CN113628066B - Multi-dimensional feature vector data loose coupling extraction method for wind generating set - Google Patents
Multi-dimensional feature vector data loose coupling extraction method for wind generating set Download PDFInfo
- Publication number
- CN113628066B CN113628066B CN202110719126.4A CN202110719126A CN113628066B CN 113628066 B CN113628066 B CN 113628066B CN 202110719126 A CN202110719126 A CN 202110719126A CN 113628066 B CN113628066 B CN 113628066B
- Authority
- CN
- China
- Prior art keywords
- fan
- data
- measurement point
- time
- feature vector
- 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
- 239000013598 vector Substances 0.000 title claims abstract description 32
- 230000008878 coupling Effects 0.000 title claims abstract description 8
- 238000010168 coupling process Methods 0.000 title claims abstract description 8
- 238000005859 coupling reaction Methods 0.000 title claims abstract description 8
- 238000000605 extraction Methods 0.000 title claims description 7
- 238000013507 mapping Methods 0.000 claims abstract description 38
- 238000005259 measurement Methods 0.000 claims abstract description 33
- 238000013500 data storage Methods 0.000 claims abstract description 20
- 238000000034 method Methods 0.000 claims abstract description 10
- 230000017525 heat dissipation Effects 0.000 description 8
- 238000005070 sampling Methods 0.000 description 4
- 238000013075 data extraction Methods 0.000 description 3
- 238000012544 monitoring process Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000001419 dependent effect Effects 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- 238000012952 Resampling Methods 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 101150086656 dim1 gene Proteins 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000004927 fusion Effects 0.000 description 1
- 238000004519 manufacturing process Methods 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
- 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
- G06F16/22—Indexing; Data structures therefor; Storage structures
- G06F16/2282—Tablespace storage structures; Management thereof
-
- 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
- G06F16/24—Querying
- G06F16/245—Query processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/253—Fusion techniques of extracted features
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Biology (AREA)
- Public Health (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computational Linguistics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Software Systems (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Water Supply & Treatment (AREA)
- General Health & Medical Sciences (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Testing Or Calibration Of Command Recording Devices (AREA)
Abstract
The invention discloses a method for loosely coupling and extracting multidimensional feature vector data of a wind generating set, which comprises the following steps: 1) According to the measurement point data storage format of each fan, a mapping system between the multidimensional characteristics of the prediction target and the measurement point data storage information stored in a scattered manner is established, and the measurement point data of each fan are stored in a history table; (2) Configuring a mapping information table according to the predicted targets, and establishing a one-to-one correspondence between the predicted targets and the mapping information table; (3) And acquiring a multidimensional feature vector data set of the wind generating set prediction target according to the prediction target, each fan ID, the starting time, the ending time and the time interval. According to the invention, object modeling and data source definition required by various application scenes in wind power application business are integrated, so that the contradiction between the feature measurement point scattered storage of a data storage layer and the multi-dimensional feature vector data set requiring semanteme of the application layer is solved.
Description
Technical Field
The invention relates to a method for loosely coupled fusion extraction of multidimensional feature vector data, in particular to a method for loosely coupled extraction of multidimensional feature vector data of a wind generating set.
Background
In the wind power monitoring system, the fan equipment objects of the same type have multiple types, and the characteristic measuring points of each type are different, so that the system cannot solidify the semantic characteristics of the measuring points according to the types of the fan equipment objects; in the data center, different characteristic data of the same fan object originate from different subsystems, and the semantic characteristics of the data are not solidified because the data are compatible with subsystem data newly accessed subsequently. Meanwhile, for specific fan objects, feature dimensions used in different application scenes are different. For the above reasons, in the monitoring system and the data center, a distributed storage method is generally adopted for each feature measurement point, and each feature measurement point value at each sampling time is stored as a single record. However, in an object-oriented application scenario, feature vector data of different dimensions of a fan object are required, and the number of feature measurement points of the fan is generally hundreds, but in the application of heat dissipation analysis of a generator of the fan and heat dissipation analysis of a variable pitch motor, a part of features of different dimensions can be used. How to extract feature vector data sets with different dimensions according to needs based on data stored in a scattered manner is a technical problem to be solved.
Disclosure of Invention
The invention aims to: the invention aims to provide a method for extracting object multidimensional feature vector data loose coupling of a corresponding feature vector data set by fusing and extracting data stored in a scattered manner according to object multidimensional features required by configuration according to specific requirements.
The technical scheme is as follows: the invention relates to a method for loosely coupling and extracting multidimensional feature vector data, which comprises the following steps:
(1) According to the measurement point data storage format of each fan, a mapping system between the multidimensional characteristics of the prediction target and the measurement point data storage information stored in a scattered manner is established, and the measurement point data of each fan are stored in a history table;
(2) Configuring a mapping information table according to the predicted targets, and establishing a one-to-one correspondence between the predicted targets and the mapping information table;
(3) And acquiring a multidimensional feature vector data set of the wind generating set prediction target according to the prediction target, each fan ID, the starting time, the ending time and the time interval.
Further, in the step (1), the history table name stored in the measurement point data of each fan is composed of a table model name and a sub-table time; the table model names are named by a specific application under the condition of ensuring that the table names do not conflict; the sub-table time is to construct a new table according to the period of year/month/day, and store the application data of different time periods into different history tables.
Further, in the step (2), according to the requirement of each prediction target, configuring the dimension of the fan object and a data source; under the requirement of each prediction target, the names of mapping information tables between the multidimensional features of the fan object and the scattered and stored measurement point historical data storage tables are different but have the same structure; and configuring the characteristics and the numerical information of the measuring points corresponding to the characteristics in a mapping information table according to the characteristic dimension required by each fan object under the specific prediction target requirement.
Further, in the step (3), the prediction target is embodied in a mapping information table name manner, and the step of obtaining the multidimensional feature vector data set of the fan object is as follows:
(31) Obtaining all features of a fan object and measuring point data information corresponding to the features in a mapping information table according to the name of the mapping information table and the ID of the fan object;
(32) Acquiring data of each feature in the query time range according to the measurement point data information of each feature, and organizing query sentences of a query relation library;
(33) Repeating the step (32) to obtain time sequence data of all the characteristics of the fan object;
(34) And transversely combining the time sequence data of each feature by taking time as an index to form a two-dimensional data set which takes time as a row index and takes dimensions as a column index, wherein the two-dimensional data set is a required fan object multidimensional feature vector data set.
Compared with the prior art, the invention has the following remarkable effects: 1. according to the specific requirements, configuring the required multidimensional features of the fan object, fusing and extracting the data stored in a scattered manner into corresponding feature vector data sets, and providing a general multidimensional feature vector data set extraction framework; 2. object modeling and data source definition required by various application scenes in wind power application business are integrated, so that the contradiction between the feature measurement point scattered storage of a data storage layer and the multi-dimensional feature vector data set needing semanteme of an application layer is solved, and the realization of application functions is facilitated.
Drawings
FIG. 1 is a data flow diagram of the present invention.
Detailed Description
The invention is described in further detail below with reference to the drawings and the detailed description.
Taking loose coupling extraction of multidimensional feature vector data in the wind power field as an example, as shown in fig. 1, configuring required object multidimensional features according to specific requirements, and fusing and extracting scattered stored data into corresponding feature vector data sets; starting from the format characteristics of the measurement point historical data storage, a mapping system between a set of fan object multidimensional semantic features and the scattered stored measurement point historical data storage information is provided, and a set of fan object multidimensional features definition and data extraction flow is designed according to the mapping system, and the method specifically comprises the following three parts:
(1) Starting from a measuring point data storage format, a mapping system between multidimensional semantic features of a fan object and scattered and stored measuring point historical data storage information is established;
(2) Based on the mapping system, configuring a specific mapping information table in a specific application scene, and establishing a one-to-one correspondence between the application scene and a mapping information set;
(3) In the data extraction link, a fan object multidimensional feature vector data set is obtained according to an application scene, a fan object ID, a start time, an end time and a time interval and is used for object-oriented data processing.
Firstly, a mapping system between a set of multidimensional semantic features of a fan object and a scattered measurement point historical data storage table is established from a measurement point data storage format. In the production environment, the historical table names of the fan measuring point data storage are composed of table model names and sub-table time, wherein the table model names are named by a user under the condition that the table names are ensured not to conflict, the sub-table time is used for avoiding the accumulation along with time, the operation efficiency is influenced by excessive number of single data table entries, a new table is built according to the period of year/month/day and the like, and the data of different time periods are stored in different historical tables. For example, the table model name of the fan measuring point history sampling record is analog, and the table is divided by month, and the actual scattered and stored history data table names are analog202001, analog202002 and …. The main attributes contained in the fan measurement point history table are shown in table 1 as an example of the format; the mapping relation between the multidimensional semantic features of the designed fan object and the scattered stored historical data storage table of the fan measuring points is shown in table 2.
Table 1 fan measurement Point data sheet Property example
Attribute meaning | Whether or not to use the primary key |
Fan measurement point ID | Y |
Data time stamp | Y |
Data value |
Table 2 definition of mapping relationship between multidimensional semantic features of fan objects and measurement point historic data storage tables
Secondly, the invention can flexibly configure the dimension and the data source of the fan object in each advanced application scene according to the requirement of the scene. In each scene, the names of mapping information tables (hereinafter referred to as mapping information tables) between the multidimensional semantic features of the fan object and the measurement point history data storage tables stored in a scattered manner are different, but the table structures are the same (the table structures of table 2). According to the feature dimension required by each fan object in a specific scene, the storage information of the features and the corresponding measurement points thereof can be configured in the mapping information table, the feature dimension is not limited by the application scene layer and the fan object layer, loose coupling feature definition and measurement point numerical value source definition can be performed under the granularity of the fan object in the specific application scene, and application object modeling and data source definition are integrated.
Taking intelligent pre-warning of heat dissipation of a converter of a fan of a certain model in a certain wind field as an example, taking the measured point 'motor side inverter temperature' as a prediction target, periodically and circularly predicting the temperature value of the converter for 30 minutes in the future, so as to pre-warn in advance before the temperature exceeds the limit. The number of the measuring points of the model fan is 110, and the measuring points related to the predicted target are shown in a table 3 after business and data analysis, and are feature vectors of heat dissipation prediction of the model fan converter; the sampling history data format examples of the measuring points of the fans are shown in table 4, and each measuring point of each fan has 1 unique identification ID in a monitoring system; and constructing the mapping relation between the characteristic vector of each fan converter heat dissipation intelligent early warning and the measuring point based on the data, wherein the mapping relation is shown in table 5.
Table 3 heat dissipation prediction independent variable (eigenvector) and dependent variable of fan converter of a certain model
Sequence number | Measuring point | Variable type |
1 | Grid side inverter temperature | Independent variable |
2 | Network side reactor temperature | Independent variable |
3 | Converter brake unit temperature | Independent variable |
4 | Converter ambient temperature | Independent variable |
5 | Wind speed | Independent variable |
6 | Active power of converter | Independent variable |
7 | Cabin temperature | Independent variable |
8 | Water temperature of water-cooled outlet valve | Independent variable |
9 | Water temperature of water-cooled inlet valve | Independent variable |
10 | Motor side inverter temperature | Dependent variable |
Table 4 fan measurement point sampling history data format examples
Table 5 mapping relation between characteristic vector and measuring point of intelligent heat dissipation early warning of fan converter
Finally, in the data extraction link, a fan object multidimensional feature vector data set is obtained according to the application scene, the object ID, the starting time, the ending time and the time interval and is used for object-oriented data processing. The application scene is embodied in a mode of mapping information table names. In the case of setting the foregoing fan converter heat dissipation intelligent early warning case, the mapping relationship table of table 5 is named convdispeat_info_cfg, for the #3 fan object (ID is 3), taking the start time=2021, 4, 9, 00, the end time=2021, 4, 19, 00, and the time interval=900 seconds, the steps of obtaining the object multidimensional feature vector data set are as follows:
and step 31, acquiring all the characteristics of the fan object and the measuring point data information corresponding to the characteristics in the mapping information table according to the input mapping information table and the fan object ID.
Step 32, according to the measurement point data information of each feature, obtaining the data of the feature in the query time range, taking feature dimension 1 of fan object 3 in search table 5 as an example, a query statement of the query relation library can be organized, and the schematic format of the sql statement is as follows:
Select fvalue from analog202104where attr_oid=101and attr_time>=“2021-4-9
00:00:00”and attr_time<=“2021-4-19 00:00:00”
the queried data is time sequence data with 600 seconds as interval, and then is converted into 900 seconds as interval time sequence data according to interpolation/resampling mode.
Step 33, in the same manner as step 32, time series data of the feature 1 and the feature 2 of the fan object 1 are acquired.
Step 34, the time sequence data of each feature of the fan object is transversely combined by taking time as an index to form a two-dimensional data set by taking time as a row index and taking dimension as a column index, and an example format is shown in table 6, namely the data set required by the advanced application scene.
Table 6#3 blower multidimensional feature vector data set example
index | dim1 | … | dim6 | … | dim10 |
2021/4/9 00:00 | 51.2 | 1300 | 58.5 | ||
2021/4/9 00:15 | xx | xx | xx | ||
2021/4/9 00:30 | xx | xx | xx | ||
… | … | … | … | ||
2021/4/18 23:45 | xx | xx | xx | ||
2021/4/19 00:00 | xx | xx | xx |
Claims (2)
1. A method for loosely coupling and extracting multidimensional feature vector data of a wind generating set is characterized by comprising the following steps:
(1) According to the measurement point data storage format of each fan, a mapping system between the multidimensional characteristics of the prediction target and the measurement point data storage information stored in a scattered manner is established, and the measurement point data of each fan are stored in a history table;
(2) Configuring a mapping information table according to the predicted targets, and establishing a one-to-one correspondence between the predicted targets and the mapping information table; the detailed implementation process is as follows:
configuring the dimension and the data source of a fan object according to the requirement of each prediction target; under the requirement of each prediction target, the names of mapping information tables between the multidimensional features of the fan object and the scattered and stored measurement point historical data storage tables are different and have the same structure; according to the feature dimension required by each fan object under the specific prediction target requirement, configuring the feature and the numerical information of the measuring point corresponding to the feature in a mapping information table;
(3) According to the prediction targets, the ID of each fan, the starting time, the ending time and the time interval, a multidimensional feature vector data set of the wind generating set prediction targets is obtained; the prediction target is embodied in a mode of mapping information table names, and the step of obtaining the multidimensional feature vector data set of the fan object is as follows:
(31) Obtaining all features of a fan object and measuring point data information corresponding to the features in a mapping information table according to the name of the mapping information table and the ID of the fan object;
(32) Acquiring data of each feature in the query time range according to the measurement point data information of each feature, and organizing query sentences of a query relation library;
(33) Repeating the step (32) to obtain time sequence data of all the characteristics of the fan object;
(34) And transversely combining the time sequence data of each feature by taking time as an index to form a two-dimensional data set which takes time as a row index and takes dimensions as a column index, wherein the two-dimensional data set is a required fan object multidimensional feature vector data set.
2. The method for loosely coupled extraction of multidimensional feature vector data of wind generating set according to claim 1, wherein in the step (1), a history table name stored in measurement point data of each fan consists of a table model name and a sub-table time; the table model names are named by a specific application under the condition of ensuring that the table names do not conflict; the sub-table time is to construct a new table according to the period of year/month/day, and store the application data of different time periods into different history tables.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110719126.4A CN113628066B (en) | 2021-06-28 | 2021-06-28 | Multi-dimensional feature vector data loose coupling extraction method for wind generating set |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110719126.4A CN113628066B (en) | 2021-06-28 | 2021-06-28 | Multi-dimensional feature vector data loose coupling extraction method for wind generating set |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113628066A CN113628066A (en) | 2021-11-09 |
CN113628066B true CN113628066B (en) | 2024-03-19 |
Family
ID=78378454
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110719126.4A Active CN113628066B (en) | 2021-06-28 | 2021-06-28 | Multi-dimensional feature vector data loose coupling extraction method for wind generating set |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113628066B (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102262563A (en) * | 2011-08-09 | 2011-11-30 | 南京南瑞继保电气有限公司 | Method for data interaction and modeling of dispatching subsystem |
CN103309977A (en) * | 2013-06-14 | 2013-09-18 | 广东电网公司电力科学研究院 | Heterogeneous data resource integration method |
CN104200402A (en) * | 2014-09-11 | 2014-12-10 | 国家电网公司 | Publishing method and system of source data of multiple data sources in power grid |
CN111598296A (en) * | 2019-10-16 | 2020-08-28 | 中国南方电网有限责任公司 | Power load prediction method, power load prediction device, computer equipment and storage medium |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10540458B2 (en) * | 2016-04-26 | 2020-01-21 | Sejong Industry-Academia Cooperation Foundation Hongik University | System and method for monitoring photovoltaic power generation |
-
2021
- 2021-06-28 CN CN202110719126.4A patent/CN113628066B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102262563A (en) * | 2011-08-09 | 2011-11-30 | 南京南瑞继保电气有限公司 | Method for data interaction and modeling of dispatching subsystem |
CN103309977A (en) * | 2013-06-14 | 2013-09-18 | 广东电网公司电力科学研究院 | Heterogeneous data resource integration method |
CN104200402A (en) * | 2014-09-11 | 2014-12-10 | 国家电网公司 | Publishing method and system of source data of multiple data sources in power grid |
CN111598296A (en) * | 2019-10-16 | 2020-08-28 | 中国南方电网有限责任公司 | Power load prediction method, power load prediction device, computer equipment and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN113628066A (en) | 2021-11-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110618983B (en) | JSON document structure-based industrial big data multidimensional analysis and visualization method | |
CN105357311B (en) | A kind of storage of secondary device big data and processing method of cloud computing technology | |
CN102750406B (en) | Multi-version management method for model set and difference model-based power grid model | |
CN108509566B (en) | Network topology visualization method based on-cloud 95598 data release service system | |
CN108833523A (en) | A kind of heat supply wisdom operating system based on Internet of Things | |
Boussuge et al. | Capturing simulation intent in an ontology: CAD and CAE integration application | |
CN113221297B (en) | Method for converting power grid information model into FBX three-dimensional model and storing attributes | |
CN112783901A (en) | Internet of things time sequence big data processing method based on Internet of things middleware | |
CN115115796A (en) | Building digital twin frame based on semantic Web technology and modeling method | |
CN108228726A (en) | The increment unusual fluctuation content acquisition method and storage medium of power distribution network red and black figure | |
CN112580319A (en) | Data processing method, device, equipment and computer readable storage medium | |
CN113628066B (en) | Multi-dimensional feature vector data loose coupling extraction method for wind generating set | |
CN110502543B (en) | Equipment performance data storage method, device, equipment and storage medium | |
CN104299065A (en) | Dispatching automation host and backup model correctness checking method | |
CN111078731A (en) | Hbase-based power grid operation data collaborative query method and device and storage medium | |
CN116826709A (en) | Automatic matching method, device and storage medium for adjustable load resource and scheduling master station model | |
CN116319892A (en) | Data processing method and device based on object model and electronic equipment | |
CN115309747A (en) | Fire fighting management method and platform based on spatial grid data and electronic equipment | |
CN114493088A (en) | Information model construction method for electric power internet of things terminal equipment | |
CN113342807A (en) | Knowledge graph based on mixed database and construction method thereof | |
CN112001008A (en) | Bridge construction monitoring data integrated management system based on Revit software | |
CN117592315B (en) | Battery design management method, device, system, electronic equipment and storage medium | |
CN117271480B (en) | Data processing method, device, electronic equipment and medium | |
CN114416738B (en) | Data aggregation method and device based on relational database | |
Hislop et al. | Issues in mechanical engineering design management |
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 |