WO2023246070A1 - 异常用电站点检测方法、装置、电子设备及可读存储介质 - Google Patents

异常用电站点检测方法、装置、电子设备及可读存储介质 Download PDF

Info

Publication number
WO2023246070A1
WO2023246070A1 PCT/CN2022/143575 CN2022143575W WO2023246070A1 WO 2023246070 A1 WO2023246070 A1 WO 2023246070A1 CN 2022143575 W CN2022143575 W CN 2022143575W WO 2023246070 A1 WO2023246070 A1 WO 2023246070A1
Authority
WO
WIPO (PCT)
Prior art keywords
change rate
power consumption
data
consumption data
abnormal
Prior art date
Application number
PCT/CN2022/143575
Other languages
English (en)
French (fr)
Inventor
庄丹
马方明
邓金荣
许学研
刘吉宁
曹海山
郭向荣
张歌
陈庆专
黄伟群
Original Assignee
中国移动通信集团广东有限公司
中国移动通信集团有限公司
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 中国移动通信集团广东有限公司, 中国移动通信集团有限公司 filed Critical 中国移动通信集团广东有限公司
Publication of WO2023246070A1 publication Critical patent/WO2023246070A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
    • 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
    • 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

Definitions

  • the present disclosure relates to the field of communications, and specifically to an abnormal power site detection method, device, electronic equipment and readable storage medium.
  • the main solutions for detecting abnormal power consumption in the computer room include: (1) Installing a smart meter inside the computer room, collecting the power consumption data of all equipment in the computer room, and matching it with the smart meter power consumption data to obtain abnormal power consumption data. 2. Use equipment rated power detection. Each equipment in the computer room has its maximum rated power. Normal power consumption of the equipment cannot exceed the maximum rated power. Obtain the maximum power value of all equipment in the computer room, and use monthly power consumption data to compare with the entire Compare the rated power of all equipment in the monthly computer room, and identify the sites corresponding to equipment that exceeds the rated power as abnormal sites to obtain abnormal power consumption data.
  • the disadvantage of option (1) is that it requires a large investment, there are computer rooms that are not suitable for installing smart meters, and it is difficult to collect power consumption data of all equipment in the computer room.
  • the flaw of solution (2) is that the comparison dimension is single and other factors that affect the power consumption of the computer room equipment are not considered. Moreover, the rated power value of the computer room equipment itself is relatively large. Even if electricity theft occurs, it will rarely exceed the rated power. Therefore, The detection accuracy of scheme (2) is insufficient.
  • Embodiments of the present disclosure provide an abnormal power consumption site detection method, device, electronic equipment and readable storage medium to solve the above problems existing in related computer room abnormal power consumption detection solutions.
  • embodiments of the present disclosure provide a method for detecting abnormal power consumption sites, including:
  • the change rate correlation coefficient and the change rate anti-correlation time the abnormal power consumption site in the site to be detected is determined.
  • the step of collecting power consumption data, business data and energy consumption data of the site to be detected includes:
  • the data sequence corresponding to the first time is obtained, and the missing data is supplemented according to the preset interpolation method and the data sequence.
  • the power consumption data, business data and energy consumption data of the site to be detected are collected, and the first change rate corresponding to the power consumption data, the second change rate corresponding to the business data and the
  • the third change rate step corresponding to the energy consumption data includes:
  • the step of determining the abnormal time according to the target normal distribution includes:
  • the collection time of the deviation data is determined, and the collection time is regarded as an abnormal time.
  • the step of determining the change rate correlation coefficient and the change rate anti-correlation time based on the first change rate, the second change rate and the third change rate includes:
  • a second change rate anti-correlation time is determined according to the second change rate correlation coefficient.
  • the step of determining the abnormal power consumption site in the site to be detected based on the abnormal time, the change rate correlation coefficient and the change rate anti-correlation time includes:
  • the sites to be detected are sorted according to the radian value of the included angle, and the abnormal power consumption sites in each of the sites to be detected are determined according to the sorting results.
  • the step of determining the first change rate corresponding to the power consumption data, the second change rate corresponding to the business data, and the third change rate corresponding to the energy consumption data includes:
  • the business data and the energy consumption data have the same collection time and the same collection period, calculate the first change rate according to the power consumption data corresponding to each of the collection times;
  • the second change rate is calculated based on the business data corresponding to each collection time
  • the third change rate is calculated based on the energy consumption data corresponding to each collection time.
  • an abnormal power consumption site detection device including:
  • a change rate calculation module used to collect power consumption data, business data and energy consumption data of the site to be detected, and determine the first change rate corresponding to the power consumption data, the second change rate corresponding to the business data and the energy consumption data.
  • An abnormal time determination module configured to determine the abnormal time according to the target normal distribution when the power consumption data conforms to the target normal distribution
  • a correlation coefficient calculation module configured to determine the change rate correlation coefficient and the change rate anti-correlation time according to the first change rate, the second change rate and the third change rate;
  • An abnormal power consumption site determination module is used to determine the abnormal power consumption site in the site to be detected based on the abnormal time, the change rate correlation coefficient and the change rate anti-correlation time.
  • an embodiment of the present disclosure provides an electronic device, including a processor and a memory storing a computer program.
  • the processor executes the computer program, the steps of the abnormal power site detection method described in the first aspect are implemented. .
  • an embodiment of the present disclosure provides a readable storage medium, which includes a computer program.
  • the computer program is executed by a processor, the steps of the abnormal power site detection method described in the first aspect are implemented.
  • the abnormal power consumption site detection method, device, electronic equipment and readable storage medium collect the power consumption data, business data and energy consumption data of the site to be detected, and determine the change rate corresponding to each data.
  • the abnormal time is determined based on the target normal distribution, and then the change rate correlation coefficient and the change rate anti-correlation time are determined based on the change rate corresponding to each data, and finally based on the abnormal time and change rate correlation coefficient Anti-correlation time with the change rate, determine the abnormal power consumption site in the site to be detected, calculate and detect abnormal power consumption sites through the power consumption data, business data and energy consumption data of the site to be detected, to achieve accurate identification of abnormal power consumption sites.
  • Figure 1 is one of the flow diagrams of the method for detecting abnormal power consumption sites provided by an embodiment of the present disclosure
  • Figure 2 is the second schematic flowchart of the method for detecting abnormal power consumption sites provided by an embodiment of the present disclosure
  • Figure 3 is a schematic structural diagram of an abnormal power site detection device provided by an embodiment of the present disclosure.
  • FIG. 4 is a schematic structural diagram of an electronic device provided by an embodiment of the present disclosure.
  • FIG. 1 is one of the schematic flow diagrams of a method for detecting abnormal power consumption sites in an embodiment of the present disclosure.
  • the method for detecting abnormal power consumption sites provided by embodiments of the present disclosure may include:
  • Step S100 Collect the power consumption data, business data and energy consumption data of the site to be detected, and determine the first change rate corresponding to the power consumption data, the second change rate corresponding to the business data, and the second change rate corresponding to the energy consumption data. third rate of change;
  • the site to be detected in this embodiment refers to the computer room where the phenomenon of electricity theft is to be detected.
  • the power consumption data, business data and energy consumption data of the site to be detected are collected through equipment (such as electricity meters) in the computer room.
  • the collection period can be 30 days. or 60 days.
  • the collection time point can be 0 o'clock.
  • the collected power consumption data includes the daily power consumption of the site to be detected during the collection period.
  • the business data and energy consumption data are the same.
  • the power consumption data and business data are compared in chronological order. Sort the energy consumption data to obtain the power consumption data, business data and energy consumption data of each site to be detected.
  • the calculation method for the second change rate corresponding to the business data and the third change rate corresponding to the energy consumption data is the same as the first change rate corresponding to the power consumption data.
  • Step S200 when the power consumption data conforms to the target normal distribution, determine the abnormal time according to the target normal distribution
  • the power consumption data, business data and energy consumption data are sorted in chronological order. After obtaining the power consumption data, business data and energy consumption data of each site to be detected, it is judged whether the power consumption data conforms to the requirements in this embodiment.
  • the target normal distribution the specific method is as follows: Randomly extract multiple power consumption data on different dates from the overall power consumption data each time to find the average, randomly extract multiple times, record the average value of each extraction, and extract the required multiple times. Covers power consumption data for all dates. Verify whether the average value drawn each time conforms to the normal distribution.
  • n is the number of extractions
  • x i is the average value of i-th sampling
  • x j is the average value of j-th sampling
  • exp is the exponential function with the natural constant e as the base.
  • ⁇ is set to 0.05.
  • the dates of potential abnormal data are screened based on the standard deviation that deviates from the sample mean by more than 3 times, and the number of days when potentially abnormal data occurs is counted, that is, the number of days of abnormal time in this embodiment. .
  • Step S300 determine the change rate correlation coefficient and the change rate anti-correlation time according to the first change rate, the second change rate and the third change rate;
  • the change rate correlation coefficient in this embodiment includes a first change rate correlation coefficient and a second change rate correlation coefficient
  • the change rate anti-correlation time includes the first change rate anti-correlation time and the second change rate anti-correlation time.
  • the method for determining the second change rate correlation coefficient r (E, C) is: Cov is used to calculate covariance and Var is used to calculate variance.
  • the calculation method of the first change rate correlation coefficient is the same as the second change rate correlation coefficient.
  • the day is determined to be the second change rate.
  • Anti-correlation time The method for determining the first change rate anti-correlation time is the same as the method for determining the second change rate anti-correlation time.
  • Step S400 Determine the abnormal power consumption site in the site to be detected based on the abnormal time, the change rate correlation coefficient and the change rate anti-correlation time.
  • all site power consumption characteristics are abstracted into groups of vectors.
  • the content of the vector is as follows: [mean, variance, kurtosis, skewness statistical information of site power consumption data, first change rate correlation coefficient, second change rate correlation coefficient, change rate anti-correlation time and days, anti-correlation deviation amplitude value] .
  • the radian value of the angle between the feature vector of the site to be detected and the feature vector of sites known to have electricity theft is calculated, and the site to be predicted corresponding to the smallest radian value among the sites to be detected is taken as the detection result.
  • a ⁇ x 1 x 2 ...
  • a is the eigenvector of the site to be detected
  • b is the eigenvector of the site where electricity theft is known to exist
  • cos ⁇ is the radian value of the angle between the eigenvector of the site to be detected and the eigenvector of the site where electricity theft is known to exist.
  • the arc values of each included angle are sorted in order from small to large. The sites to be detected that are ranked higher are more likely to have electricity theft, and the output is an abnormal power consumption site.
  • This embodiment collects the power consumption data, business data and energy consumption data of the site to be detected, and determines the change rate corresponding to each data.
  • the abnormal time is determined according to the target normal distribution.
  • the change rate correlation coefficient and the change rate anti-correlation time are determined.
  • the change rate correlation coefficient and the change rate anti-correlation time are determined.
  • the abnormal power consumption site detection method provided by the embodiment of the present disclosure may also include:
  • Step S101 when there is missing data in the power consumption data, determine the first time corresponding to the missing data
  • Step S102 Obtain the data sequence corresponding to the first time, and supplement the missing data according to the preset interpolation method and the data sequence.
  • this embodiment uses mean imputation to fill in missing data. Sort the power consumption data, business data and energy consumption data in chronological order (that is, the data sequence in this embodiment). After obtaining the power consumption data, business data and energy consumption data of each site to be detected, the following mean value is used to compensate method (i.e., the preset interpolation method in this implementation) to fill in the missing data.
  • the average value of the data before and after the time sequence of the missing data is used as the missing data to supplement the missing data.
  • This embodiment uses preset interpolation methods and time sorting to supplement missing data, ensuring the integrity of the data and laying a data foundation for subsequent accurate identification of abnormal power consumption sites.
  • the abnormal power consumption site detection method provided by the embodiment of the present disclosure may also include:
  • Step S110 determine the mean sequence corresponding to the power consumption data according to the preset sampling method
  • Step S120 Check the mean sequence according to the preset calibration formula to obtain the calibration result
  • Step S130 Determine whether the power consumption data conforms to the target normal distribution according to the test result.
  • the power consumption data, business data and energy consumption data are sorted in chronological order. After obtaining the power consumption data, business data and energy consumption data of each site to be detected, it is judged whether the power consumption data conforms to the requirements in this embodiment.
  • the target normal distribution the specific method is as follows: randomly select multiple (for example, 5-8) power consumption data on different dates from the overall power consumption data according to the preset sampling method to find the average, and randomly select multiple times ( For example, 200 times), record the average value of each extraction to obtain the mean sequence corresponding to the power consumption data, and extract the power consumption data multiple times to cover all dates.
  • the verification formula (ie, the preset verification formula in this embodiment) is as follows: Among them, n is the number of extractions, is the average value of n-th sampling, x i is the average value of i-th sampling, x j is the average value of j-th sampling, and exp is the exponential function with the natural constant e as the base. Assume that the average value of each draw conforms to the normal distribution, and the confidence ⁇ is set to 0.05. Calculate T ep .
  • This embodiment uses a preset sampling method and a preset verification formula to determine whether the power consumption data conforms to the target normal distribution, and preliminary verification of the power consumption data lays a data foundation for the subsequent accurate identification of abnormal power consumption sites.
  • the abnormal power consumption site detection method provided by the embodiment of the present disclosure may also include:
  • Step S210 determine the deviation data in the power consumption data according to the standard deviation corresponding to the target normal distribution
  • Step S220 Determine the collection time of the deviation data, and use the collection time as an abnormal time.
  • the date of the potential abnormal data i.e., the deviation data in this embodiment
  • the number of days is the number of abnormal time days in this embodiment.
  • This embodiment screens potential abnormal data through the target normal distribution and initially screens the power consumption data, laying a data foundation for subsequent accurate identification of abnormal power consumption sites.
  • the abnormal power consumption site detection method provided by the embodiment of the present disclosure may also include:
  • Step S310 determine the first change rate correlation coefficient according to the first change rate, the second change rate and the preset coefficient calculation formula
  • Step S320 determine the second change rate correlation coefficient according to the first change rate, the third change rate and the preset coefficient calculation formula
  • Step S330 If the first change rate correlation coefficient is less than the preset threshold, determine the first change rate anti-correlation time according to the first change rate correlation coefficient;
  • Step S340 If the second change rate correlation coefficient is less than the preset threshold, determine the second change rate anti-correlation time according to the second change rate correlation coefficient.
  • the method for determining the second change rate correlation coefficient r (E, C) is: Cov is used to calculate covariance and Var is used to calculate variance.
  • the calculation method of the first change rate correlation coefficient is the same as the second change rate correlation coefficient.
  • the day is determined to be the second change rate.
  • Anti-correlation time The method for determining the first change rate anti-correlation time is the same as the method for determining the second change rate anti-correlation time.
  • This embodiment uses the change rate and the preset coefficient calculation formula to determine the change rate correlation coefficient and the change rate anti-correlation time, thereby obtaining data for determining abnormal power consumption sites, and achieving accurate identification of abnormal power consumption sites.
  • Figure 2 is a schematic flowchart 2 of a method for detecting abnormal power consumption sites in an embodiment of the present disclosure.
  • the method for detecting abnormal power consumption sites provided by an embodiment of the present disclosure may also include:
  • Step S410 Determine the first feature vector corresponding to the site to be detected based on the power consumption data, the abnormal time, the change rate correlation coefficient and the change rate anti-correlation time;
  • Step S420 Obtain the second feature vector corresponding to the preset abnormal site, and calculate the radian value of the angle between the first feature vector and the second feature vector;
  • Step S430 Sort the sites to be detected according to the radian value of the included angle, and determine the abnormal power consumption sites in each of the sites to be detected based on the sorting results.
  • the characteristic vector of the station to be detected and the characteristic vector of the station known to have electricity theft are calculated (i.e., the second eigenvector in this embodiment).
  • the radian value of the angle between them is taken as the detection result.
  • the site to be predicted corresponding to the smallest radian value among the sites to be detected is taken.
  • a ⁇ x 1 x 2 ... x n ⁇
  • b ⁇ y 1 y 2 ... y n ⁇
  • a is the eigenvector of the site to be detected
  • b is the eigenvector of the site where electricity theft is known to exist
  • cos ⁇ is the radian value of the angle between the eigenvector of the site to be detected and the eigenvector of the site where electricity theft is known to exist.
  • the arc values of each included angle are sorted in order from small to large. The sites to be detected that are ranked higher are more likely to have electricity theft, and the output is an abnormal power consumption site.
  • This embodiment uses characteristic data such as abnormal time, change rate correlation coefficient, and change rate anti-correlation time to determine abnormal power consumption sites and achieve accurate identification of abnormal power consumption sites.
  • the abnormal power consumption site detection method provided by the embodiment of the present disclosure may also include:
  • Step S103 When the power consumption data, the business data and the energy consumption data have the same collection time and the same collection period, calculate the first change rate based on the power consumption data corresponding to each collection time;
  • Step S104 Calculate the second change rate based on the business data corresponding to each collection time, and calculate the third change rate based on the energy consumption data corresponding to each collection time.
  • the power consumption data, business data and energy consumption data have the same collection time and the same collection cycle, taking the power consumption data obtained by 60-day collection cycle, 0 o'clock collection time, and sorted by time as an example, determine the consumption data.
  • the calculation method for the second change rate corresponding to the business data and the third change rate corresponding to the energy consumption data is the same as the first change rate corresponding to the power consumption data.
  • this embodiment lays a data foundation for the subsequent accurate identification of abnormal power consumption sites.
  • Figure 3 is a schematic structural diagram of an abnormal power usage site detection device in an embodiment of the present disclosure.
  • the abnormal power usage site detection device provided by an embodiment of the present disclosure will be described below.
  • the abnormal power usage site detection device described below is the same as the above.
  • the described detection methods for abnormal power consumption sites can be referenced accordingly.
  • Abnormal power consumption site detection devices include:
  • the change rate calculation module 301 is used to collect the power consumption data, business data and energy consumption data of the site to be detected, and determine the first change rate corresponding to the power consumption data, the second change rate corresponding to the business data and the The third change rate corresponding to the energy consumption data;
  • the abnormal time determination module 302 is configured to determine the abnormal time according to the target normal distribution when the power consumption data conforms to the target normal distribution;
  • the correlation coefficient calculation module 303 is used to determine the change rate correlation coefficient and the change rate anti-correlation time according to the first change rate, the second change rate and the third change rate;
  • the abnormal power consumption site determination module 304 is used to determine the abnormal power consumption site in the site to be detected based on the abnormal time, the change rate correlation coefficient and the change rate anti-correlation time.
  • the abnormal power consumption site detection device also includes:
  • a first time determination module configured to determine the first time corresponding to the missing data when there is missing data in the power consumption data
  • a missing data supplement module is used to obtain the data sequence corresponding to the first time, and supplement the missing data according to the preset interpolation method and the data sequence.
  • the abnormal power consumption site detection device also includes:
  • a mean sequence determination module configured to determine the mean sequence corresponding to the power consumption data according to a preset sampling method
  • a verification result determination module used to verify the mean sequence according to a preset verification formula to obtain a verification result
  • a judgment module configured to judge whether the power consumption data conforms to the target normal distribution according to the test results.
  • the abnormal time determination module includes:
  • a deviation data determination unit configured to determine the deviation data in the power consumption data based on the standard deviation corresponding to the target normal distribution
  • An abnormal time determination unit is used to determine the collection time of the deviation data, and use the collection time as the abnormal time.
  • the correlation coefficient calculation module includes:
  • a first change rate correlation coefficient determination unit configured to determine a first change rate correlation coefficient based on the first change rate, the second change rate and a preset coefficient calculation formula
  • a second change rate correlation coefficient determination unit configured to determine a second change rate correlation coefficient based on the first change rate, the third change rate and a preset coefficient calculation formula
  • a first change rate anti-correlation time determination unit configured to determine a first change rate anti-correlation time based on the first change rate correlation coefficient when the first change rate correlation coefficient is less than a preset threshold
  • a second change rate anti-correlation time determination unit is configured to determine a second change rate anti-correlation time based on the second change rate correlation coefficient when the second change rate correlation coefficient is less than a preset threshold.
  • the abnormal power consumption site determination module includes:
  • a first feature vector determination unit configured to determine the first feature vector corresponding to the site to be detected based on the power consumption data, the abnormal time, the change rate correlation coefficient, and the change rate anti-correlation time;
  • An included angle radian value calculation unit is used to obtain the second feature vector corresponding to the preset abnormal site, and calculate the included angle radian value between the first feature vector and the second feature vector;
  • An abnormal power consumption site determination unit is configured to sort the sites to be detected according to the radian value of the included angle, and determine the abnormal power consumption sites in each of the sites to be detected according to the sorting results.
  • the change rate calculation module includes:
  • a first change rate calculation unit configured to calculate the power consumption data corresponding to each collection time when the power consumption data, the business data and the energy consumption data have the same collection time and the same collection period.
  • the third change rate calculation unit is configured to calculate the second change rate based on the business data corresponding to each collection time, and calculate the third change rate based on the energy consumption data corresponding to each collection time.
  • Figure 4 illustrates a schematic diagram of the physical structure of an electronic device.
  • the electronic device may include: a processor (processor) 410, a communication interface (Communication Interface) 420, a memory (memory) 430 and a communication bus 440.
  • the processor 410, the communication interface 420, and the memory 430 complete communication with each other through the communication bus 440.
  • the processor 410 can call the computer program in the memory 430 to perform the steps of the abnormal power consumption site detection method.
  • the above-mentioned logical instructions in the memory 430 can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product.
  • the technical solution of the present disclosure is essentially or the part that contributes to the relevant technology or the part of the technical solution can be embodied in the form of a software product.
  • the computer software product is stored in a storage medium and includes several The instructions are used to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present disclosure.
  • the aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program code. .
  • embodiments of the present disclosure also provide a readable storage medium, the readable storage medium includes a computer program, the computer program can be stored on a non-transitory computer readable storage medium, and the computer program is processed When the computer is executed, the computer can execute the steps of the abnormal power consumption site detection method provided by the above embodiments.
  • the device embodiments described above are only illustrative.
  • the units described as separate components may or may not be physically separated.
  • the components shown as units may or may not be physical units, that is, they may be located in One location, or it can be distributed across multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Persons of ordinary skill in the art can understand and implement the method without any creative effort.
  • each embodiment can be implemented by software plus a necessary general hardware platform, and of course, it can also be implemented by hardware.
  • the computer software products can be stored in computer-readable storage media, such as ROM/RAM, disks, etc. , CD, etc., including a number of instructions to cause a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the abnormal power site detection method described in each embodiment or some parts of the embodiment.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Business, Economics & Management (AREA)
  • Data Mining & Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Analysis (AREA)
  • Pure & Applied Mathematics (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Economics (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Human Resources & Organizations (AREA)
  • Tourism & Hospitality (AREA)
  • Strategic Management (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Primary Health Care (AREA)
  • Marketing (AREA)
  • Evolutionary Biology (AREA)
  • General Business, Economics & Management (AREA)
  • General Health & Medical Sciences (AREA)
  • Operations Research (AREA)
  • Probability & Statistics with Applications (AREA)
  • Water Supply & Treatment (AREA)
  • Algebra (AREA)
  • Public Health (AREA)
  • Databases & Information Systems (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

本公开涉及通信领域，提供一种异常用电站点检测方法、装置、电子设备及可读存储介质。异常用电站点检测方法包括：采集待检测站点的耗电数据、业务数据和能耗数据，确定耗电数据对应的第一变化率，业务数据对应的第二变化率以及能耗数据对应的第三变化率；在耗电数据符合目标正态分布的情况下，根据目标正态分布确定异常时间；根据第一变化率、第二变化率和第三变化率，确定变化率相关系数和变化率反相关时间；根据异常时间、变化率相关系数和变化率反相关时间，确定待检测站点中的异常用电站点。

Description

异常用电站点检测方法、装置、电子设备及可读存储介质
相关申请的交叉引用
本申请主张在2022年06月22日在中国提交的中国专利申请号No.202210713265.0的优先权，其全部内容通过引用包含于此。
技术领域
本公开涉及通信领域，具体涉及一种异常用电站点检测方法、装置、电子设备及可读存储介质。
背景技术
通信运营商网络机房广泛分布，由于无人看管，基站时长出现偷电漏电的现象，不仅侵害了通信运营商的合法权益，还造成了资源损失，而且偷电漏电现象引起的电压不稳问题，还对机房设备造成较大的安全隐患。由于偷电漏电的隐秘性，很难及时和精准地检测机房用电异常情况。
相关的机房用电异常检测主要方案包括：(1)在机房内部加装智能电表，收集机房内部所有设备用电数据，并与智能电表用电数据进行匹配，从而获得异常用电的数据。2、利用设备额定功率检测，机房的每个设备都有其最大额定功率，设备正常用电不可能突破最大额定功率，获取机房内所有设备的最大功率值，用每月的用电数据与整个月机房所有设备的额定功率比对，将超过额定功率的设备对应的站点标识为异常站点，从而获得异常用电数据。方案(1)的缺陷在于投资大，存在不适合加装智能电表的机房且收集机房内部所有设备用电数据较困难。方案(2)的缺陷在于比对维度单一，没有考虑影响机房设备用电量的其他因素，且机房设备额定功率值本身就比较大，即使出现偷电现象，也很少会超过额定功率，因此方案(2)的检测精度不足。
发明内容
本公开实施例提供一种异常用电站点检测方法、装置、电子设备及可读存储介质，用以解决相关机房用电异常检测方案中存在的上述问题。
第一方面，本公开实施例提供一种异常用电站点检测方法，包括：
采集待检测站点的耗电数据、业务数据和能耗数据，确定所述耗电数据对应的第一变化率，所述业务数据对应的第二变化率以及所述能耗数据对应的第三变化率；
在所述耗电数据符合目标正态分布的情况下，根据所述目标正态分布确定异常时间；
根据所述第一变化率、所述第二变化率和所述第三变化率，确定变化率相关系数和变化率反相关时间；
根据所述异常时间、所述变化率相关系数和所述变化率反相关时间，确定所述待检测站点中的异常用电站点。
在一个实施例中，所述采集待检测站点的耗电数据、业务数据和能耗数据的步骤之后包括：
在所述耗电数据中存在缺失数据的情况下，确定所述缺失数据对应的第一时间；
获取所述第一时间对应的数据序列，根据预设插值法和所述数据序列，对所述缺失数据进行补充。
在一个实施例中，所述采集待检测站点的耗电数据、业务数据和能耗数据，确定所述耗电数据对应的第一变化率，所述业务数据对应的第二变化率以及所述能耗数据对应的第三变化率的步骤之后包括：
根据预设抽样方式确定所述耗电数据对应的均值序列；
根据预设校验公式对所述均值序列进行检验，得到校验结果；
根据所述检验结果判断所述耗电数据是否符合目标正态分布。
在一个实施例中，所述根据所述目标正态分布确定异常时间的步骤包括：
根据所述目标正态分布对应的标准差，确定所述耗电数据中的偏离数据；
确定所述偏离数据的采集时间，并将所述采集时间作为异常时间。
在一个实施例中，所述根据所述第一变化率、所述第二变化率和所述第三变化率，确定变化率相关系数和变化率反相关时间的步骤包括：
根据所述第一变化率、所述第二变化率和预设系数计算公式，确定第一变化率相关系数；
根据所述第一变化率、所述第三变化率和预设系数计算公式，确定第二变化率相关系数；
在所述第一变化率相关系数小于预设阈值的情况下，根据所述第一变化率相关系数确定第一变化率反相关时间；
在所述第二变化率相关系数小于预设阈值的情况下，根据所述第二变化率相关系数确定第二变化率反相关时间。
在一个实施例中，所述根据所述异常时间、所述变化率相关系数和所述变化率反相关时间，确定所述待检测站点中的异常用电站点的步骤包括：
根据所述耗电数据、所述异常时间、所述变化率相关系数和所述变化率反相关时间，确定所述待检测站点对应的第一特征向量；
获取预设异常站点对应的第二特征向量，计算所述第一特征向量与所述第二特征向量之间的夹角弧度值；
根据所述夹角弧度值的大小对各所述待检测站点进行排序，根据排序结果确定各所述待检测站点中的异常用电站点。
在一个实施例中，所述确定所述耗电数据对应的第一变化率，所述业务数据对应的第二变化率以及所述能耗数据对应的第三变化率的步骤包括：
在所述耗电数据、所述业务数据以及所述能耗数据具有相同采集时间和相同采集周期的情况下，根据各所述采集时间对应的耗电数据计算第一变化率；
根据各所述采集时间对应的业务数据计算第二变化率，根据各所述采集时间对应的能耗数据计算第三变化率。
第二方面，本公开实施例提供一种异常用电站点检测装置，包括：
变化率计算模块，用于采集待检测站点的耗电数据、业务数据和能耗数据，确定所述耗电数据对应的第一变化率，所述业务数据对应的第二变化率以及所述能耗数据对应的第三变化率；
异常时间确定模块，用于在所述耗电数据符合目标正态分布的情况下，根据所述目标正态分布确定异常时间；
相关系数计算模块，用于根据所述第一变化率、所述第二变化率和所述第三变化率，确定变化率相关系数和变化率反相关时间；
异常用电站点确定模块，用于根据所述异常时间、所述变化率相关系数和所述变化率反相关时间，确定所述待检测站点中的异常用电站点。
第三方面，本公开实施例提供一种电子设备，包括处理器和存储有计算机程序的存储器，所述处理器执行所述计算机程序时实现第一方面所述的异常用电站点检测方法的步骤。
第四方面，本公开实施例提供一种可读存储介质，包括计算机程序，所述计算机程序被处理器执行时实现第一方面所述的异常用电站点检测方法的步骤。
本公开实施例提供的异常用电站点检测方法、装置、电子设备及可读存储介质，通过采集待检测站点的耗电数据、业务数据和能耗数据，确定各数据对应的变化率，在耗电数据符合目标正态分布的情况下，根据目标正态分布确定异常时间，然后根据各数据对应的变化率，确定变化率相关系数和变化率反相关时间，最终根据异常时间、变化率相关系数和变化率反相关时间，确定待检测站点中的异常用电站点，通过待检测站点的耗电数据、业务数据和能耗数据计算并检测异常用电站点，实现异常用电站点的精确识别。
附图说明
为了更清楚地说明本公开或相关技术中的技术方案，下面将对实施例或相关技术描述中所需要使用的附图作一简单地介绍，显而易见地，下面描述中的附图是本公开的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1是本公开实施例提供的异常用电站点检测方法的流程示意图之一；
图2是本公开实施例提供的异常用电站点检测方法的流程示意图之二；
图3是本公开实施例提供的异常用电站点检测装置的结构示意图；
图4是本公开实施例提供的电子设备的结构示意图。
具体实施方式
为使本公开的目的、技术方案和优点更加清楚，下面将结合本公开实施例中的附图，对本公开中的技术方案进行清楚、完整地描述，显然，所描述 的实施例是本公开一部分实施例，而不是全部的实施例。基于本公开中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本公开保护的范围。
参照图1，图1是本公开实施例中异常用电站点检测方法的流程示意图之一。本公开实施例提供的异常用电站点检测方法，可以包括：
步骤S100，采集待检测站点的耗电数据、业务数据和能耗数据，确定所述耗电数据对应的第一变化率，所述业务数据对应的第二变化率以及所述能耗数据对应的第三变化率；
具体地，通信运营商网络机房广泛分布在全国各处，机房用电无人看管，偷漏电现象时有发生，因此，及时准确地检测出网络机房中出现偷漏电现象的机房便成为了亟待解决的技术问题。本实施例中的待检测站点即是指待检测偷漏电现象的机房，通过机房内的设备(例如电表)采集待检测站点的耗电数据、业务数据和能耗数据，采集周期可以是30天或者60天，采集时间点可以是0点，采集得到的耗电数据包括采集周期内待检测站点每天的耗电量，业务数据和能耗数据相同，以时间先后顺序对耗电数据、业务数据和能耗数据进行排序，得到各待检测站点的耗电数据、业务数据和能耗数据。
以按时间排序得到的耗电数据为例，确定耗电数据对应的第一变化率的方法是：第N天的变化率＝(第N天的耗电数据－第N-1天的耗电数据)×100÷第N天的耗电数据。业务数据对应的第二变化率以及能耗数据对应的第三变化率计算方法与上述耗电数据对应的第一变化率相同。
步骤S200，在所述耗电数据符合目标正态分布的情况下，根据所述目标正态分布确定异常时间；
具体地，以时间先后顺序对耗电数据、业务数据和能耗数据进行排序，得到各待检测站点的耗电数据、业务数据和能耗数据后，判断耗电数据是否符合本实施例中的目标正态分布，具体方法如下：每次从总体耗电数据中随机抽取多个不同日期的耗电数据求出平均值，随机抽取多次，记录每次抽取的平均值，并且多次抽取需覆盖所有日期的耗电数据。校验每次抽取的平均值是否符合正态分布，校验公式如下：
Figure PCTCN2022143575-appb-000001
其中，n为抽 取次数，
Figure PCTCN2022143575-appb-000002
为n次抽取的均值，x i为第i次抽取的平均值,x j为第j次抽取的平均值，exp为以自然常数e为底的指数函数。
Figure PCTCN2022143575-appb-000003
假设每次抽取的平均值符合正态分布，置信度α设定为0.05，计算T ep，如果T ep落入拒绝域，{T EP≥T 1-α,EP(n)}，则不属于正态分布，拒绝域和EP检验查表信息如下表1所示。
Figure PCTCN2022143575-appb-000004
表1
如果待检测站点的耗电数据属于正态分布，则按照偏离样本均值超过3倍的标准差筛选潜在异常数据的日期，统计发生潜在发生异常数据的天数，即本实施例中的异常时间的天数。
步骤S300，根据所述第一变化率、所述第二变化率和所述第三变化率，确定变化率相关系数和变化率反相关时间；
具体地，本实施例中的变化率相关系数包括第一变化率相关系数和第二变化率相关系数，变化率反相关时间包括第一变化率反相关时间和第二变化率反相关时间。确定耗电数据对应的第一变化率(以E表示)的方法是：第N天的变化率＝(第N天的耗电数据－第N-1天的耗电数据)×100÷第N天的耗电数据；确定能耗数据对应的第三变化率(以C表示)的方法是：第N天的变化率＝(第N天的能耗数据－第N-1天的能耗数据)×100÷第N天的能耗数据。则根据第一变化率、第三变化率和预设系数计算公式，确定第二变化 率相关系数r(E，C)的方法是：
Figure PCTCN2022143575-appb-000005
Cov用于计算协方差，Var用于计算方差。第一变化率相关系数的计算方法与第二变化率相关系数相同。得到第一变化率相关系数与第二变化率相关系数后，在第一变化率相关系数小于预设阈值的情况下，根据第一变化率相关系数确定第一变化率反相关时间的方法是：按照时间顺序对齐耗电数据对应的第一变化率和能耗数据对应的第三变化率，若时间对齐的第一变化率和第三变化率正负相反，则确定该天为第二变化率反相关时间。第一变化率反相关时间的确定方法与第二变化率反相关时间的确定方法相同。
步骤S400，根据所述异常时间、所述变化率相关系数和所述变化率反相关时间，确定所述待检测站点中的异常用电站点。
具体地，得到异常时间、变化率相关系数和变化率反相关时间等特征数据后，将所有的站点用电特征，抽象为一组一组的向量。向量包含的内容如下：[站点耗电数据均值，方差，峰度，偏度统计信息，第一变化率相关系数，第二变化率相关系数，变化率反相关时间天数，反相关偏离幅度值]。通过向量内积，计算出待检测站点的特征向量与已知存在偷电行为的站点的特征向量之间的夹角弧度值，取待检测站点中最小弧度值对应的待预测站点作为检测结果。a＝{x 1 x 2 ... x n}，b＝{y 1 y 2 ... y n}，
Figure PCTCN2022143575-appb-000006
Figure PCTCN2022143575-appb-000007
a为待检测站点的特征向量，b为已知存在偷电行为的站点的特征向量，cosθ为待检测站点的特征向量与已知存在偷电行为的站点的特征向量之间的夹角弧度值。按照从小到大的顺序对各夹角弧度值排序，其中排序靠前的待检测站点存在偷电行为的可能性越大，输出为异常用电站点。
本实施例通过采集待检测站点的耗电数据、业务数据和能耗数据，确定各数据对应的变化率，在耗电数据符合目标正态分布的情况下，根据目标正态分布确定异常时间，然后根据各数据对应的变化率，确定变化率相关系数和变化率反相关时间，最终根据异常时间、变化率相关系数和变化率反相关时间，确定待检测站点中的异常用电站点，通过待检测站点的耗电数据、业务数据和能耗数据计算并检测异常用电站点，实现异常用电站点的精确识别。
在一个实施例中，本公开实施例提供的异常用电站点检测方法，还可以包括：
步骤S101，在所述耗电数据中存在缺失数据的情况下，确定所述缺失数 据对应的第一时间；
步骤S102，获取所述第一时间对应的数据序列，根据预设插值法和所述数据序列，对所述缺失数据进行补充。
具体地，由于数据采集不稳定等原因，某些待检测站点的耗电数据、业务数据和能耗数据可能存在数据缺失的情况，针对缺失的数据(例如某天耗电量为空值)，鉴于数据在短时间内相对稳定的特性，本实施例采用均值补差的方式对缺失数据进行填充处理。以时间先后顺序对耗电数据、业务数据和能耗数据进行排序(即本实施例中的数据序列)，得到各待检测站点的耗电数据、业务数据和能耗数据之后，采用以下均值补差的方式(即本实施中的预设插值法)对缺失数据进行填充处理，首先确定缺失数据在按照时间排序的耗电数据或业务数据或能耗数据中的位置，计算缺失数据时间顺序前后数据的平均值，将缺失数据时间顺序前后数据的平均值作为缺失数据，对缺失数据进行补充。
本实施例通过预设插值法和时间排序，对缺失数据进行补充，保证了数据的完整性，为后续实现异常用电站点的精确识别奠定了数据基础。
在一个实施例中，本公开实施例提供的异常用电站点检测方法，还可以包括：
步骤S110，根据预设抽样方式确定所述耗电数据对应的均值序列；
步骤S120，根据预设校验公式对所述均值序列进行检验，得到校验结果；
步骤S130，根据所述检验结果判断所述耗电数据是否符合目标正态分布。
具体地，以时间先后顺序对耗电数据、业务数据和能耗数据进行排序，得到各待检测站点的耗电数据、业务数据和能耗数据后，判断耗电数据是否符合本实施例中的目标正态分布，具体方法如下：根据预设抽样方式每次从总体耗电数据中随机抽取多个(例如，5-8个)不同日期的耗电数据求出平均值，随机抽取多次(例如，200次)，记录每次抽取的平均值，得到耗电数据对应的均值序列，并且多次抽取需覆盖所有日期的耗电数据。校验耗电数据对应的均值序列是否符合正态分布，校验公式(即本实施例中的预设校验公式)如下：
Figure PCTCN2022143575-appb-000008
其中，n为抽取次数，
Figure PCTCN2022143575-appb-000009
为n次抽取的均值，x i为第i次抽取的平均值,x j为第j次抽 取的平均值，exp为以自然常数e为底的指数函数。
Figure PCTCN2022143575-appb-000010
假设每次抽取的平均值符合正态分布，置信度α设定为0.05，计算T ep，如果T ep落入拒绝域，{T EP≥T 1-α,EP(n)}，则不属于正态分布，假设n＝200，α＝0.05，T 1-α， EP(n)＝0.379，将n＝200，α＝0.05代入上述公式(1)，若得到的T ep≥T 1-α， EP(n)，则判断耗电数据符合目标正态分布。
本实施例通过预设抽样方式和预设校验公式判断耗电数据是否符合目标正态分布，初步对耗电数据进行校验，为后续实现异常用电站点的精确识别奠定了数据基础。
在一个实施例中，本公开实施例提供的异常用电站点检测方法，还可以包括：
步骤S210，根据所述目标正态分布对应的标准差，确定所述耗电数据中的偏离数据；
步骤S220，确定所述偏离数据的采集时间，并将所述采集时间作为异常时间。
具体地，如果待检测站点的耗电数据属于正态分布，则按照偏离样本均值超过3倍的标准差筛选潜在异常数据(即本实施例中的偏离数据)的日期，统计发生潜在发生异常数据的天数，即本实施例中的异常时间的天数。
本实施例通过目标正态分布筛选潜在异常数据，初步对耗电数据进行筛选，为后续实现异常用电站点的精确识别奠定了数据基础。
在一个实施例中，本公开实施例提供的异常用电站点检测方法，还可以包括：
步骤S310，根据所述第一变化率、所述第二变化率和预设系数计算公式，确定第一变化率相关系数；
步骤S320，根据所述第一变化率、所述第三变化率和预设系数计算公式，确定第二变化率相关系数；
步骤S330，在所述第一变化率相关系数小于预设阈值的情况下，根据所述第一变化率相关系数确定第一变化率反相关时间；
步骤S340，在所述第二变化率相关系数小于预设阈值的情况下，根据所述第二变化率相关系数确定第二变化率反相关时间。
具体地，确定耗电数据对应的第一变化率(以E表示)的方法是：第N天的变化率＝(第N天的耗电数据－第N-1天的耗电数据)×100÷第N天的耗电数据；确定能耗数据对应的第三变化率(以C表示)的方法是：第N天的变化率＝(第N天的能耗数据－第N-1天的能耗数据)×100÷第N天的能耗数据。则根据第一变化率、第三变化率和预设系数计算公式，确定第二变化率相关系数r(E，C)的方法是：
Figure PCTCN2022143575-appb-000011
Cov用于计算协方差，Var用于计算方差。第一变化率相关系数的计算方法与第二变化率相关系数相同。得到第一变化率相关系数与第二变化率相关系数后，在第一变化率相关系数小于预设阈值的情况下，根据第一变化率相关系数确定第一变化率反相关时间的方法是：按照时间顺序对齐耗电数据对应的第一变化率和能耗数据对应的第三变化率，若时间对齐的第一变化率和第三变化率正负相反，则确定该天为第二变化率反相关时间。第一变化率反相关时间的确定方法与第二变化率反相关时间的确定方法相同。
本实施例通过变化率和预设系数计算公式确定变化率相关系数和变化率反相关时间，得到确定异常用电站点的数据，实现异常用电站点的精确识别。
参照图2，图2是本公开实施例中异常用电站点检测方法的流程示意图之二，在一个实施例中，本公开实施例提供的异常用电站点检测方法，还可以包括：
步骤S410，根据所述耗电数据、所述异常时间、所述变化率相关系数和所述变化率反相关时间，确定所述待检测站点对应的第一特征向量；
步骤S420，获取预设异常站点对应的第二特征向量，计算所述第一特征向量与所述第二特征向量之间的夹角弧度值；
步骤S430，根据所述夹角弧度值的大小对各所述待检测站点进行排序，根据排序结果确定各所述待检测站点中的异常用电站点。
具体地，得到异常时间、变化率相关系数和变化率反相关时间等特征数据后，将所有的站点用电特征，抽象为一组一组的向量，即本实施例中的第一特征向量。第一特征向量包含的内容如下：[站点耗电数据均值，方差，峰度，偏度统计信息，第一变化率相关系数，第二变化率相关系数，变化率反相关时间天数，反相关偏离幅度值]。通过向量内积，计算出待检测站点的特征向量与已知存在偷电行为的站点(即本实施例中的预设异常站点)的特 征向量(即本实施例中的第二特征向量)之间的夹角弧度值，取待检测站点中最小弧度值对应的待预测站点作为检测结果。a＝{x 1 x 2 ... x n}，b＝{y 1 y 2 ... y n}，
Figure PCTCN2022143575-appb-000012
a为待检测站点的特征向量，b为已知存在偷电行为的站点的特征向量，cosθ为待检测站点的特征向量与已知存在偷电行为的站点的特征向量之间的夹角弧度值。按照从小到大的顺序对各夹角弧度值排序，其中排序靠前的待检测站点存在偷电行为的可能性越大，输出为异常用电站点。
本实施例通过异常时间、变化率相关系数和变化率反相关时间等特征数据，确定异常用电站点，实现异常用电站点的精确识别。
在一个实施例中，本公开实施例提供的异常用电站点检测方法，还可以包括：
步骤S103，在所述耗电数据、所述业务数据以及所述能耗数据具有相同采集时间和相同采集周期的情况下，根据各所述采集时间对应的耗电数据计算第一变化率；
步骤S104，根据各所述采集时间对应的业务数据计算第二变化率，根据各所述采集时间对应的能耗数据计算第三变化率。
具体地，在耗电数据、业务数据以及能耗数据具有相同采集时间和相同采集周期的情况下，以60天采集周期，0点采集时间，按时间排序得到的耗电数据为例，确定耗电数据对应的第一变化率的方法是：第N天的变化率＝(第N天的耗电数据－第N-1天的耗电数据)×100÷第N天的耗电数据。业务数据对应的第二变化率以及能耗数据对应的第三变化率计算方法与上述耗电数据对应的第一变化率相同。
本实施例通过计算各数据对应的变化率，为后续实现异常用电站点的精确识别奠定了数据基础。
参考图3，图3是本公开实施例中异常用电站点检测装置的结构示意图，下面对本公开实施例提供的异常用电站点检测装置进行描述，下文描述的异常用电站点检测装置与上文描述的异常用电站点检测方法可相互对应参照。
异常用电站点检测装置，包括：
变化率计算模块301，用于采集待检测站点的耗电数据、业务数据和能耗数据，确定所述耗电数据对应的第一变化率，所述业务数据对应的第二变化率以及所述能耗数据对应的第三变化率；
异常时间确定模块302，用于在所述耗电数据符合目标正态分布的情况下，根据所述目标正态分布确定异常时间；
相关系数计算模块303，用于根据所述第一变化率、所述第二变化率和所述第三变化率，确定变化率相关系数和变化率反相关时间；
异常用电站点确定模块304，用于根据所述异常时间、所述变化率相关系数和所述变化率反相关时间，确定所述待检测站点中的异常用电站点。
可选地，所述异常用电站点检测装置，还包括：
第一时间确定模块，用于在所述耗电数据中存在缺失数据的情况下，确定所述缺失数据对应的第一时间；
缺失数据补充模块，用于获取所述第一时间对应的数据序列，根据预设插值法和所述数据序列，对所述缺失数据进行补充。
可选地，所述异常用电站点检测装置，还包括：
均值序列确定模块，用于根据预设抽样方式确定所述耗电数据对应的均值序列；
校验结果确定模块，用于根据预设校验公式对所述均值序列进行检验，得到校验结果；
判断模块，用于根据所述检验结果判断所述耗电数据是否符合目标正态分布。
可选地，所述异常时间确定模块，包括：
偏离数据确定单元，用于根据所述目标正态分布对应的标准差，确定所述耗电数据中的偏离数据；
异常时间确定单元，用于确定所述偏离数据的采集时间，并将所述采集时间作为异常时间。
可选地，所述相关系数计算模块，包括：
第一变化率相关系数确定单元，用于根据所述第一变化率、所述第二变化率和预设系数计算公式，确定第一变化率相关系数；
第二变化率相关系数确定单元，用于根据所述第一变化率、所述第三变化率和预设系数计算公式，确定第二变化率相关系数；
第一变化率反相关时间确定单元，用于在所述第一变化率相关系数小于预设阈值的情况下，根据所述第一变化率相关系数确定第一变化率反相关时间；
第二变化率反相关时间确定单元，用于在所述第二变化率相关系数小于预设阈值的情况下，根据所述第二变化率相关系数确定第二变化率反相关时间。
可选地，所述异常用电站点确定模块，包括：
第一特征向量确定单元，用于根据所述耗电数据、所述异常时间、所述变化率相关系数和所述变化率反相关时间，确定所述待检测站点对应的第一特征向量；
夹角弧度值计算单元，用于获取预设异常站点对应的第二特征向量，计算所述第一特征向量与所述第二特征向量之间的夹角弧度值；
异常用电站点确定单元，用于根据所述夹角弧度值的大小对各所述待检测站点进行排序，根据排序结果确定各所述待检测站点中的异常用电站点。
可选地，所述变化率计算模块，包括：
第一变化率计算单元，用于在所述耗电数据、所述业务数据以及所述能耗数据具有相同采集时间和相同采集周期的情况下，根据各所述采集时间对应的耗电数据计算第一变化率；
第三变化率计算单元，用于根据各所述采集时间对应的业务数据计算第二变化率，根据各所述采集时间对应的能耗数据计算第三变化率。
图4示例了一种电子设备的实体结构示意图，如图4所示，该电子设备可以包括：处理器(processor)410、通信接口(Communication Interface)420、存储器(memory)430和通信总线440，其中，处理器410，通信接口420，存储器430通过通信总线440完成相互间的通信。处理器410可以调用存储器430中的计算机程序，以执行异常用电站点检测方法的步骤。
此外，上述的存储器430中的逻辑指令可以通过软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读取存储介质 中。基于这样的理解，本公开的技术方案本质上或者说对相关技术做出贡献的部分或者该技术方案的部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行本公开各个实施例所述方法的全部或部分步骤。而前述的存储介质包括：U盘、移动硬盘、只读存储器(ROM，Read-Only Memory)、随机存取存储器(RAM，Random Access Memory)、磁碟或者光盘等各种可以存储程序代码的介质。
另一方面，本公开实施例还提供一种可读存储介质，所述可读存储介质包括计算机程序，所述计算机程序可存储在非暂态计算机可读存储介质上，所述计算机程序被处理器执行时，计算机能够执行上述各实施例所提供的异常用电站点检测方法的步骤。
以上所描述的装置实施例仅仅是示意性的，其中所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部模块来实现本实施例方案的目的。本领域普通技术人员在不付出创造性的劳动的情况下，即可以理解并实施。
通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到各实施方式可借助软件加必需的通用硬件平台的方式来实现，当然也可以通过硬件。基于这样的理解，上述技术方案本质上或者说对相关技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品可以存储在计算机可读存储介质中，如ROM/RAM、磁碟、光盘等，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行各个实施例或者实施例的某些部分所述的异常用电站点检测方法。
最后应说明的是：以上实施例仅用以说明本公开的技术方案，而非对其限制；尽管参照前述实施例对本公开进行了详细的说明，本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本公开各实施例技术方案的精神和范围。

Claims (10)

  1. 一种异常用电站点检测方法，包括：
    采集待检测站点的耗电数据、业务数据和能耗数据，确定所述耗电数据对应的第一变化率，所述业务数据对应的第二变化率以及所述能耗数据对应的第三变化率；
    在所述耗电数据符合目标正态分布的情况下，根据所述目标正态分布确定异常时间；
    根据所述第一变化率、所述第二变化率和所述第三变化率，确定变化率相关系数和变化率反相关时间；
    根据所述异常时间、所述变化率相关系数和所述变化率反相关时间，确定所述待检测站点中的异常用电站点。
  2. 根据权利要求1所述的异常用电站点检测方法，其中，所述采集待检测站点的耗电数据、业务数据和能耗数据的步骤之后包括：
    在所述耗电数据中存在缺失数据的情况下，确定所述缺失数据对应的第一时间；
    获取所述第一时间对应的数据序列，根据预设插值法和所述数据序列，对所述缺失数据进行补充。
  3. 根据权利要求1所述的异常用电站点检测方法，其中，所述采集待检测站点的耗电数据、业务数据和能耗数据，确定所述耗电数据对应的第一变化率，所述业务数据对应的第二变化率以及所述能耗数据对应的第三变化率的步骤之后包括：
    根据预设抽样方式确定所述耗电数据对应的均值序列；
    根据预设校验公式对所述均值序列进行检验，得到校验结果；
    根据所述检验结果判断所述耗电数据是否符合目标正态分布。
  4. 根据权利要求1所述的异常用电站点检测方法，其中，所述根据所述 目标正态分布确定异常时间的步骤包括：
    根据所述目标正态分布对应的标准差，确定所述耗电数据中的偏离数据；
    确定所述偏离数据的采集时间，并将所述采集时间作为异常时间。
  5. 根据权利要求1所述的异常用电站点检测方法，其中，所述根据所述第一变化率、所述第二变化率和所述第三变化率，确定变化率相关系数和变化率反相关时间的步骤包括：
    根据所述第一变化率、所述第二变化率和预设系数计算公式，确定第一变化率相关系数；
    根据所述第一变化率、所述第三变化率和预设系数计算公式，确定第二变化率相关系数；
    在所述第一变化率相关系数小于预设阈值的情况下，根据所述第一变化率相关系数确定第一变化率反相关时间；
    在所述第二变化率相关系数小于预设阈值的情况下，根据所述第二变化率相关系数确定第二变化率反相关时间。
  6. 根据权利要求1所述的异常用电站点检测方法，其中，所述根据所述异常时间、所述变化率相关系数和所述变化率反相关时间，确定所述待检测站点中的异常用电站点的步骤包括：
    根据所述耗电数据、所述异常时间、所述变化率相关系数和所述变化率反相关时间，确定所述待检测站点对应的第一特征向量；
    获取预设异常站点对应的第二特征向量，计算所述第一特征向量与所述第二特征向量之间的夹角弧度值；
    根据所述夹角弧度值的大小对各所述待检测站点进行排序，根据排序结果确定各所述待检测站点中的异常用电站点。
  7. 根据权利要求1所述的异常用电站点检测方法，其中，所述确定所述耗电数据对应的第一变化率，所述业务数据对应的第二变化率以及所述能耗数据对应的第三变化率的步骤包括：
    在所述耗电数据、所述业务数据以及所述能耗数据具有相同采集时间和相同采集周期的情况下，根据各所述采集时间对应的耗电数据计算第一变化率；
    根据各所述采集时间对应的业务数据计算第二变化率，根据各所述采集时间对应的能耗数据计算第三变化率。
  8. 一种异常用电站点检测装置，包括：
    变化率计算模块，用于采集待检测站点的耗电数据、业务数据和能耗数据，确定所述耗电数据对应的第一变化率，所述业务数据对应的第二变化率以及所述能耗数据对应的第三变化率；
    异常时间确定模块，用于在所述耗电数据符合目标正态分布的情况下，根据所述目标正态分布确定异常时间；
    相关系数计算模块，用于根据所述第一变化率、所述第二变化率和所述第三变化率，确定变化率相关系数和变化率反相关时间；
    异常用电站点确定模块，用于根据所述异常时间、所述变化率相关系数和所述变化率反相关时间，确定所述待检测站点中的异常用电站点。
  9. 一种电子设备，包括处理器和存储有计算机程序的存储器，所述处理器执行所述计算机程序时实现权利要求1至7任一项所述的异常用电站点检测方法的步骤。
  10. 一种非暂态计算机可读存储介质，其上存储有计算机程序，所述计算机程序被处理器执行时实现如权利要求1至7任一项所述异常用电站点检测方法。
PCT/CN2022/143575 2022-06-22 2022-12-29 异常用电站点检测方法、装置、电子设备及可读存储介质 WO2023246070A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210713265.0 2022-06-22
CN202210713265.0A CN117312744A (zh) 2022-06-22 2022-06-22 异常用电站点检测方法、装置、电子设备及可读存储介质

Publications (1)

Publication Number Publication Date
WO2023246070A1 true WO2023246070A1 (zh) 2023-12-28

Family

ID=89279967

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/143575 WO2023246070A1 (zh) 2022-06-22 2022-12-29 异常用电站点检测方法、装置、电子设备及可读存储介质

Country Status (2)

Country Link
CN (1) CN117312744A (zh)
WO (1) WO2023246070A1 (zh)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2006146459A (ja) * 2004-11-18 2006-06-08 Renesas Technology Corp 半導体デバイスの製造方法および製造システム
CN103488867A (zh) * 2013-07-16 2014-01-01 深圳市航天泰瑞捷电子有限公司 一种用电异常用户自动筛选方法
CN105228175A (zh) * 2015-09-17 2016-01-06 福建新大陆软件工程有限公司 一种基于决策树的基站能耗优化方法及系统
CN110458230A (zh) * 2019-08-12 2019-11-15 江苏方天电力技术有限公司 一种基于多判据融合的配变用采数据异常甄别方法
JP2020150692A (ja) * 2019-03-14 2020-09-17 三菱電機株式会社 モータ劣化傾向監視システム

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2006146459A (ja) * 2004-11-18 2006-06-08 Renesas Technology Corp 半導体デバイスの製造方法および製造システム
CN103488867A (zh) * 2013-07-16 2014-01-01 深圳市航天泰瑞捷电子有限公司 一种用电异常用户自动筛选方法
CN105228175A (zh) * 2015-09-17 2016-01-06 福建新大陆软件工程有限公司 一种基于决策树的基站能耗优化方法及系统
JP2020150692A (ja) * 2019-03-14 2020-09-17 三菱電機株式会社 モータ劣化傾向監視システム
CN110458230A (zh) * 2019-08-12 2019-11-15 江苏方天电力技术有限公司 一种基于多判据融合的配变用采数据异常甄别方法

Also Published As

Publication number Publication date
CN117312744A (zh) 2023-12-29

Similar Documents

Publication Publication Date Title
WO2021073114A1 (zh) 基于统计的异常流量监测方法、装置、设备及存储介质
CN109474483A (zh) 一种设备异常情况的检测方法、检测装置及终端设备
CN113127305B (zh) 异常检测方法及装置
CN106548402B (zh) 资源转移监控方法及装置
CN112198375A (zh) 单相用户的相位识别方法、装置、设备和存储介质
CN103731866A (zh) 检测用户终端性能的方法和系统
CN108537692A (zh) 小区楼盘数据采集方法及装置
CN114978877A (zh) 一种异常处理方法、装置、电子设备及计算机可读介质
CN111951125A (zh) 一种基于大数据分析的台区异常户变关系识别方法
CN116523140A (zh) 窃电检测方法、装置、电子设备和存储介质
WO2023246070A1 (zh) 异常用电站点检测方法、装置、电子设备及可读存储介质
CN114595765A (zh) 数据处理方法、装置、电子设备及存储介质
CN117235731B (zh) 一种保密装备大数据监控预警系统
CN111612054B (zh) 一种基于非负矩阵分解和密度聚类的用户窃电行为识别方法
CN105630762A (zh) 辨识机密数据的方法及电子装置
CN110502346A (zh) 一种集群环境下资源信息管理系统及方法
CN114219245B (zh) 基于大数据的乡村电力指数评价方法、装置及存储介质
CN116342305A (zh) 一种差旅费报销方法、装置、计算机设备及存储介质
Chu et al. Anomaly detection and clustering‐based identification method for consumer–transformer relationship and associated phase in low‐voltage distribution systems
JP7156737B2 (ja) ゴミ回収の要否の判定装置、判定方法、プログラム、および記録媒体
CN109391738B (zh) 一种根据移动终端信息进行预警的方法
CN112417007A (zh) 一种数据分析方法、装置、电子设备及存储介质
CN105938610A (zh) 一种用电采集系统中违约用电判别与报表自动生成方法
CN105337994B (zh) 基于网络流量的恶意代码检测方法及装置
CN110163083A (zh) 用户信息的匹配方法、装置及终端设备

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22947805

Country of ref document: EP

Kind code of ref document: A1