management, visualization and storage are carried out on the energy situation predicted value data and the corresponding analysis result, and the energy situation predicted by the aid of the intelligent management system is improved, and the energy situation is improved to be used by the aid of a theoretical energy management system, and the energy situation is predicted from the cloud management system, and the energy situation is improved, and the energy situation is predicted according to the energy situation is predicted and the energy situation predicted.
Drawings
For ease of illustration, the application is described in detail by the following detailed description and the accompanying drawings.
FIG. 1 is a general flow chart of a user energy situation deduction method based on group intelligent analysis;
FIG. 2 is a network security situation value calculation process diagram of a user energy situation deduction method based on group intelligent analysis;
FIG. 3 is a sliding time window diagram of a user energy situation deduction method based on group intelligent analysis;
fig. 4 is a flow chart of energy situation value prediction of a Boosting-based load cluster based on a user energy situation deduction method based on group intelligent analysis.
Detailed Description
As shown in fig. 1, 2, 3 and 4, the following technical solutions are adopted in this embodiment: a user energy situation deduction method based on group intelligent analysis comprises the following specific steps:
s1, in a first stage, setting an energy consumption situation index system;
s2, in a second stage, researching the coupling relation between the power consumption situation data sequences and the time windows facing different time scales, constructing a time sequence sample set of power consumption situations, and completing power consumption situation prediction of a key load cluster of a future time node by fitting a regression function;
and S3, in a third stage, researching a key load cluster based on a multi-agent mode by adopting a mode of combining an electricity consumption situation prediction method and a park user electricity consumption situation combination prediction method, wherein each key load cluster corresponds to one agent, the energy consumption situations of the key load clusters are respectively predicted, and then the energy consumption situation predicted values of the clusters are integrated to obtain the energy consumption situation predicted value of the park user.
Further, the energy consumption situation index system in S1 comprises energy consumption sensing data, demand response characteristics, park production process, load control characteristics, response will and producersCondition and perception system evaluation index system, in the energy consumption situation index system, different load clusters have corresponding weights, and the condition and perception system evaluation index system is known asEach key load cluster is corresponding to the weight of +.>，/>A certain time monitoring point->The energy consumption measurement values of the key load clusters are respectively as follows。
Further, the time monitoring point in S1The calculation formula of the energy situation value V of the park user is as follows:。
further, the step S2 further includes the specific steps of:
s201, predicting the power consumption situation of a key load cluster, wherein input quantity data are the power consumption situations of a plurality of continuous discrete time points, and constructing a sample set by utilizing a data analysis means in the time sequence field when in use;
s202, utilizing a sliding time window method to monitor each discrete time pointThe key load clusters are constructed in time series with the power state values Vi>If the sliding time window size is set to be 4 and the sliding step length is set to be 1, then，/>，/>。
Further, the time sequence S is constructed into a data sample set when the energy situation predicted value is obtained in the step S3, and the weak learning machine sequence is obtained by training the time sequence S through Boosting algorithmAnd each weak learning machine corresponds to a key load cluster, T is the maximum iteration number of the Boosting algorithm, and the strong learning machine H, namely the energy situation predicted value of the park user, is obtained by using a method of weighting and summing the weak learning machine sequences H.
Further, the method for obtaining the energy consumption situation predicted value of the park user further comprises the following steps:
s301, setting the maximum iteration number T of a Boosting algorithm, and setting a weak learning machine as CVM;
s302, constructing a data sample set, if the time sequence isV1, V2, V3 and V4 are input vectors of a weak learning machine CVM, the energy situation value V5 for the load cluster predicted according to S301 is used as an output vector of the CVM, and the data sample set is divided into a training sample set and a test sample set according to the open set test requirement;
s303, setting a training sample set as an original data set D of a Boosting algorithm;
，/>initial weight for each sample in the original dataset D +.>，/>M is the total number of samples;
s304, extracting the first 20% from the original data set DThe samples with the largest values form a training sample subset Df, and the training sample subset Df is trained by the CVM to obtain a weak learning machine hf, ">，/>The selection condition of (c) is that,(4-2) if this condition is satisfied, +.>Calculating a result for the same; if the condition is not satisfied, thenWherein->Energy situation value for load cluster +.>A predicted value;
s305, calculating training errors of weak learning machine hf，/>Wherein->In order for the training error to be a function of the user,，/>is the extraction probability;
s306, calculating the weight of the weak learning machine hf，
；
S307, updating the weight of the training sample,
；
s308, executing the next link when one of the following two conditions is met, otherwise returning to S303;
condition 1: the Boosting algorithm reaches the maximum iteration times T;
condition 2: training sample weights in sample subset DfThe value no longer changes;
s309, outputting a strong learning machine H,sign is a sign function;
s3010, performing performance test on the strong learning machine H, and inputting a test sample set into the strong learning machine H to detect whether the performance of the strong learning machine H meets the initial setting requirement.
Further, when executing the step S3010, it should be determined whether to execute the step S3010 according to the actual situation.
Further, the third stage is entered after the step S3010 is completed, and the method is used for predicting the energy situation value of the load cluster, and the strong learning machine H is used for completing the energy situation value prediction of the load cluster at the future time monitoring point.
Further, after the three stages, a Boosting-based energy situation value of the load cluster is deduced, and the energy situation value of the load cluster of the future time monitoring point is accurately predicted.
Further, what is said isIn S307: wherein the method comprises the steps ofFor updated training sample weights, +.>For training the weights of the samples, +.>For the calculated value obtained by the weak learning machine according to the input quantity xl, vl is the actual energy situation value, mf is the normalization coefficient, ensuring +.>。
Examples: when the group intelligent analysis is connected to the user energy situation deduction:
s1, checking equipment by staff, starting after checking, setting up a power consumption situation index system, and evaluating the index system by using power consumption sensing data, demand response characteristics, load control characteristics, response willingness, production working conditions and sensing systems, wherein in the power consumption situation index system, different load clusters have corresponding weights, and the power consumption situation index system is known as followsEach key load cluster is corresponding to the weight of +.>，/>A certain time monitoring point->The energy consumption measurement value of each key load cluster is +.>；
Time monitoring pointPark user of (C)The energy situation value V is calculated as follows: />；
S2, researching coupling relations between power consumption situation data sequences and time windows for different time scales, constructing a time sequence sample set of power consumption situations, and completing power consumption situation prediction of a key load cluster of a future time node by fitting a regression function; the power consumption situation prediction input quantity data of the key load cluster are the power consumption situations of a plurality of continuous discrete time points, and a sample set is constructed by utilizing a data analysis means in the time sequence field when in use; utilizing sliding time window method to make each discrete time monitoring pointThe key load clusters are constructed in time series with the power state values Vi>If the sliding time window size is set to 4 and the sliding step length is 1, then +.>，，/>；
S3, researching key load clusters based on multi-agent modes by adopting a mode of combining an electricity consumption situation prediction method and a park user electricity consumption situation combination prediction method, wherein each key load cluster corresponds to one agent, the energy consumption situation of the key load cluster is respectively predicted, then the energy consumption situation predicted values of a plurality of clusters are integrated to obtain an energy consumption situation predicted value of the park user, a time sequence S is constructed into a data sample set when the energy consumption situation predicted value is obtained, and a weak learning machine sequence is obtained by training the time sequence by using a Boosting algorithmEach weak schoolThe learning machine corresponds to a key load cluster, T is the maximum iteration number of the Boosting algorithm, and the strong learning machine H, namely the energy consumption situation predicted value of the park user, is obtained by using a method of weighting and summing the weak learning machine sequence H:
s301, setting the maximum iteration number T of a Boosting algorithm, and setting a weak learning machine as CVM;
s302, constructing a data sample set, if the time sequence isV1, V2, V3 and V4 are input vectors of a weak learning machine CVM, the energy situation value V5 for the load cluster predicted according to S301 is used as an output vector of the CVM, and the data sample set is divided into a training sample set and a test sample set according to the open set test requirement;
s303, setting a training sample set as an original data set D of a Boosting algorithm;
，/>initial weight for each sample in the original dataset D +.>，/>M is the total number of samples;
s304, extracting the first 20% from the original data set DThe samples with the largest values form a training sample subset Df, and the training sample subset Df is trained by the CVM to obtain a weak learning machine hf, ">，/>The selection condition of (c) is that,
（4-2）
if this condition is satisfied, thenCalculating a result for the same; if this condition is not met +.>WhereinEnergy situation value for load cluster +.>A predicted value;
s305, calculating training errors of weak learning machine hf，/>Wherein->In order for the training error to be a function of the user,，/>is the extraction probability;
s306, calculating the weight of the weak learning machine hf，
；
S307, updating the weight of the training sample,
；
wherein the method comprises the steps ofFor updated training sample weights, +.>For training the weights of the samples, +.>For the calculated value obtained by the weak learning machine according to the input quantity xl, vl is the actual energy situation value, mf is the normalization coefficient, and the assurance is ensured；
S308, executing the next link when one of the following two conditions is met, otherwise returning to S303;
condition 1: the Boosting algorithm reaches the maximum iteration times T;
condition 2: training sample weights in sample subset DfThe value no longer changes;
s309, outputting a strong learning machine H,sign is a sign function;
s3010, a strong learning machine H performance test is carried out, a test sample set is input into the strong learning machine H to detect whether the performance of the strong learning machine H meets the initial setting requirement, whether the S3010 link is executed or not is determined according to actual conditions when the S3010 is executed, a third stage is entered after the S3010 is completed and used for predicting the energy situation value of the load cluster, the load cluster energy situation value of a future time monitoring point is predicted by utilizing the strong learning machine H, and after the three stages, the energy situation value of the load cluster based on Boosting is deduced and accurately predicted for the load cluster energy situation value of the future time monitoring point.
Specific: according to the method, a transition processing is carried out by setting three stages, an energy situation index system is set in actual use, the coupling relation between an energy situation data sequence and a time window is researched, a time sequence sample set of energy situation is constructed, the energy situation prediction of a key load cluster of a future time node is completed by fitting a regression function, the key load clusters based on a multi-agent mode are researched by adopting a mode of combining an energy situation prediction method and an energy situation combination prediction method of a park user, each key load cluster corresponds to one agent, the energy situation of the key load cluster is respectively predicted, the energy situation predicted values of a plurality of clusters are merged to obtain the energy situation predicted value of the park user, the change trend of the energy situation value is accurately predicted, the energy rule of the park user is effectively known, the requirements of constructing a accurate energy regulation model of the park are met, management, visualization and storage are carried out on the energy situation predicted value data and the corresponding analysis result, and the energy situation predicted by the aid of the intelligent management system is improved, and the energy situation is improved to be used by the aid of a theoretical energy management system, and the energy situation is predicted from the cloud management system, and the energy situation is improved, and the energy situation is predicted according to the energy situation is predicted and the energy situation predicted.
Those of ordinary skill in the art will appreciate that the modules and method steps of the examples described in connection with the embodiments disclosed herein can be implemented as electronic hardware, or as a combination of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the solution. 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 application.
It will be clearly understood by those skilled in the art that, for convenience and brevity of description, specific working procedures of the apparatus, device and module described above may refer to corresponding procedures in the foregoing method embodiments, which are not repeated herein.
In the several embodiments provided by the present application, it should be understood that the disclosed apparatus, system, and method may be implemented in other ways. For example, the system embodiments described above are merely illustrative, e.g., the division of the modules is merely a logical function division, and there may be additional divisions when actually implemented, e.g., multiple modules or units may be combined or integrated into another device, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be an indirect coupling or communication connection via some interfaces, devices or apparatuses, which may be in electrical, mechanical or other form.
In addition, it should be noted that the combination of the technical features described in the present application is not limited to the combination described in the claims or the combination described in the specific embodiments, and all the technical features described in the present application may be freely combined or combined in any manner unless contradiction occurs between them.
It should be noted that the above-mentioned embodiments are merely examples of the present application, and it is obvious that the present application is not limited to the above-mentioned embodiments, and many similar variations are possible. All modifications attainable or obvious from the present disclosure set forth herein should be deemed to be within the scope of the present disclosure.
The foregoing is merely illustrative of the preferred embodiments of the present application, and is not intended to limit the scope of the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the protection scope of the present application.
Claims (10)
1. The user energy situation deduction method based on the group intelligent analysis is characterized by comprising the following specific steps of:
s1, in a first stage, setting an energy consumption situation index system;
s2, in a second stage, researching the coupling relation between the power consumption situation data sequences and the time windows facing different time scales, constructing a time sequence sample set of power consumption situations, and completing power consumption situation prediction of a key load cluster of a future time node by fitting a regression function;
and S3, in a third stage, researching a key load cluster based on a multi-agent mode by adopting a mode of combining an electricity consumption situation prediction method and a park user electricity consumption situation combination prediction method, wherein each key load cluster corresponds to one agent, the energy consumption situations of the key load clusters are respectively predicted, and then the energy consumption situation predicted values of the clusters are integrated to obtain the energy consumption situation predicted value of the park user.
2. The user energy situation deduction method based on group intelligent analysis according to claim 1, wherein the method comprises the following steps: the energy consumption situation index system in S1 comprises energy consumption sensing data, demand response characteristics, park production technology, load control characteristics, response will, production working conditions and a sensing system evaluation index system, and in the energy consumption situation index system, different load clusters have corresponding weights and are known to beEach key load cluster is corresponding to the weight of +.>，A certain time monitoring point->The energy consumption measurement value of each key load cluster is +.>。
3. The user energy situation deduction method based on group intelligent analysis according to claim 2, wherein: the time monitoring point in the S1The calculation formula of the energy situation value V of the park user is as follows: />。
4. The method for deducting the user energy situation based on intelligent group analysis according to claim 3, wherein the method comprises the following steps: the step S2 further comprises the following steps:
s201, predicting the power consumption situation of a key load cluster, wherein input quantity data are the power consumption situations of a plurality of continuous discrete time points, and constructing a sample set by utilizing a data analysis means in the time sequence field when in use;
s202, utilizing a sliding time window method to monitor each discrete time pointThe key load clusters are constructed in time series with the power state values Vi>If the sliding time window size is set to be 4 and the sliding step length is set to be 1, then，/>，/>。
5. The method for deducting the user energy situation based on intelligent group analysis according to claim 4, wherein the method comprises the following steps: the time sequence S is constructed into a data sample set when the energy situation predicted value is obtained in the S3, and the weak learning machine sequence is obtained by training the time sequence S through Boosting algorithmEach weak learning machine corresponds to a key load cluster, T is the maximum iteration number of the Boosting algorithm, and the method for weighting and summing the weak learning machine sequences h is utilized to obtainAnd (5) reaching a strong learning machine H, namely, predicting the energy situation of the park user.
6. The method for deducing the energy consumption situation of the users based on the intelligent group analysis according to claim 5, wherein the method for obtaining the predicted value of the energy consumption situation of the users in the park further comprises the following steps:
s301, setting the maximum iteration number T of a Boosting algorithm, and setting a weak learning machine as CVM;
s302, constructing a data sample set, if the time sequence isV1, V2, V3 and V4 are input vectors of a weak learning machine CVM, the energy situation value V5 for the load cluster predicted according to S301 is used as an output vector of the CVM, and the data sample set is divided into a training sample set and a test sample set according to the open set test requirement;
s303, setting a training sample set as an original data set D of a Boosting algorithm;
，/>for the initial weight of each sample in the original dataset D,，/>m is the total number of samples;
s304, extracting the first 20% from the original data set DThe samples with the largest values form a training sample subset Df, and the training sample subset Df is trained by the CVM to obtain a weak learning machine hf, ">，/>The selection condition of (1) is->(4-2) if this condition is satisfied, then +.>Calculating a result for the same; if this condition is not met +.>Wherein->Energy situation value for load cluster +.>A predicted value;
s305, calculating training errors of weak learning machine hf，/>Wherein->In order for the training error to be a function of the user,，/>is the extraction probability;
s306, calculating the weight of the weak learning machine hf，
；
S307, updating the weight of the training sample,
；
s308, executing the next link when one of the following two conditions is met, otherwise returning to S303;
condition 1: the Boosting algorithm reaches the maximum iteration times T;
condition 2: training sample weights in sample subset DfThe value no longer changes;
s309, outputting a strong learning machine H,sign is a sign function;
s3010, performing performance test on the strong learning machine H, and inputting a test sample set into the strong learning machine H to detect whether the performance of the strong learning machine H meets the initial setting requirement.
7. The method for deducting the user energy situation based on intelligent group analysis according to claim 6, wherein the method comprises the following steps: when executing the step S3010, it should be determined whether to execute the step S3010 according to the actual situation.
8. The method for deducting the user energy situation based on intelligent group analysis according to claim 6, wherein the method comprises the following steps: and after the S3010 is finished, entering a third stage for predicting the energy situation value of the load cluster, and finishing the energy situation value prediction of the load cluster of the future time monitoring point by using a strong learning machine H.
9. The method for deducting the user energy situation based on intelligent group analysis according to claim 8, wherein the method comprises the following steps: after the three phases, the energy situation value of the load cluster based on Boosting is deduced, and the energy situation value of the load cluster of the future time monitoring point is accurately predicted.
10. The method for deducting the user energy situation based on intelligent group analysis according to claim 6, wherein the method comprises the following steps: in S307: wherein the method comprises the steps ofFor updated training sample weights, +.>For training the weights of the samples, +.>For the calculated value obtained by the weak learning machine according to the input quantity xl, vl is the actual energy situation value, mf is the normalization coefficient, and the assurance is ensured。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311033746.8A CN116757874A (en) | 2023-08-17 | 2023-08-17 | User energy situation deduction method based on group intelligent analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311033746.8A CN116757874A (en) | 2023-08-17 | 2023-08-17 | User energy situation deduction method based on group intelligent analysis |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116757874A true CN116757874A (en) | 2023-09-15 |
Family
ID=87961228
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311033746.8A Pending CN116757874A (en) | 2023-08-17 | 2023-08-17 | User energy situation deduction method based on group intelligent analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116757874A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117710153A (en) * | 2024-02-06 | 2024-03-15 | 深圳市先行电气技术有限公司 | Energy utilization auxiliary decision-making method and system based on multi-terminal equipment |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102185735A (en) * | 2011-04-26 | 2011-09-14 | 华北电力大学 | Network security situation prediction method |
| CN102231144A (en) * | 2011-06-03 | 2011-11-02 | 中国电力科学研究院 | Method for predicting theoretical line loss of power distribution network based on Boosting algorithm |
| CN102263790A (en) * | 2011-07-18 | 2011-11-30 | 华北电力大学 | Intrusion detecting method based on integrated learning |
| CN105550946A (en) * | 2016-01-28 | 2016-05-04 | 东北电力大学 | Multi-agent based electricity utilization strategy capable of enabling residential users to participate in automated demand response |
| CN110222910A (en) * | 2019-06-20 | 2019-09-10 | 武汉大学 | A kind of active power distribution network Tendency Prediction method and forecasting system |
| CN111652413A (en) * | 2020-05-20 | 2020-09-11 | 国网江苏省电力有限公司南通供电分公司 | Industrial power load prediction method based on multi-Agent distributed mass data processing |
| CN113111943A (en) * | 2021-04-13 | 2021-07-13 | 深圳供电局有限公司 | Analysis method for data full life cycle safety |
-
2023
- 2023-08-17 CN CN202311033746.8A patent/CN116757874A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102185735A (en) * | 2011-04-26 | 2011-09-14 | 华北电力大学 | Network security situation prediction method |
| CN102231144A (en) * | 2011-06-03 | 2011-11-02 | 中国电力科学研究院 | Method for predicting theoretical line loss of power distribution network based on Boosting algorithm |
| CN102263790A (en) * | 2011-07-18 | 2011-11-30 | 华北电力大学 | Intrusion detecting method based on integrated learning |
| CN105550946A (en) * | 2016-01-28 | 2016-05-04 | 东北电力大学 | Multi-agent based electricity utilization strategy capable of enabling residential users to participate in automated demand response |
| CN110222910A (en) * | 2019-06-20 | 2019-09-10 | 武汉大学 | A kind of active power distribution network Tendency Prediction method and forecasting system |
| CN111652413A (en) * | 2020-05-20 | 2020-09-11 | 国网江苏省电力有限公司南通供电分公司 | Industrial power load prediction method based on multi-Agent distributed mass data processing |
| CN113111943A (en) * | 2021-04-13 | 2021-07-13 | 深圳供电局有限公司 | Analysis method for data full life cycle safety |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117710153A (en) * | 2024-02-06 | 2024-03-15 | 深圳市先行电气技术有限公司 | Energy utilization auxiliary decision-making method and system based on multi-terminal equipment |
| CN117710153B (en) * | 2024-02-06 | 2024-05-10 | 深圳市先行电气技术有限公司 | Energy utilization auxiliary decision-making method and system based on multi-terminal equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113805064B (en) | Lithium ion battery pack health state prediction method based on deep learning | |
| Escanciano et al. | Uniform convergence of weighted sums of non and semiparametric residuals for estimation and testing | |
| CN109818349B (en) | Power grid robust state prediction method based on multidimensional state matrix sliding matching | |
| CN116757874A (en) | User energy situation deduction method based on group intelligent analysis | |
| CN106934242B (en) | The health degree appraisal procedure and system of equipment under multi-mode based on Cross-Entropy Method | |
| CN116150897A (en) | Machine tool spindle performance evaluation method and system based on digital twin | |
| Kong et al. | A cumulative-exposure-based algorithm for failure data from a load-sharing system | |
| CN111369120A (en) | Non-invasive load monitoring method based on equipment transfer learning | |
| Chen et al. | A deep learning feature fusion based health index construction method for prognostics using multiobjective optimization | |
| CN114936504A (en) | Equipment residual life prediction method and system based on Bayesian multi-source data fusion | |
| CN115204491A (en) | Production line working condition prediction method and system based on digital twinning and LSTM | |
| CN115114128A (en) | Satellite health state evaluation system and evaluation method | |
| CN112149953B (en) | Electromechanical equipment operation safety assessment method based on multimode linkage and multistage cooperation | |
| CN106056305A (en) | Power generation system reliability rapid assessment method based on state clustering | |
| WO2024056051A1 (en) | Non-intrusive flexible load aggregation characteristic identification and optimization method, apparatus, and device | |
| CN110887652B (en) | Interactive multi-model detection method for vibration detection and displacement extraction of accelerometer | |
| CN112131716A (en) | Random electromagnetic transient analysis method and system for power electronic power system | |
| CN112988527A (en) | GPU management platform anomaly detection method and device and storage medium | |
| CN109117352B (en) | Server performance prediction method and device | |
| CN115809818A (en) | Multidimensional diagnosis and evaluation method and device for auxiliary equipment of pumped storage power station | |
| CN111723857B (en) | Intelligent monitoring method and system for running state of process production equipment | |
| CN110020407B (en) | Data compression iterative computation method based on principal component analysis | |
| CN114331349A (en) | Scientific research project management method and system based on Internet of things technology | |
| CN116707331B (en) | Inverter output voltage high-precision adjusting method and system based on model prediction | |
| CN109598644A (en) | Stealing user identification method and terminal device based on Gaussian Profile |
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 |