Search

CN-122022860-A - Industrial user portrait construction method based on non-invasive sensing technology

CN122022860ACN 122022860 ACN122022860 ACN 122022860ACN-122022860-A

Abstract

The invention provides an industrial user portrait construction method based on a non-invasive sensing technology, which comprises the steps of determining original electricity utilization data of an industrial user, carrying out sliding window detection on the original electricity utilization data based on a detection threshold value to obtain a load event list, wherein the load event list comprises a first starting time pair of a load switching event or a second starting time pair of a load mode adjustment event, determining response time of the load event list, carrying out feature matrix construction based on the response time and the original electricity utilization data to obtain an electricity utilization feature matrix, carrying out similarity matching based on the electricity utilization feature matrix to obtain an industrial user portrait, and carrying out industry classification on the industrial user by the industrial user portrait. The method and the device can train without depending on a large amount of data, are high in algorithm calculation efficiency, and can effectively characterize industrial user resources of different types of sub-industries.

Inventors

  • ZHU SHIJIA
  • ZHAO JINCHAO
  • YANG KAIXUAN
  • SUN HAIBIN
  • CAI SHENGLIANG
  • ZHANG ZUOTING
  • TAN DASHUAI
  • TIAN YOUJIA
  • WANG JING
  • GAO LIYUAN
  • WANG SHAOFEI
  • SONG RUI
  • WANG XUEBIN
  • CHENG LIANG
  • Xing Fahui
  • Li Shute

Assignees

  • 国网信息通信产业集团有限公司
  • 国网青海省电力公司
  • 国网青海省电力公司电力科学研究院

Dates

Publication Date
20260512
Application Date
20251023

Claims (10)

  1. 1. An industrial user portrait construction method based on a non-invasive sensing technology is characterized by comprising the following steps: Determining original electricity utilization data of an industrial user, and detecting a sliding window of the original electricity utilization data based on a detection threshold value to obtain a load event list, wherein the load event list comprises a first starting and ending time pair of a load switching event or a second starting and ending time pair of a load mode adjustment event; determining the response time of the load event list, and constructing a feature matrix based on the response time and the original electricity utilization data to obtain an electricity utilization feature matrix; And performing similarity matching based on the electricity utilization feature matrix to obtain an industrial user portrait, wherein the industrial user portrait is configured to perform industry classification on the industrial user.
  2. 2. The method of claim 1, wherein the detection threshold comprises a load cut-in detection threshold and a load mode adjustment detection threshold.
  3. 3. The method according to claim 2, characterized in that the first start-stop time is obtained by: and determining an electrical parameter difference value of an adjacent window of the original power utilization data, and carrying out threshold detection on the electrical parameter difference value based on a load switching detection threshold value to obtain the first starting time, wherein the electrical parameter difference value comprises a current effective value difference value and a power average value difference value.
  4. 4. The method according to claim 2, characterized in that the second start-stop time pair is obtained by: and the number of the crossing points of the load power and the load average power of the original electricity data is truly determined, and threshold detection is carried out on the number of the crossing points based on the load mode adjustment detection threshold value, so that the second starting and ending time pair is obtained.
  5. 5. The method of claim 1, wherein the constructing the feature matrix based on the response time and the raw electricity consumption data to obtain an electricity consumption feature matrix comprises: determining the power consumption information and the power consumption information of the original power consumption data, and extracting the characteristics of the response time, the original power consumption data and the power consumption information to obtain the power consumption characteristics of a user; and constructing a matrix based on the electricity utilization characteristics of the user to obtain the electricity utilization characteristic matrix.
  6. 6. The method of claim 1, wherein the performing similarity matching based on the electricity utilization feature matrix to obtain an industrial user representation comprises: constructing based on the electricity utilization characteristic matrix to obtain an industrial user behavior vector; performing similarity measurement on the industrial user behavior vector to obtain an industrial user Euclidean distance; And clustering the industrial users based on the Euclidean distance of the industrial users to obtain the portrait of the industrial users.
  7. 7. An industrial consumer representation construction device based on non-invasive sensing technology, comprising: The system comprises an event list determining module, a load event list generating module and a load event judging module, wherein the event list determining module is configured to determine original electricity utilization data of an industrial user, and perform sliding window detection on the original electricity utilization data based on a detection threshold value to obtain a load event list, and the load event list comprises a first starting time pair of a load switching event or a second starting time pair of a load mode adjusting event; The characteristic matrix determining module is configured to determine the response time of the load event list, and perform characteristic matrix construction based on the response time and the original electricity consumption data to obtain an electricity consumption characteristic matrix; and the user portrait determining module is configured to perform similarity matching based on the electricity utilization characteristic matrix to obtain industrial user portraits, and the industrial user portraits are configured to perform industry classification on the industrial users.
  8. 8. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the method of any one of claims 1 to 6 when the program is executed.
  9. 9. A non-transitory computer readable storage medium storing computer instructions for causing a computer to perform the method of any one of claims 1 to 6.
  10. 10. A computer program product comprising computer program instructions which, when run on a computer, cause the computer to perform the method of any one of claims 1 to 6.

Description

Industrial user portrait construction method based on non-invasive sensing technology Technical Field The disclosure relates to the technical field of power systems, in particular to an industrial user portrait construction method based on a non-invasive sensing technology. Background This section is intended to provide a background or context to the embodiments of the disclosure recited in the claims. The description herein is not admitted to be prior art by inclusion in this section. Currently, the power grid is being changed from energy consumption control to carbon emission control, and the energy structure is changed to clean low carbon, so that new challenges are brought to power supply and demand balance, and management on the demand side is increasingly important. The traditional invasive load monitoring is difficult to popularize due to high cost and complex installation, and the non-invasive load sensing technology can analyze the equipment state and the energy consumption trend through total data of users and is widely focused. The existing user portrait method based on electricity utilization time and power is difficult to accurately distinguish industrial users in different industries, core production equipment in different industries is relatively fixed, and the start-stop and operation mode change time becomes a key distinguishing feature. However, in the related art, there are problems of complex types, poor interpretability, dependence on a large amount of history data to train a model, and the like. Disclosure of Invention In view of the above, an object of the present disclosure is to provide an industrial user portrait construction method based on a non-invasive sensing technology, which at least solves one of the technical problems in the related art to a certain extent. With the above object in view, a first aspect of exemplary embodiments of the present disclosure provides a method for industrial user portrayal construction based on a non-invasive sensing technique, the method comprising: Determining original electricity utilization data of an industrial user, and detecting a sliding window of the original electricity utilization data based on a detection threshold value to obtain a load event list, wherein the load event list comprises a first starting and ending time pair of a load switching event or a second starting and ending time pair of a load mode adjustment event; determining the response time of the load event list, and constructing a feature matrix based on the response time and the original electricity utilization data to obtain an electricity utilization feature matrix; And performing similarity matching based on the electricity utilization feature matrix to obtain an industrial user portrait, wherein the industrial user portrait is configured to perform industry classification on the industrial user. Based on the same inventive concept, a second aspect of exemplary embodiments of the present disclosure provides an industrial user portrayal construction device based on a non-invasive sensing technique, comprising: The system comprises an event list determining module, a load event list generating module and a load event judging module, wherein the event list determining module is configured to determine original electricity utilization data of an industrial user, and perform sliding window detection on the original electricity utilization data based on a detection threshold value to obtain a load event list, and the load event list comprises a first starting time pair of a load switching event or a second starting time pair of a load mode adjusting event; The characteristic matrix determining module is configured to determine the response time of the load event list, and perform characteristic matrix construction based on the response time and the original electricity consumption data to obtain an electricity consumption characteristic matrix; and the user portrait determining module is configured to perform similarity matching based on the electricity utilization characteristic matrix to obtain industrial user portraits, and the industrial user portraits are configured to perform industry classification on the industrial users. Based on the same inventive concept, a third aspect of exemplary embodiments of the present disclosure provides an electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the method according to the first aspect when executing the program. Based on the same inventive concept, a fourth aspect of the exemplary embodiments of the present disclosure provides a non-transitory computer-readable storage medium storing computer instructions for causing a computer to perform the method according to the first aspect. Based on the same inventive concept, a fifth aspect of exemplary embodiments of the present disclosure provides a computer program product compri