CN112257964B - Load-intensive urban intelligent park demand aggregation modeling method - Google Patents
Load-intensive urban intelligent park demand aggregation modeling method Download PDFInfo
- Publication number
- CN112257964B CN112257964B CN202011321937.0A CN202011321937A CN112257964B CN 112257964 B CN112257964 B CN 112257964B CN 202011321937 A CN202011321937 A CN 202011321937A CN 112257964 B CN112257964 B CN 112257964B
- Authority
- CN
- China
- Prior art keywords
- data
- load
- curve
- intensive
- intelligent park
- 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
- 238000000034 method Methods 0.000 title claims abstract description 63
- 238000004220 aggregation Methods 0.000 title claims abstract description 40
- 230000002776 aggregation Effects 0.000 title claims abstract description 40
- 230000010354 integration Effects 0.000 claims abstract description 8
- 238000012545 processing Methods 0.000 claims abstract description 7
- 239000011159 matrix material Substances 0.000 claims description 32
- 238000004458 analytical method Methods 0.000 claims description 18
- 230000008569 process Effects 0.000 claims description 18
- 238000005259 measurement Methods 0.000 claims description 12
- 238000005265 energy consumption Methods 0.000 claims description 10
- 230000008859 change Effects 0.000 claims description 9
- 238000012937 correction Methods 0.000 claims description 9
- 238000007405 data analysis Methods 0.000 claims description 7
- 238000013075 data extraction Methods 0.000 claims description 7
- 238000013500 data storage Methods 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000010606 normalization Methods 0.000 claims description 4
- 238000012546 transfer Methods 0.000 claims description 4
- 238000006243 chemical reaction Methods 0.000 claims description 3
- 238000006116 polymerization reaction Methods 0.000 claims description 3
- 238000001308 synthesis method Methods 0.000 claims description 3
- 238000001514 detection method Methods 0.000 claims description 2
- 238000005457 optimization Methods 0.000 abstract description 13
- 238000005516 engineering process Methods 0.000 abstract description 3
- 238000003672 processing method Methods 0.000 abstract 1
- 238000010187 selection method Methods 0.000 abstract 1
- 238000007726 management method Methods 0.000 description 7
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- 238000005520 cutting process Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 230000005611 electricity Effects 0.000 description 2
- 238000003860 storage Methods 0.000 description 2
- 230000004931 aggregating effect Effects 0.000 description 1
- 125000003275 alpha amino acid group Chemical group 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007621 cluster analysis Methods 0.000 description 1
- 150000001875 compounds Chemical class 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000011217 control strategy Methods 0.000 description 1
- 238000001816 cooling Methods 0.000 description 1
- 238000004146 energy storage Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000007710 freezing Methods 0.000 description 1
- 230000008014 freezing Effects 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 239000013598 vector Substances 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 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
- G06Q10/00—Administration; Management
- G06Q10/04—Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem"
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
- G06F16/215—Improving data quality; Data cleansing, e.g. de-duplication, removing invalid entries or correcting typographical errors
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/28—Databases characterised by their database models, e.g. relational or object models
- G06F16/284—Relational databases
- G06F16/285—Clustering or classification
-
- 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
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0631—Resource planning, allocation, distributing or scheduling for enterprises or organisations
-
- 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
- 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
- G06Q50/10—Services
- G06Q50/26—Government or public services
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Economics (AREA)
- Physics & Mathematics (AREA)
- Strategic Management (AREA)
- General Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Quality & Reliability (AREA)
- Health & Medical Sciences (AREA)
- Development Economics (AREA)
- Educational Administration (AREA)
- Game Theory and Decision Science (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Operations Research (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Public Health (AREA)
- Water Supply & Treatment (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a load-intensive urban intelligent park demand aggregation modeling method, which is characterized in that an integration and selection method of original data is provided by starting from a database structure and a data source form of an energy management platform in an intelligent park, and an energy data set needing to be mined and processed is extracted; on the basis, a processing method of the demand aggregation original data of the load-intensive city intelligent park is provided, the energy utilization data of the load-intensive city intelligent park is processed, and the utilization efficiency of the data is improved by processing, identifying and correcting the bad source data; and then, the energy utilization characteristics of the load-intensive city intelligent park demand data are mined by using a fuzzy clustering technology, and a load-intensive city intelligent park demand aggregation model is established. The method has the advantages that the result of demand aggregation modeling is applied to the energy optimization scheduling of the intelligent parks, the load optimization operation of multiple types of users in the intelligent parks in the load-intensive city can be realized, and then the overall energy efficiency of the intelligent parks is improved.
Description
Technical Field
The invention relates to the technical field of comprehensive energy utilization, in particular to comprehensive energy utilization of a load-intensive city intelligent park, and specifically relates to a demand aggregation modeling method of the load-intensive city intelligent park.
Background
End users on the energy demand side of the current load-intensive city smart park are gradually an aggregate of various energy utilization forms such as energy storage, combined cooling, heating and power supply, load and the like, and have certain self energy coordination capability, and the load-intensive city smart park mainly comprises three typical end users: industrial users, institutions and data centers, each having different energy usage devices and features, are embodied in three aspects:
(1) There are significant differences in the cold-hot electrical load characteristics of various typical end users. The loads of the data center are mainly electric load and space cold load, the space heat load is less, and the hot water load and the freezing and refrigerating load are avoided. The load of the industrial user is complete in variety, and the energy utilization curve shows obvious two-shift characteristic; the load changes at 8-24 are small, and the load in the working period and the non-working period are obviously different; the overall load characteristics of the public institution show extremely strong timeliness and seasonality, the energy consumption curve is stable from 9 am to 10 pm, and the load level is low in other time periods; the commercial off-season load levels are higher, especially holiday load levels, while the commercial off-season load levels are lower. The characteristic curves of the cold and hot electric loads of various end users are different, so that different energy supply equipment operation control strategies are required to be formulated for different end users, and the requirements of the cold and hot electric loads of the various end users are met.
(2) There are variations in the various typical end user power devices and forms of power. Industrial users and public institutions employ ac power systems, while data centers employ ac-dc hybrid power systems. In addition, data centers are also equipped with UPS's and large capacity batteries that can meet the system off-grid for a period of time, which is not available to other types of end users. The types and models of the energy supply devices of various typical end users are different, and the operation characteristics of the energy supply devices are different, so that different energy supply and energy utilization optimization coordination control models are required to be established for different end users. (3) There are differences in the solving methods for the energy optimization of different end users using the coordinated control model.
Because the mathematical models of the energy supply equipment in different end users are different, different solving methods are needed for different end user energy optimization and utilization of the coordination control model, for example, public institutions and data centers can adopt linear programming algorithm to solve, and industrial users can only adopt interior point method for solving nonlinear optimization problem. By solving the respective energy optimization and utilization coordination control model of each terminal user, the economic operation and energy optimization and utilization of the system can be realized on the basis of meeting various load demands.
The types and the number of users in the park are numerous, such as industrial users, data center users, public institution users, resident users and the like, and the classified aggregation control of the users has better park-level control effect than the single control. The aggregation theory is derived from an element-based load modeling method and is a theoretical basis of a unit user-based modeling method. The demand modeling method based on the aggregation theory is a precondition for realizing the classified aggregation control. Complex systems often contain a large number of variables and parameters, and it is difficult to determine which of these variables and parameters are necessary and which are negligible during the modeling process. The aggregation theory firstly considers a model on a microscopic level, and the influence of various factors on the whole system is considered in the model, so that a model containing a large number of variables and equations is established; and then selecting proper aggregation variables, simplifying a huge model by using a certain aggregation method, namely selecting the aggregation variables, and aggregating a microsystem, so as to obtain a simplified model on a corresponding macroscopic level. The simplified model on the macro level makes the analysis of the behavior characteristics of the system more convenient, and the micro model is beneficial to comprehensively describing the characteristics of the system.
In summary, in order to improve the energy utilization efficiency of multiple users in the load-intensive city smart park, construct the urban power grid dispatching operation ecological service system of the internet +, support the deep construction of the comprehensive energy demand response system, and the demand aggregation modeling method suitable for multiple typical user loads in the load-intensive city smart park is urgently needed to provide a basis for further operation dispatching, so that the overall energy utilization efficiency of the smart park is improved.
Disclosure of Invention
The invention aims at solving the problems in the prior art, and provides a data fuzzy clustering-based load-intensive city intelligent park demand aggregation modeling method so as to optimize load operation of multiple users in the load-intensive city intelligent park and further improve the overall energy efficiency of the intelligent park.
The invention aims at solving the problems through the following technical scheme:
a method for modeling demand aggregation of a load-intensive city intelligent park is characterized by comprising the following steps: the modeling method comprises the following steps:
(1) Inputting a sample data matrix X, and setting n sample sets of load analysis days as follows: x= [ X ] 1 ,X 2 ,···,X n ]Each sample X j With m characteristic indices, i.e. sample X j Can be represented as X j ＝[X j1 ,X j2 ,···,X jm ],j＝1,2,···,n；
(2) Normalization of data, processing of sample data is performed using the following equation:
x' jk ＝(x jk -x kmin )/(x kmax -x kmin ) (7)
in the formula (4), x jk Kth data for the jth sample; x is x kmax And x kmin Respectively the maximum and minimum values of the corresponding components; x's' jk Is normalized data;
(3) Establishing a fuzzy similarity relation matrix R, wherein the fuzzy similarity relation matrix R= { R ij -determining r by absolute value exponential method ij The formula is:
in the formula (5), x' ik For the ith sampleThe normalized data; x's' jk Normalizing the data of the j-th sample;
(4) After a fuzzy equivalent distance matrix R 'is established to obtain a fuzzy similar relation matrix R, constructing a transfer closure R of R by a square self-synthesis method, and modifying the fuzzy similar relation matrix R into a fuzzy equivalent matrix R'; when R' =r, go to step 5); when R ' +.r, circularly establishing a fuzzy equivalent matrix R ' in step 4), until R ' =r, and then entering step 5);
(5) Dynamic clustering is carried out on the intercept matrix, a proper threshold lambda is selected, and fuzzy equivalent matrix R' is cut; the size of the threshold lambda directly influences the clustering result, and when lambda is reduced from 1 to 0, classification is gradually merged by coarse thinning, so that a dynamic cluster map is formed; selection of the optimal lambda value, based on the lambda rate of change C i And (3) determining:
in the formula (6), i is the polymerization sequence number from high to low of lambda; n is n i And n i-1 The number of elements in the ith and the i-1 th clusters respectively; lambda (lambda) i And lambda is i-1 Confidence levels at i and i-1 clustering, respectively;
if C i ＝max(C j ) The confidence level lambda of the ith cluster is considered i Is the optimal threshold.
The obtaining process of the sample data matrix X in the step (1) comprises the following steps:
(11) Determining a source form of load data of the load-intensive city smart park;
(12) Establishing an integration and selection mode of load data of the load-intensive city intelligent park;
(13) Bad data identification and correction of load data of the load-intensive city intelligent park;
(14) Valid data is obtained.
The source form of the load data of the load-intensive city intelligent park in the step (11) comprises: industrial user data, resident user data, public institution data, data center data and other data are respectively analyzed and integrated for various equipment types, field names of collected data, field types and data sources, and required data field collection tables are listed.
The integration and selection of the load data of the load-intensive city intelligent park in the step (12) comprises a data extraction and data analysis process, wherein the data extraction is responsible for acquiring source data available for analysis from various data environments of the park and uploading the source data to a distributed file system in a distributed data storage and calculation platform through a client program; the source data in the distributed file system is loaded into a distributed database of a distributed data storage and calculation platform through analysis and conversion.
The source data includes usage information data, user details, and business data associated therewith.
The bad data identification and correction process in the step (13) is as follows:
(131) Identification process of bad data: firstly, analyzing the validity of a data format of the collected energy basic data, and further judging whether the data is bad data from the perspective of validity of a data value on the basis of the validity of the data format;
(132) And (3) correcting bad data: after bad data detection and identification, finding out a curve P to be detected d P of (2) 1 ,p 2 ,···,p q The load data of the point is bad data, and the corrected value of the bad data of the ith point of the curve to be measured is:
in the formula (3), P c Is the corrected load curve; p (P) t Is a characteristic load curve; k (K) j The j-th measuring point, p of the curve to be measured j Difference coefficients of points; q is the number of bad data measurement points, and q is more than or equal to 1 and less than or equal to n; n is the number of load curve measurements, such as n=24, 48 or 96.
The data format validity analysis in the step (131) includes two angles of a message format and a model format, and the requirements of the validity of the message format are as follows: the acquired energy consumption data must meet the wsdl message format standard, meet the definition of IEC61986 on various message bodies and meet the definition of CIM UML model; the valid requirements of the model format are: the acquired energy consumption data must meet the business scene model, conform to XML definition, conform to characteristic curve and CIM UML definition.
The judging process of the validity of the data value in the step (131) is as follows: firstly, according to the characteristics of various load curves, determining the range of the load value allowable change of various load curves, and judging the load data exceeding the range as bad data.
In the step (131), the difference coefficient K of the ith measurement point of the load curve is determined first i The method comprises the following steps:
in the formula (1), P d (i) The load value of the ith measuring point of the curve to be measured; p (P) t (i) Load values of the ith measuring points of the corresponding similar day characteristic curves; n is the number of load curve measurement points, such as n=24, 48 or 96;
determining the threshold value of the differential coefficient of the load curve as K according to the similar daily load curve al ；
K i ＞K al (12)
When the difference coefficient K of the ith measuring point on the load curve to be measured i And (3) satisfying the formula (2), preliminarily judging that the ith point load data of the load curve to be tested is bad data.
Compared with the prior art, the invention has the following advantages:
according to the method, the utilization efficiency and the accuracy of the data of the load-intensive city smart park are improved by analyzing, integrating, selecting, identifying and correcting the original data in the data source of the energy management platform in the smart park; the built load-intensive city smart park demand aggregation model can fully mine the energy utilization characteristics of the load-intensive city smart park demand data, realize the combination and aggregation classification of multiple types of users in the load-intensive city smart park, apply the demand aggregation modeling result to park energy optimization scheduling, realize the group load optimization operation of multiple types of users in the load-intensive city smart park, reduce the operation and management difficulty of an energy management platform, and improve the overall energy efficiency of the smart park.
Drawings
FIG. 1 is a diagram of a user data acquisition and storage architecture in a load-intensive urban intelligent park demand aggregation modeling method of the present invention;
FIG. 2 is a diagram of a user data analysis process in the load-intensive urban intelligent park demand aggregation modeling method of the present invention;
FIG. 3 is a dimension of user data format validity determination in the load-intensive urban intelligent park demand aggregation modeling method of the present invention;
FIG. 4 is a flow chart of the load-intensive urban intelligent park demand aggregation modeling method of the invention;
FIG. 5 is a graph of daily load for a typical daytime production industrial user for a load-intensive urban smart park in an example;
FIG. 6 is a graph of typical institutional customer daily loads for a load-intensive urban intelligent park in an example;
FIG. 7 is a graph of daily load for a typical two-shift production industry user for a load-intensive urban intelligent park in an example;
figure 8 is a graph of daily load for a typical data center user for a load-intensive urban intelligent park in an example.
Detailed Description
The method comprises the steps of carrying out energy consumption data analysis and data processing on a plurality of terminal users in a certain area, carrying out aggregation classification on the user loads by using a fuzzy clustering technology, and then obtaining a total load predicted value of the area by integrating daily load prediction results of each type, wherein the energy consumption curves of users such as industrial users, data centers, public institutions and the like are different, the demand conditions of cold and hot electric loads are greatly different, and the method is further described below with reference to a drawing and an example.
Establishing a load data source form of a load-intensive city smart park
The energy consumption and the energy supply of various users in the load-intensive city smart park coexist, and various users in the load-intensive city successively enter the smart park, so that the data sources of the energy management platform of the park are divided into the following five types: industrial user data, resident user data, public institution data, data center data and other data are respectively analyzed and integrated with various equipment types, field names, field types and data sources of collected data, required data field collection tables are listed, and the data types and sources of various users are respectively shown in tables 1 to 5.
TABLE 1 Industrial user data types and sources
TABLE 2 resident user data types and sources
TABLE 3 institutional data types and sources
Table 4 data center data types and sources
TABLE 5 other data types and sources
(II) establishing integration and selection mode of load data of load-intensive city intelligent park
The integration and selection of the load data of the load-intensive city intelligent park comprises a data extraction and data analysis process, wherein the data extraction is responsible for acquiring source data available for analysis from various data environments of the park and mainly comprises energy utilization information data, user details and various service data related to the energy utilization information data, and the source data is uploaded to a distributed file system in a distributed data storage and calculation platform through a client program; the source data in the distributed file system is analyzed, converted and loaded into a distributed database HBase of a distributed data storage and calculation platform, and a data acquisition and storage structure is shown in figure 1.
1) Flat File: the GoldenGate tool of Oracle is selected, oracle GoldenGate can capture the changed data from the online log in near real time, and the captured changed data is stored on the distributed file system in the form of Trail format. Oracle GoldenGate for Flat File provides the ability to analyze Trail as Flat File.
2) Data synchronization: and selectively extracting source data from each service system database according to a predefined data extraction specification, and extracting the source data into a distributed file system through a client program.
3) Data analysis conversion: the extracted source data is loaded into a distributed database HBase according to analysis rules by utilizing the distributed data storage and the distributed processing function of the computing platform; the method comprises the following steps: the Flat File uploaded by the client is distributed in the cluster machine in the form of blocks, a Map function is started according to a server where the Flat File Block is located, the Map function is processed according to analysis rules and loaded into HBase, and a data analysis process is shown in figure 2.
(III) bad data identification and correction of load data of load-intensive city intelligent park
And aiming at the collected energy basic data, carrying out data filtering and correction on invalid data and bad data so as to realize accurate presentation and prediction of the data. Firstly, carrying out validity analysis of a data format, namely judging whether the data is valid from two angles of a message format and a model format, wherein the validity judgment dimension of the data format is shown in figure 3; on the basis of the validity of the data format, it is further determined from the viewpoint of the data value whether the data is bad data.
1) Identification process of bad data
1. Data format validity
Data format validity analysis is performed from several dimensions:
(1) Message format
The obtained energy consumption data must meet wsdl message format standards, comply with IEC61986 definition for various message bodies, comply with CIM UML model definition.
(2) Model format
The acquired energy consumption data must meet the business scene model, conform to XML definition, conform to characteristic curve and CIM UML definition.
2. Data value validity
In the invention, when judging the validity of the data value, in order to avoid the error when calculating the load change rate of the statistic historical data, according to the characteristics of various load curves, the power company determines the allowable change range of the load values of various load curves according to the similar daily load curves, and the load data exceeding the range can be initially judged as bad data, wherein the difference coefficient of the ith measuring point of the load curve is as follows:
in the formula (1), P d (i) The load value of the ith measuring point of the curve to be measured; p (P) t (i) Load values of the ith measuring points of the corresponding similar day characteristic curves; n is the number of load curve measurements, such as n=24, 48 or 96.
Electric power companies based on similar typesDaily load curve characteristic, determining the threshold value of the differential coefficient of the load curve as K al 。
The determination of the threshold value of the differential coefficient is affected by various factors, and the power area, the load type, the purpose of bad data identification and correction, the correction method, the experience knowledge and the like need to be considered. If the difference coefficient of the ith measurement point on the load curve to be measured meets the following formula, the ith load data of the load curve to be measured can be initially judged to be bad data. On this basis, the electric power company can further check and confirm whether the point is bad data by other technical means or methods such as field investigation and analysis, and investigate and analyze the cause of the bad data.
K i ＞K al (14)
K in the formula i The method is used for detecting and identifying bad data, is simple and feasible, has small calculated amount and strong practicability, and is used for the difference coefficient of the ith measuring point.
2) Correction process of bad data
After the bad data is detected and identified, the curve P to be detected is found d P of (2) 1 ,p 2 ,···,p q The load data of the point is bad data, and the corrected value of the bad data of the ith point of the curve to be measured is:
in the formula (3), P c Is the corrected load curve; p (P) t Is a characteristic load curve; k (K) j The j-th measuring point, p of the curve to be measured j Difference coefficients of points; q is the number of bad data measurement points, and q is more than or equal to 1 and less than or equal to n; n is the number of load curve measurements, such as n=24, 48 or 96.
(IV) establishing a demand aggregation model of load data of the load-intensive city intelligent park
Screening load data sources of the load-intensive city intelligent park, selecting data integration, and identifying and correcting bad data of the data; according to the similarity of load characteristics of multiple types of users, merging comprehensive loads with similar or similar load characteristics in different load points of the same power grid into one type, describing the classified load characteristics by using the same load model, and modeling a demand aggregation model of load data of the load-intensive city intelligent park by using a dynamic clustering algorithm.
The load is aggregated and modeled by using the load-intensive city intelligent park demand aggregation modeling method described in the summary of the invention, wherein the load aggregation modeling process is shown in fig. 4.
The specific process of demand aggregation modeling is as follows:
1) A matrix of sample data X is input.
Let n sample sets of load analysis day be: x= [ X ] 1 ,X 2 ,···,X n ]Each sample X j With m characteristic indices, i.e. sample X j Can be represented as X j ＝[X j1 ,X j2 ,···,X jm ],j＝1,2,···,n。
2) Normalization of data
The dimension and magnitude of each characteristic index are different, and normalization processing is required to be carried out on the input sample data; the processing of the sample data is performed using the following equation:
x' jk ＝(x jk -x kmin )/(x kmax -x kmin ) (16)
in the formula (4), x jk Kth data for the jth sample; x is x kmax And x kmin Respectively the maximum and minimum values of the corresponding components, x' jk Is normalized data.
3) Establishing a fuzzy similarity relation matrix
To measure the degree of similarity between classified samples, a fuzzy similarity relation matrix R= { R is established ij -determining r by absolute value exponential method ij The formula is:
in the formula (5), the amino acid sequence of the compound,x′ ik normalizing the data of the ith sample; x's' jk Normalized data for the j-th sample.
4) Establishing fuzzy equivalent distance array
After obtaining a fuzzy similarity relation matrix R, constructing a transfer closure R of R (a minimum transfer relation for a relation) by using a square self-synthesis method, and modifying the fuzzy similarity relation matrix R into a fuzzy equivalence matrix R'; when R' =r, go to step 5); when R ' +.r, the fuzzy equivalence matrix R ' is circularly established in step 4), until R ' =r, then step 5 is entered.
5) Dynamic clustering intercept array
Selecting a proper threshold lambda and cutting the fuzzy equivalent matrix R'; selecting a proper threshold lambda and cutting the fuzzy equivalent matrix R'; the magnitude of the threshold lambda directly influences the clustering result, and when lambda is reduced from 1 to 0, classification is gradually merged by coarse thinning, so that a dynamic cluster map is formed.
Selection of the optimal lambda value, based on the lambda rate of change C i And (3) determining:
in the formula (6), i is the polymerization sequence number from high to low of lambda; n is n i And n i-1 The number of elements in the ith and the i-1 th clusters respectively; lambda (lambda) i And lambda is i-1 Confidence levels for i and i-1 clusters, respectively.
If C i ＝max(C j ) The confidence level lambda of the ith cluster is considered i Is the optimal threshold.
Fifthly, setting up an example scene to simulate and describe the effect of the invention
In order to analyze the demand aggregation modeling method of the load-intensive city intelligent park in detail, 16 household electric power users in the intelligent park are taken as example scenes, 24-point load data of the 16 household electric power users are taken as clustering feature vectors, and a plurality of users comprise industrial users, data centers and public institutions, and the daily air temperature is 15 ℃. According to the similarity of the load characteristics of multiple types of users, a dynamic clustering algorithm is used for modeling a demand aggregation model of load data of the load-intensive city intelligent park, the data are imported into a load aggregation modeling program for cluster analysis, and the load aggregation modeling program is divided into 4 types: the first category includes users 1, 6, 10 and 11; the second category includes users 2, 7 and 16; the third class includes users 3, 5, 13 and 15; the fourth class includes users 4, 8, 9, 12 and 14.
According to the result of the load aggregation modeling analysis, daily loads of 18 home power users are classified as shown in fig. 5 to 8.
The first class belongs to typical industrial users, and represents typical peak load curves for daytime production and night rest, the potential of the users for shifting peaks and filling valleys is smaller, and only partial air conditioners and lighting loads can be reduced, so that certain peak shifting capability is realized; the second category belongs to a typical public institution, and represents that the daytime is at a peak of electricity consumption, the coincidence change is small, and the night coincidence is almost zero, and typical industries include office places, supermarkets, shopping malls and the like; the third class belongs to typical industrial users, and represents a typical two-shift reverse enterprise, wherein two power consumption peaks exist according to a curve, and the power consumption at the shift switching period and night is in a valley; the fourth category belongs to a typical data center, and characterizes that the load change of 24 hours a day of a user is not large, only small fluctuation exists, and no electricity consumption peaks and valleys exist.
According to the analysis, the load-intensive urban intelligent park demand aggregate modeling is conducted, one of the most important and critical methods is to introduce node load fuzzy clustering, namely, the node load is classified, a foundation is laid for the follow-up intelligent park energy optimization utilization, a basis is provided for further operation scheduling, and the overall energy efficiency of the intelligent park is further improved.
According to the method, the utilization efficiency and the accuracy of the data of the load-intensive city smart park are improved by analyzing, integrating, selecting, identifying and correcting the original data in the data source of the energy management platform in the smart park; the built load-intensive city smart park demand aggregation model can fully mine the energy utilization characteristics of the load-intensive city smart park demand data, realize the combination and aggregation classification of multiple types of users in the load-intensive city smart park, apply the demand aggregation modeling result to park energy optimization scheduling, realize the group load optimization operation of multiple types of users in the load-intensive city smart park, reduce the operation and management difficulty of an energy management platform, and improve the overall energy efficiency of the smart park.
The above embodiments are only for illustrating the technical idea of the present invention, and the protection scope of the present invention is not limited by the above embodiments, and any modification made on the basis of the technical scheme according to the technical idea of the present invention falls within the protection scope of the present invention; the technology not related to the invention can be realized by the prior art.
Claims (8)
1. A method for modeling demand aggregation of a load-intensive city intelligent park is characterized by comprising the following steps: the modeling method comprises the following steps:
(1) Inputting a sample data matrix X, and setting n sample sets of load analysis days as follows: x= [ X ] 1 ,X 2 ,···,X n ]Each sample X j With m characteristic indices, i.e. sample X j Can be represented as X j ＝[X j1 ,X j2 ,···,X jm ]J=1, 2, carrying out the following steps; the source data adopted by the sample data matrix X comprises energy utilization information data, user details and service data related to the energy utilization information data in the park;
(2) Normalization of data, processing of sample data is performed using the following equation:
x' jk ＝(x jk -x kmin )/(x kmax -x kmin ) (1)
in the formula (4), x jk Kth data for the jth sample; x is x kmax And x kmin Respectively the maximum and minimum values of the corresponding components; x's' jk Is normalized data;
(3) Establishing a fuzzy similarity relation matrix R, wherein the fuzzy similarity relation matrix R= { R ij -determining r by absolute value exponential method ij The formula is:
in the formula (5), x' ik Normalizing the data of the ith sample; x's' jk Normalizing the data of the j-th sample;
(4) After a fuzzy equivalent distance matrix R 'is established to obtain a fuzzy similar relation matrix R, constructing a transfer closure R of R by a square self-synthesis method, and modifying the fuzzy similar relation matrix R into a fuzzy equivalent matrix R'; when R' =r, go to step 5); when R ' +.r, circularly establishing a fuzzy equivalent matrix R ' in step 4), until R ' =r, and then entering step 5);
(5) Dynamic clustering is carried out on the intercept matrix, a proper threshold lambda is selected, and fuzzy equivalent matrix R' is cut; the size of the threshold lambda directly influences the clustering result, and when lambda is reduced from 1 to 0, classification is gradually merged by coarse thinning, so that a dynamic cluster map is formed; selection of the optimal lambda value, based on the lambda rate of change C i And (3) determining:
in the formula (6), i is the polymerization sequence number from high to low of lambda; n is n i And n i-1 The number of elements in the ith and the i-1 th clusters respectively; lambda (lambda) i And lambda is i-1 Confidence levels at i and i-1 clustering, respectively;
if C i ＝max(C j ) The confidence level lambda of the ith cluster is considered i Is the optimal threshold.
2. The load-intensive urban intelligent park demand aggregate modeling method of claim 1, wherein: the obtaining process of the sample data matrix X in the step (1) comprises the following steps:
(11) Determining a source form of load data of the load-intensive city smart park;
(12) Establishing an integration and selection mode of load data of the load-intensive city intelligent park;
(13) Bad data identification and correction of load data of the load-intensive city intelligent park;
(14) Valid data is obtained.
3. The load-intensive city intelligent park demand aggregate modeling method of claim 2, wherein: the source form of the load data of the load-intensive city intelligent park in the step (11) comprises: industrial user data, resident user data, public institution data, data center data and other data are respectively analyzed and integrated for various equipment types, field names of collected data, field types and data sources, and required data field collection tables are listed.
4. The load-intensive city intelligent park demand aggregate modeling method of claim 2, wherein: the integration and selection of the load data of the load-intensive city intelligent park in the step (12) comprises a data extraction and data analysis process, wherein the data extraction is responsible for acquiring source data available for analysis from various data environments of the park and uploading the source data to a distributed file system in a distributed data storage and calculation platform through a client program; the source data in the distributed file system is loaded into a distributed database of a distributed data storage and calculation platform through analysis and conversion.
5. The load-intensive city intelligent park demand aggregate modeling method of claim 2, wherein: the bad data identification and correction process in the step (13) is as follows:
(131) Identification process of bad data: firstly, analyzing the validity of a data format of the collected energy basic data, and further judging whether the data is bad data from the perspective of validity of a data value on the basis of the validity of the data format;
(132) And (3) correcting bad data: after bad data detection and identification, finding out a curve P to be detected d P of (2) 1 ,p 2 ,···,p q The load data of the point is bad data, and the corrected value of the bad data of the ith point of the curve to be measured is:
in the formula (3), P c Is the corrected load curve; p (P) t Is a characteristic load curve; k (K) j The j-th measuring point, p of the curve to be measured j Difference coefficients of points; q is the number of bad data measurement points, and q is more than or equal to 1 and less than or equal to n; n is the number of load curve measurements, such as n=24, 48 or 96.
6. The load-intensive city intelligent park demand aggregate modeling method of claim 5, wherein: the data format validity analysis in the step (131) includes two angles of a message format and a model format, and the requirements of the validity of the message format are as follows: the acquired energy consumption data must meet the wsdl message format standard, meet the definition of IEC61986 on various message bodies and meet the definition of CIM UML model; the valid requirements of the model format are: the acquired energy consumption data must meet the business scene model, conform to XML definition, conform to characteristic curve and CIM UML definition.
7. The load-intensive city intelligent park demand aggregate modeling method of claim 5, wherein: the judging process of the validity of the data value in the step (131) is as follows: firstly, according to the characteristics of various load curves, determining the range of the load value allowable change of various load curves, and judging the load data exceeding the range as bad data.
8. The load-intensive city intelligent park demand aggregate modeling method of claim 5 or 7, wherein: in the step (131), the difference coefficient K of the ith measurement point of the load curve is determined first i The method comprises the following steps:
in the formula (1), P d (i) The load value of the ith measuring point of the curve to be measured; p (P) t (i) Load values of the ith measuring points of the corresponding similar day characteristic curves; n is the number of load curve measurement points, such as n=24, 48 or 96;
determining the threshold value of the differential coefficient of the load curve as K according to the similar daily load curve al ；
K i ＞K al (6)
When the difference coefficient K of the ith measuring point on the load curve to be measured i And (3) satisfying the formula (2), preliminarily judging that the ith point load data of the load curve to be tested is bad data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011321937.0A CN112257964B (en) | 2020-11-23 | 2020-11-23 | Load-intensive urban intelligent park demand aggregation modeling method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011321937.0A CN112257964B (en) | 2020-11-23 | 2020-11-23 | Load-intensive urban intelligent park demand aggregation modeling method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112257964A CN112257964A (en) | 2021-01-22 |
| CN112257964B true CN112257964B (en) | 2024-04-09 |
Family
ID=74225342
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011321937.0A Active CN112257964B (en) | 2020-11-23 | 2020-11-23 | Load-intensive urban intelligent park demand aggregation modeling method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112257964B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114416869A (en) * | 2021-12-31 | 2022-04-29 | 新奥数能科技有限公司 | Load aggregation scheduling platform and system |
| CN117392853B (en) * | 2023-12-11 | 2024-04-12 | 山东通维信息工程有限公司 | Big data intelligent lane control system based on high in clouds |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105069519A (en) * | 2015-07-16 | 2015-11-18 | 国网天津市电力公司 | Intelligent power grid park terminal user energy demand condition dynamic prediction system and method |
| CN105631764A (en) * | 2015-12-31 | 2016-06-01 | 国网电力科学研究院武汉南瑞有限责任公司 | Smart power grid big data application system orienting smart city |
| CN108110766A (en) * | 2018-01-23 | 2018-06-01 | 华翔翔能电气股份有限公司 | A kind of transformer load management method and device towards wisdom garden |
| WO2019174040A1 (en) * | 2018-03-16 | 2019-09-19 | 大连理工大学 | Coupling and clustering analysis and decision-making tree based short-term power generation scheduling method for cascaded hydroelectric station group |
| CN110879547A (en) * | 2019-11-07 | 2020-03-13 | 上海航天智慧能源技术有限公司 | Complementary garden comprehensive energy supply system of multipotency |
-
2020
- 2020-11-23 CN CN202011321937.0A patent/CN112257964B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105069519A (en) * | 2015-07-16 | 2015-11-18 | 国网天津市电力公司 | Intelligent power grid park terminal user energy demand condition dynamic prediction system and method |
| CN105631764A (en) * | 2015-12-31 | 2016-06-01 | 国网电力科学研究院武汉南瑞有限责任公司 | Smart power grid big data application system orienting smart city |
| CN108110766A (en) * | 2018-01-23 | 2018-06-01 | 华翔翔能电气股份有限公司 | A kind of transformer load management method and device towards wisdom garden |
| WO2019174040A1 (en) * | 2018-03-16 | 2019-09-19 | 大连理工大学 | Coupling and clustering analysis and decision-making tree based short-term power generation scheduling method for cascaded hydroelectric station group |
| CN110879547A (en) * | 2019-11-07 | 2020-03-13 | 上海航天智慧能源技术有限公司 | Complementary garden comprehensive energy supply system of multipotency |
Non-Patent Citations (1)
| Title |
|---|
| 探析火灾自动报警系统设计;周健;魅力中国;20151231(第014期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112257964A (en) | 2021-01-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116646933B (en) | Big data-based power load scheduling method and system | |
| Li et al. | When weather matters: IoT-based electrical load forecasting for smart grid | |
| CN110097297B (en) | Multi-dimensional electricity stealing situation intelligent sensing method, system, equipment and medium | |
| Teichgraeber et al. | Time-series aggregation for the optimization of energy systems: Goals, challenges, approaches, and opportunities | |
| CN103559655B (en) | The Forecasting Methodology of the novel feeder line load of microgrid based on data mining | |
| CN112257964B (en) | Load-intensive urban intelligent park demand aggregation modeling method | |
| CN110782153A (en) | Modeling method and system for comprehensive energy efficiency assessment system of enterprise park | |
| CN110796307A (en) | Distributed load prediction method and system for comprehensive energy system | |
| CN111754037A (en) | Long-term load hybrid prediction method for regional terminal integrated energy supply system | |
| CN112149890A (en) | Comprehensive energy load prediction method and system based on user energy label | |
| CN114764521A (en) | Garden building shape optimization method and system based on genetic algorithm | |
| CN117200217A (en) | Power system scheduling method based on load classification | |
| CN110991638A (en) | Generalized load modeling method based on clustering and neural network | |
| CN114298165B (en) | Construction method of regional comprehensive energy system multipotency load situation awareness model | |
| Yu | Mining hidden knowledge from measured data for improving building energy performance | |
| CN115829418A (en) | Power consumer load characteristic portrait construction method and system suitable for load management | |
| CN115730731A (en) | Automatic identification method and display platform for urban high-carbon emptying room unit | |
| Gong et al. | An energy efficiency solution based on time series data mining algorithm on elementary school building | |
| CN114662922A (en) | Resident demand response potential evaluation method and system considering photovoltaic uncertainty | |
| CN111815022A (en) | Power load prediction method based on time-delay coordinate embedding method | |
| CN117220417B (en) | Dynamic monitoring method and system for consumer-side electrical load | |
| Borretta et al. | Software-based solutions for the optimization of a building electric bill using integrated PV and storage systems: a case study | |
| CN104751233A (en) | Contract capacity optimization system and optimization method | |
| Fahad et al. | Big Data Aggregation, Visualization and Clustering for Smart Grid in Smart City using Machine Learning | |
| Basnet et al. | Predictive Analytics to Estimate Level of Residential Participation in Residential Demand Response Program |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |