CN109272564B - Plant station diagram automatic generation method based on machine learning - Google Patents

Plant station diagram automatic generation method based on machine learning Download PDF

Info

Publication number
CN109272564B
CN109272564B CN201811108036.6A CN201811108036A CN109272564B CN 109272564 B CN109272564 B CN 109272564B CN 201811108036 A CN201811108036 A CN 201811108036A CN 109272564 B CN109272564 B CN 109272564B
Authority
CN
China
Prior art keywords
voltage side
characteristic
interval
main
graphic
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
Application number
CN201811108036.6A
Other languages
Chinese (zh)
Other versions
CN109272564A (en
Inventor
罗俊
封�波
刘翌
熊浩
蒋宇
范青
王元
徐善荣
江华
姜骞
苏运光
葛王飞
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
State Grid Corp of China SGCC
Nari Technology Co Ltd
NARI Nanjing Control System Co Ltd
Lianyungang Power Supply Co of State Grid Jiangsu Electric Power Co Ltd
Original Assignee
State Grid Corp of China SGCC
Nari Technology Co Ltd
NARI Nanjing Control System Co Ltd
Lianyungang Power Supply Co of State Grid Jiangsu Electric Power Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by State Grid Corp of China SGCC, Nari Technology Co Ltd, NARI Nanjing Control System Co Ltd, Lianyungang Power Supply Co of State Grid Jiangsu Electric Power Co Ltd filed Critical State Grid Corp of China SGCC
Priority to CN201811108036.6A priority Critical patent/CN109272564B/en
Publication of CN109272564A publication Critical patent/CN109272564A/en
Application granted granted Critical
Publication of CN109272564B publication Critical patent/CN109272564B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T11/002D [Two Dimensional] image generation
    • G06T11/20Drawing from basic elements, e.g. lines or circles
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T11/002D [Two Dimensional] image generation
    • G06T11/60Editing figures and text; Combining figures or text

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Supply And Distribution Of Alternating Current (AREA)

Abstract

The invention discloses a plant station diagram automatic generation method based on machine learning, which comprises the steps of carrying out topology analysis on stock graphs, identifying main characteristics and auxiliary characteristics of a wiring diagram, writing the main characteristics into a graph characteristic library file after carrying out index coding, and writing the auxiliary characteristics and a main characteristic index into a characteristic description file; inputting target graphic parameters, and performing characterization processing on the parameters to form characteristic parameters; comparing and evaluating the formed characteristic parameters with the graphic characteristics of a graphic characteristic library to obtain the graphic characteristics most similar to the characteristic parameters; and multiplexing, modifying and expanding the obtained graphic feature units to complete the layout and wiring of the graphics. The method can generate the accurate high-quality plant station wiring diagram which conforms to the drawing habits of users and integrates the diagram models, effectively solves a series of problems that static drawings are large in operation and maintenance workload, prone to errors and the like in dispatching automation application, and provides auxiliary technical support for automatic operation and maintenance of the power grid plant station wiring diagram.

Description

Plant station diagram automatic generation method based on machine learning
Technical Field
The invention belongs to the technical field of power systems, and particularly relates to a plant station diagram automatic generation method based on machine learning.
Background
The power grid station wiring diagram is very important for operation management personnel of a power grid, and the operation management personnel can intuitively manage, schedule and handle accidents and the like for the power grid through the power grid station diagram. At present, a plant station wiring diagram is mostly drawn manually, model information of each electrical device is firstly established in a database through a dbi tool, then a power grid plant station wiring diagram is drawn through a manual drawing tool provided in an EMS, and finally association between the device in the diagram and the model device in the database is achieved through a retriever. Although the manually drawn tidal current diagram is more in line with the cognitive habits of dispatching operation personnel, the maintenance workload is overlarge in the mode, the problems of untimely and inaccurate performances and the like can occur, and inconvenience is brought to the work of the dispatching operation personnel of the power grid.
In many academic institutions and scientific research institutes in China, a lot of researches for automatically generating graphs of power systems mainly focus on the aspects of automatic layout of distribution network single line diagrams and power grid tidal current diagrams, and relatively few researches for automatically generating station diagrams are performed. The plant station wiring diagram is multiple in equipment types, complex in bus wiring mode, and each place has own drawing habit, and based on the existing topological model information, the power grid plant station wiring diagram is realized by simply utilizing a layout and wiring algorithm, so that the application requirement is difficult to meet and manual drawing is replaced.
Disclosure of Invention
The invention aims to provide a plant diagram automatic generation method based on machine learning.
The technical solution for realizing the purpose of the invention is as follows: a plant station diagram automatic generation method based on machine learning comprises the following steps:
step 1, establishing a characteristic model: performing topology analysis on the stock graph, identifying main characteristics and auxiliary characteristics of the wiring graph, performing index coding on the main characteristics, writing the main characteristics into a graph characteristic library file, and writing the auxiliary characteristics and the main characteristic index into a characteristic description file;
step 2, generating characteristic parameters of the target graph: inputting target graphic parameters, and performing characterization processing on the parameters to form characteristic parameters;
step 3, characteristic evaluation: comparing and evaluating the formed characteristic parameters with the graphic characteristics of a graphic characteristic library to obtain the graphic characteristics most similar to the characteristic parameters;
step 4, generating and drawing a graph: and multiplexing, modifying and expanding the obtained graphic feature units to complete the layout and wiring of the graphics.
Compared with the prior art, the invention has the following remarkable advantages: according to the invention, the machine learning of the stock graph is adopted, the graph drawing is realized by utilizing the graph characteristics, the high-quality plant station wiring diagram which is accurate and accords with the drawing habit of the user and is integrated with the graph model is generated, a series of problems of large workload, high error possibility and the like of the static drawing operation and maintenance in the dispatching automation application are effectively solved, and an auxiliary technical support is provided for the automation operation and maintenance of the power grid plant station wiring diagram.
Drawings
FIG. 1 is a schematic diagram of an automatic factory floor diagram generation framework based on machine learning according to the invention.
FIG. 2 is a diagram of the architecture for generating a graphic feature library according to the present invention.
FIG. 3 is a schematic diagram of a graph generation process of the present invention.
Detailed Description
The invention is further illustrated by the following examples in conjunction with the accompanying drawings.
As shown in fig. 1, a plant diagram automatic generation method based on machine learning includes the following steps:
step 1, establishing a characteristic model: performing topology analysis on the stock graph, identifying main characteristics and auxiliary characteristics of the wiring graph, performing index coding on the main characteristics, writing the main characteristics into a graph characteristic library file, and writing the auxiliary characteristics and the main characteristic index into a characteristic description file;
as a specific implementation mode, the main characteristics comprise bus incoming and outgoing line interval characteristics, bus connection interval characteristics, bus division interval characteristics and bus accessory facility intervals, and the auxiliary characteristics comprise the number of main transformers, the number of buses of each voltage class and a bus connection mode.
As a specific implementation manner, the index of the main feature is composed of two segments, i.e. type: id, where type identifies the type of the interval, id identifies the constituent primitives that make up the interval, and the calculation formula is:
Figure BDA0001808345090000021
and n is the number of the types of the primitives contained in the interval, but does not comprise a connecting line, a bus and a main transformer, t is the type number of the primitives, and num is the number of the primitives with the type of t.
As a specific embodiment, each sub-inventory pattern corresponds to a pattern profile with the suffix of.fac.pic.g, sorted by voltage level (i.e., high, medium, and low voltage level) of the inventory pattern, and each type of inventory pattern corresponds to a profile with the suffix of.xml.
Step 2, generating characteristic parameters of the target graph: inputting target graphic parameters, and performing characterization processing on the parameters to form characteristic parameters;
as a specific implementation mode, the formed characteristic parameters comprise the number of main transformers, the types of the main transformers, the number of high-voltage side buses, middle-voltage side buses, low-voltage side buses, high-voltage side buses, middle-voltage side buses and low-voltage side buses, interval index values of high-voltage side buses, middle-voltage side buses and low-voltage side power supply, and interval index values of high-voltage side power supplies, middle-voltage side power supplies and low-voltage side power supplies.
Step 3, characteristic evaluation: comparing and evaluating the formed characteristic parameters with the graphic characteristics of a graphic characteristic library to obtain the graphic characteristics most similar to the characteristic parameters;
as a specific implementation, the evaluation formula is specifically:
Figure BDA0001808345090000031
wherein w1 represents the evaluation weight of the main transformer, n1 represents the number of the main transformers, s i Power matching degree score; w2 represents the busbar evaluation weight, m represents the busbar number, s j ' represents the busbar wiring matching degree score; w3 represents interval evaluation weight, p represents number of interval kinds, s k "represents the interval match score of each type; score represents the feature assessment score, with higher scores being more closely matched.
Step 4, generating and drawing a graph: and multiplexing, modifying and expanding the obtained graphic feature units to complete the layout and wiring of the graphics.
Examples
To verify the validity of the scheme of the present invention, the following simulation experiment was performed.
The graphic feature library is composed of a graphic feature file and a feature description file, as shown in fig. 2, the learning engine scans the historical stock graphs at regular time, and when a pair of stock graphs is scanned, firstly, whether the feature record of the graph exists in the feature description file is searched. If the record does not exist, performing topology analysis on the graph, sequentially extracting the features, writing the main features into a graph feature file, and writing the auxiliary features and the main feature index into a feature description file. And if no new graph or new graph characteristic needs to be extracted after the scanning is finished, sleeping for 5 minutes.
As shown in fig. 3, when a plant station diagram is automatically generated, input parameters are input through a front-end interface and are preprocessed to form target graphic features including the number of main transformers, the types of the main transformers, the wiring modes of high, middle and low voltage sides, the number of high, middle and low side buses, interval index values of high, middle and low voltage side incoming and outgoing lines, interval index values of high, middle and low voltage side bus-to-bus, interval index values of high, middle and low voltage side buses, interval index values of high, middle and low voltage side power supplies, and the like. And then comparing and evaluating the target graphic features with the graphic feature description file, finding out the feature file where the feature element closest to the target features is located, and reading the graphic feature unit from the feature files. And finally, copying, modifying, replacing and the like the graphic features acquired from the feature library by using a graphic layout and wiring algorithm according to the main features and the auxiliary features of the target graphic to generate a bus, a main transformer and an interval. After the graph is generated, the graph can be directly modified and adjusted in a graph editor, a data model of the equipment primitive in the graph is generated after the graph is edited, and the equipment primitive can be automatically associated with the data model in a remote signaling and remote measuring mode.
By the aid of the scheme, accurate high-quality plant station wiring diagrams which accord with drawing habits of users and integrate drawing models can be generated, and a series of problems that static drawings are large in operation and maintenance workload and prone to error in dispatching automation application are effectively solved.

Claims (3)

1. A plant station diagram automatic generation method based on machine learning is characterized by comprising the following steps:
step 1, establishing a characteristic model: performing topology analysis on the stock graph, identifying main characteristics and auxiliary characteristics of the wiring graph, performing index coding on the main characteristics, writing the main characteristics into a characteristic file in a graph characteristic library, and writing the auxiliary characteristics and the main characteristic index into a characteristic description file;
step 2, generating characteristic parameters of the target graph: inputting target graphic parameters, and performing characterization processing on the parameters to form characteristic parameters;
step 3, characteristic evaluation: comparing and evaluating the formed characteristic parameters with the graphic characteristics of a graphic characteristic library to obtain the graphic characteristics most similar to the characteristic parameters;
step 4, generating and drawing a graph: multiplexing, modifying and expanding the obtained graphic feature units to complete the layout and wiring of the graphics;
in the step 1, the main characteristics comprise bus incoming and outgoing line interval characteristics, bus connection interval characteristics, bus division interval characteristics and bus accessory facility intervals, and the auxiliary characteristics comprise the number of main transformers, the number of buses of each voltage class and bus connection modes;
in the step 1, the index of the main characteristic consists of two sections, namely type id, wherein the type identifies the type of the interval, and the id identifies the component primitives forming the interval;
in the step 2, the formed characteristic parameters comprise the number of main transformers, the types of the main transformers, the number of high-voltage side, middle-voltage side and low-voltage side buses, the wiring mode of the high-voltage side, middle-voltage side and low-voltage side buses, interval index values of high-voltage side, middle-voltage side and low-voltage side bus-to-bus lines, interval index values of high-voltage side, middle-voltage side and low-voltage side buses, interval index values of high-voltage side, middle-voltage side and low-voltage side power supplies and interval index values of high-voltage side, middle-voltage side and low-voltage side power supplies.
2. The method according to claim 1, wherein in step 1, each inventory pattern corresponds to a pattern feature file with a suffix of.fac.pic.g, the inventory patterns are classified according to voltage levels of the inventory patterns, and each type of inventory pattern corresponds to a feature description file with a suffix of.xml.
3. The method for automatically generating the plant station diagram based on the machine learning as claimed in claim 1, wherein in the step 3, the higher the score of the comparison evaluation is, the greater the similarity is, and the evaluation formula is:
Figure FDA0003983964780000011
wherein w1 represents the evaluation weight of the main transformer, n1 represents the number of the main transformers, S i Power matching degree score; w2 represents the busbar evaluation weight, m represents the busbar number, S j ' represents the busbar wiring matching degree score; w3 represents interval evaluation weight, p represents number of interval kinds, S k "represents the interval match score for each class; score represents the feature assessment score, with higher scores being more closely matched.
CN201811108036.6A 2018-09-21 2018-09-21 Plant station diagram automatic generation method based on machine learning Active CN109272564B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201811108036.6A CN109272564B (en) 2018-09-21 2018-09-21 Plant station diagram automatic generation method based on machine learning

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201811108036.6A CN109272564B (en) 2018-09-21 2018-09-21 Plant station diagram automatic generation method based on machine learning

Publications (2)

Publication Number Publication Date
CN109272564A CN109272564A (en) 2019-01-25
CN109272564B true CN109272564B (en) 2023-03-28

Family

ID=65198673

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201811108036.6A Active CN109272564B (en) 2018-09-21 2018-09-21 Plant station diagram automatic generation method based on machine learning

Country Status (1)

Country Link
CN (1) CN109272564B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114329856A (en) * 2021-04-28 2022-04-12 贵州电网有限责任公司 Single line diagram layout machine learning processing method and system

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105046345A (en) * 2015-05-20 2015-11-11 北京科东电力控制系统有限责任公司 Method of rapidly creating power grid plant station diagram
CN107292003A (en) * 2017-06-06 2017-10-24 南京南瑞继保电气有限公司 A kind of automatic generating method of electric network station wiring diagram

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105046345A (en) * 2015-05-20 2015-11-11 北京科东电力控制系统有限责任公司 Method of rapidly creating power grid plant station diagram
CN107292003A (en) * 2017-06-06 2017-10-24 南京南瑞继保电气有限公司 A kind of automatic generating method of electric network station wiring diagram

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
一种基于CIM的厂站接线图自动生成技术;沙树名等;《电力系统自动化》;20081110;第68-71页 *

Also Published As

Publication number Publication date
CN109272564A (en) 2019-01-25

Similar Documents

Publication Publication Date Title
CN106960103B (en) Automatic generation method and device of power grid graph
CN107292003A (en) A kind of automatic generating method of electric network station wiring diagram
CN104484738B (en) Electric network data manages method and system concentratedly
CN102403718B (en) Generating method for power grid topological relationship based on Arcgis
CN105809260A (en) Method for intelligently generating monitoring information point list of dispatching automation system
CN110046391B (en) Automatic transformer substation primary wiring diagram generation method based on monitoring information table
CN106910140B (en) SVG format-based power system wiring diagram network topology analysis method
CN108521123A (en) A kind of visual more contact point distribution network failure section correlating methods
CN105574675A (en) Rule bank-based method and system for planning and supervising constructional engineering
CN112668733A (en) Method and device for dispatching defect and fault of communication management system based on unsupervised learning
CN115204093A (en) Automatic layout method and system for plant station wiring diagram
CN117559443A (en) Ordered power utilization control method for large industrial user cluster under peak load
CN109272564B (en) Plant station diagram automatic generation method based on machine learning
CN113570345B (en) Automatic power failure range identification system based on construction project circuit diagram
CN109447466B (en) Overall process visual management and control system based on power distribution network communication network construction
CN105633970B (en) Distribution back scheduling control system building method based on major network scheduling system
CN116841985A (en) Method, system and device for building plant model topology
CN112967376B (en) Data-driven transformer substation three-dimensional model establishing method and device
CN110310048A (en) A kind of distribution planning overall process appraisal procedure and device
CN102760194B (en) Method and system for combining common information models (CIMs) of transformer substations and master scheduling station
CN112101708B (en) Expert knowledge-fused intelligent decision method and system for short circuit safety of power grid
CN115422146B (en) Construction and application method of enterprise standardized database in Xinjiang region
CN109816168A (en) Data prediction system for electrical equipment
CN109472641A (en) A kind of electricity power engineering analysis of prices method and system
CN116310258B (en) Three-dimensional construction management system and method based on transformer substation live-action

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