WO2020015060A1

WO2020015060A1 - Power consumption anomaly estimation method and apparatus, device, and computer storage medium

Info

Publication number: WO2020015060A1
Application number: PCT/CN2018/103335
Authority: WO
Inventors: 孙闳绅; 金戈; 徐亮; 肖京
Original assignee: 平安科技（深圳）有限公司
Priority date: 2018-07-17
Filing date: 2018-08-30
Publication date: 2020-01-23
Also published as: CN109726885A

Abstract

A power consumption anomaly estimation method and apparatus, a device, and a computer storage medium. The method comprises the following steps: receiving a power consumption anomaly estimation request, obtaining time point information in the power consumption anomaly estimation request, and obtaining power consumption feature information related to the time point information (S10); inputting the power consumption feature information and the time point information to a preset regression model to obtain theoretical power consumption corresponding to the time point information (S20); obtaining actual power consumption corresponding to the time point information and comparing the actual power consumption with the theoretical power consumption to determine whether the actual power consumption is abnormal (S30); and if the actual power consumption is abnormal, generating power consumption anomaly prompt information (S40).The method can improve the accuracy and efficiency of power consumption anomaly estimation.

Description

Method, device, equipment and computer storage medium for evaluating power consumption abnormality Ranch

This application claims the priority of a Chinese patent application filed with the Chinese Patent Office on July 17, 2018, with the application number 201810799777.7, and the invention name is "Method, Apparatus, Equipment, and Computer Storage Medium for Evaluating Abnormal Power Consumption." Citations are incorporated in the application.

Technical field

The present application relates to the field of power detection, and in particular, to a method, a device, a device, and a computer storage medium for evaluating abnormal power consumption.

Background technique

Due to the current dependence of people on electricity, in order to prevent the impact of sudden power outages on people's lives, people use different methods to detect abnormal power consumption.

There are different ways to evaluate current power consumption abnormalities, but they are all based on historical data. For example, the detection equipment collects data on the power consumption of each floor in real time. Take the average of the electrical data, take the average value of the electricity as the electricity constant, and set an electricity consumption interval based on the electricity constant. When the actual electricity consumption is not in this electricity interval, it is determined as an abnormal electricity consumption, that is, The current evaluation of abnormal power consumption must rely on historical data. Such an abnormal evaluation scheme of power consumption has the following deficiencies. For example, the abnormal evaluation is too one-sided due to the influence of abnormal values. In addition, the efficiency of abnormal evaluation is not high. The accuracy and efficiency of assessment has become a technical problem that needs to be solved urgently.

Summary of the invention

The main purpose of this application is to provide a method, device, equipment and computer storage medium for evaluating abnormality in power consumption, which aims to improve the accuracy and efficiency of detecting abnormality in power consumption.

In order to achieve the above object, the present application provides a method for evaluating abnormality in power consumption, which includes the following steps:

Receiving a power consumption abnormality evaluation request, acquiring time node information in the power consumption abnormality evaluation request, and acquiring power consumption characteristic information related to the time node information;

Inputting the power consumption characteristic information and the time node information into a preset regression model to obtain a theoretical power consumption corresponding to the time node information;

Acquiring the actual power consumption corresponding to the time node information, and comparing the actual power consumption with the theoretical power consumption to determine whether the actual power consumption is abnormal;

If the actual power consumption is abnormal, the power consumption abnormality prompt information is generated.

Optionally, before the steps of receiving the power consumption abnormality evaluation request, obtaining the time node information in the power consumption abnormality evaluation request, and obtaining the power consumption characteristic information related to the time node information, the steps include:

Obtaining power samples from a preset power sample set, classifying each of the power samples according to a preset classification rule, and obtaining n power sample subsets;

The following steps are performed for each of the power sample subsets:

Using the power sample subset as a target power sample subset, generating an initial regression model based on the target power sample subset, and obtaining other power sample subsets in the n power sample subsets excluding the target power sample subset Using the other power sample subsets to iteratively train the initial regression model to obtain a regression submodel corresponding to the target power sample subset;

Encapsulating the regression sub-model corresponding to each of the target power sample subsets to generate a preset regression model.

Optionally, the step of inputting the power consumption characteristic information and the time node information into a preset regression model to obtain a theoretical power consumption corresponding to the time node information includes:

Inputting the power consumption characteristic information and the time node information into each of the regression sub-models of the preset regression model to obtain a basic power consumption corresponding to each of the regression sub-models;

Sum the respective basic power consumptions and calculate an average value to obtain a theoretical power consumption corresponding to the time node information.

Optionally, after the step of acquiring the actual power consumption corresponding to the time node information, and comparing the actual power consumption with the theoretical power consumption to determine whether the actual power consumption is abnormal ,include:

If the actual power consumption is normal, using the time node information and power characteristic information related to the time node information as an updated power sample;

Saving the updated power sample to the preset power sample set to obtain an updated power sample set;

When receiving a preset regression model update request, obtaining the updated power sample in the updated power sample set;

Each regression sub-model in the preset regression model is adjusted based on the updated power sample to obtain an updated regression sub-model, and each of the updated regression sub-models is packaged to generate an updated regression model.

Optionally, the steps of receiving a power consumption abnormality evaluation request, obtaining time node information in the power consumption abnormality evaluation request, and obtaining power consumption characteristic information related to the time node information include:

Receiving a power consumption abnormality evaluation request, acquiring time node information in the power consumption abnormality evaluation request, and collecting temperature information and holiday information corresponding to the time node information;

Use the temperature information and the holiday information as the power consumption characteristic information related to the time node information.

Optionally, the step of obtaining the actual power consumption corresponding to the time node information, and comparing the actual power consumption with the theoretical power consumption to determine whether the actual power consumption is abnormal, include:

Acquiring the actual power consumption corresponding to the time node information, and calculating a ratio between the actual power consumption and the theoretical power consumption;

If the ratio between the actual power consumption and the theoretical power consumption exceeds a preset threshold, determining that the actual power consumption is abnormal;

If the ratio between the actual power consumption and the theoretical power consumption does not exceed a preset threshold, it is determined that the actual power consumption is normal.

Optionally, if the actual power consumption is abnormal, the step of generating prompt information of abnormal power consumption includes:

If the actual power consumption is abnormal, obtaining historical power consumption corresponding to the time node information;

Calculating the actual power consumption and the historical power consumption according to a preset synchronization formula to obtain a year-on-year growth rate;

Generate power consumption abnormality prompt information based on the year-on-year growth rate and the actual power consumption for users to view.

In addition, in order to achieve the foregoing object, the present application further provides an apparatus for evaluating abnormality in power consumption. The apparatus for evaluating abnormality in power consumption includes:

A receiving and acquiring module, configured to receive a request for abnormal evaluation of power consumption, acquire time node information in the request for abnormal evaluation of power consumption, and obtain power consumption characteristic information related to the time node information;

An input calculation module, configured to input the power consumption characteristic information and the time node information into a preset regression model to obtain a theoretical power consumption corresponding to the time node information;

An acquisition judgment module, configured to acquire the actual power consumption corresponding to the time node information, and compare the actual power consumption with the theoretical power consumption to determine whether the actual power consumption is abnormal;

A prompting module is configured to generate prompting information of abnormal power consumption if the actual power consumption is abnormal.

In addition, in order to achieve the above-mentioned object, the present application also provides a power consumption abnormality evaluation device;

The power consumption abnormality evaluation device includes: a memory, a processor, and computer-readable instructions stored on the memory and executable on the processor, wherein:

When the computer-readable instructions are executed by the processor, the steps of the power consumption abnormality assessment method described above are implemented.

In addition, in order to achieve the above purpose, the present application also provides a computer storage medium;

Computer-readable instructions are stored on the computer storage medium, and when the computer-readable instructions are executed by a processor, the steps of the method for evaluating abnormality in power consumption as described above are implemented.

A method, device, equipment, and computer storage medium for evaluating abnormal power consumption according to the embodiments of the present application. In this application, a power consumption calculation regression model is established in advance. When the server receives a request for abnormal power consumption evaluation, the power consumption is obtained. The abnormality evaluation requests the current actual power consumption value, calculates a theoretical power consumption value based on a preset regression model, and then compares the actual power consumption value with the theoretical power consumption value to determine whether the power consumption is abnormal. In this embodiment, the power consumption is abnormal. The evaluation does not rely on historical data, which can effectively exclude the impact of historical power abnormal values on the power abnormal evaluation and improve the accuracy of the abnormal evaluation. At the same time, the server uses a preset regression model to perform power abnormal evaluation to make the power consumption The anomaly evaluation procedure is simpler and more efficient.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic structural diagram of a device for a hardware operating environment involved in a solution according to an embodiment of the present application; FIG.

FIG. 2 is a schematic flowchart of a first embodiment of an abnormality assessment method for power consumption of this application;

FIG. 3 is a detailed flowchart of step S40 of the method for evaluating abnormality of power consumption in FIG. 2;

FIG. 4 is a schematic diagram of functional modules of an embodiment of an abnormality assessment device for power consumption of the present application.

The implementation, functional features and advantages of the purpose of this application will be further described with reference to the embodiments and the drawings.

detailed description

It should be understood that the specific embodiments described herein are only used to explain the application, and are not used to limit the application.

As shown in FIG. 1, FIG. 1 is a server (also called a power consumption abnormality evaluation device) of a hardware operating environment involved in the solution of the embodiment of the present application. The power consumption abnormality evaluation device may be a separate power consumption abnormality evaluation device. The structure may also be a schematic diagram of the structure formed by combining other devices with an abnormality evaluation device for power consumption.

In the embodiment of the present application, a server refers to a computer that manages resources and provides services to users, and is generally divided into a file server, a database server, and an application server. A computer or computer system running the above software is also called a server. Compared with ordinary PC (personal For a personal computer, the server requires higher stability, security, and performance. As shown in Figure 1, the server may include a processor 1001, such as a central processing unit (Central Processing Unit, CPU), network interface 1004, user interface 1003, memory 1005, communication bus 1002, chipset, disk system, network and other hardware. The communication bus 1002 is used to implement connection and communication between these components. The user interface 1003 may include a display, an input unit such as a keyboard, and the optional user interface 1003 may further include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a wireless fidelity WIreless-FIdelity, WIFI interface). The memory 1005 may be a high-speed random access memory (random access memory (RAM), or non-volatile memory), such as disk storage. The memory 1005 may optionally be a storage device independent of the foregoing processor 1001.

Optionally, the server may further include a camera, RF (Radio Frequency (radio frequency) circuit, sensor, audio circuit, WiFi module; input unit, display screen, touch screen; optional network interface except wireless interface except WiFi, Bluetooth and so on. Those skilled in the art can understand that the server structure shown in FIG. 1 does not constitute a limitation on the server, and may include more or fewer components than shown in the figure, or some components may be combined, or different components may be arranged.

As shown in FIG. 1, the computer software product is stored in a storage medium (storage medium: also called computer storage medium, computer medium, readable medium, readable storage medium, computer-readable storage medium, or directly called medium, etc., such as RAM , Magnetic disks, optical disks, storage media refers to non-volatile readable storage media), including a number of instructions for a terminal device (can be a mobile phone, computer, server, air conditioner, or network device, etc.) to execute this In the method described in the various embodiments of the application, the memory 1005 as a computer storage medium may include an operating system, a network communication module, a user interface module, and computer-readable instructions.

In the server shown in FIG. 1, the network interface 1004 is mainly used to connect to a background database and perform data communication with the background database; the user interface 1003 is mainly used to connect to a client (client, also called a client or a terminal. The terminal may be a fixed terminal or a mobile terminal, which is not described in detail here), and performs data communication with the client; and the processor 1001 may be used to call computer-readable instructions stored in the memory 1005 and execute the following embodiments of the present application Provides steps in the method for assessing abnormal power consumption.

In the first embodiment of the method for evaluating abnormality in power consumption in the present application, the method for evaluating abnormality in power consumption includes the following steps:

Before the steps of the first embodiment, a developer needs to establish a preset regression model before the server can calculate a theoretical power consumption based on the preset regression model to compare the theoretical power consumption with the actual power consumption for Evaluation of abnormal user volume, specific steps for establishing a preset regression model, including:

Step S01: Obtain a power sample from a preset power sample set, classify each of the power samples according to a preset classification rule, and obtain n power sample subsets.

The server obtains a power sample from a preset power sample set, where the preset power sample set refers to pre-stored historical power related information, the server obtains the included power samples from the preset power sample set, and sets each power sample as preset The classification rules are used to classify and obtain n subsets of power samples. The preset classification rules refer to the preset power sample classification rules. For example, the preset classification rules are set to the collection time classification rules. The collection time is classified to obtain a subset of power samples corresponding to each year and month.

For example, the server will collect the historical power consumption and related information: the power consumption on Tuesday, June 5, 2018 from 13:00 to 13:05, the outdoor temperature is 30 degrees Celsius, and the address is xxx office building in Shenzhen Room, Guangdong Province. Information such as the working day is saved to the memory; when receiving a request to establish a preset regression model, the server randomly extracts a certain amount of historical power consumption and its related information from the memory as a power sample, and uses the extracted power samples Forms a preset power sample set, and the server classifies each power sample in the preset power sample set according to the power sample collection time to obtain n power sample subsets in different time periods, where the power samples in each power sample subset may be the same May also be different.

Step S02, the following steps are performed for each of the power sample subsets: using the power sample subset as a target power sample subset, generating an initial regression model based on the target power sample subset, and obtaining the n power samples The other power sample subsets of the target power sample subset are removed from the subset, and the initial regression model is iteratively trained using the other power sample subsets to obtain a regression submodel corresponding to the target power sample subset.

The server uses each of the power sample subsets as a target power sample subset, and generates an initial regression model based on the target power sample subset. The initial regression model is a function of power consumption and power characteristic data to establish an initial regression model. It is the characteristic data of each sample in the target power sample subset. The characteristic data includes: time data, temperature data, and holiday data. The function relationship between time data, temperature data, holiday data, and power consumption is determined according to a preset model. This functional relationship is used as the initial regression model; specifically, a model of characteristic data and power consumption is set in advance according to experience. When the initial regression model is established, each power sample in the target power sample subset is obtained, and each power sample is passed Determine the initial value of the parameter by equal division status, and assign the determined initial value of the parameter to the preset model to obtain the initial regression model; after the initial regression model is completed; set the maximum number of iterations and the convergence threshold; the server uses to divide the target Charges other than a subset of charge samples In this subset, iterative training is performed on the initial regression model until the number of iterations previously set or has been converged. At this time, the optimal model parameters can be obtained, and then the target power sample subset is obtained according to the optimal model parameters. Corresponding regression model.

That is, the server in this embodiment uses time node information, whether it is a holiday and temperature information as feature data, and sets corresponding weights for each feature data to generate an initial regression model corresponding to each target power sample subset. The generated initial regression model is related to the above characteristic data.

After the initial regression model is generated, the server iteratively trains the initial regression model using n-1 power sample subsets other than the target power sample subset, and the server generates a regression submodel corresponding to each target power sample subset.

In step S03, the regression sub-model corresponding to each of the target power sample subsets is encapsulated to generate a preset regression model.

The server obtains the regression submodel corresponding to each target power sample subset, encapsulates each regression submodel, and generates a preset regression model, that is, in this embodiment, the n regression submodels obtained by training are packaged as a preset Regression model.

In this embodiment, a preset regression model is created and generated based on historical power consumption information. The power consumption abnormality evaluation is performed based on the generated preset regression prediction, and the power consumption abnormality evaluation based on the generated preset regression prediction can effectively consider the time series characteristics without introducing too much the strong influence of time series on the time point. Effectively detect abnormal points. In addition, in this implementation, the method of establishing multiple regression sub-models is used in the scenario where a preset regression model is established, which effectively reduces the process of calculating the theoretical value of power consumption based on the preset regression model. Possible overfitting.

Referring to FIG. 2, in a first embodiment of a method for evaluating abnormality in power consumption according to the present application, the method for evaluating abnormality in power consumption includes:

Step S10: Receive a power consumption abnormality evaluation request, acquire time node information in the power consumption abnormality evaluation request, and acquire power consumption characteristic information related to the time node information.

The user triggers a power consumption abnormality evaluation request on the terminal. The terminal sends the power consumption abnormality evaluation request to the server. When the server receives the power consumption abnormality evaluation request, the server obtains the time node information in the power consumption abnormality evaluation request and obtains the The power consumption characteristic information related to the time node information, that is, when the server receives the power consumption abnormality evaluation request, the server obtains the time node included in the power consumption abnormality evaluation request, and collects temperature information and holiday information corresponding to the time node. ; Using the temperature information and the holiday information as the power consumption characteristic information related to the time node information.

For example, the server receives the power consumption abnormality evaluation request and obtains the time node information in the power consumption abnormality evaluation request: on May 2, 2018, the server obtains the power consumption characteristic information related to the time node including: Wednesday, work day, location It is the Shenzhen xxx office building in Guangdong Province and the temperature is 30 degrees Celsius, that is, the characteristic information collected in this embodiment will affect the abnormal evaluation of power consumption, that is, whether the rest day will affect the power consumption of the Shenzhen xxx office building in Guangdong Province. To some extent, the temperature will also affect the electricity consumption.

In step S20, the power consumption characteristic information and the time node information are input into a preset regression model to obtain a theoretical power consumption corresponding to the time node information.

The server inputs the power consumption characteristic information and time node information to each regression sub-model of the preset regression model to obtain the basic power consumption corresponding to each of the regression sub-models; that is, the server evaluates the power consumption abnormally. The power consumption characteristic information and time node in the request are input into each regression sub-model, and n basic power consumptions are obtained according to the calculation formula in each regression sub-model. The average value, and the average value obtained by adding and summing the n basic power consumptions as the theoretical power consumption corresponding to the time node information.

For example, the server inputs information such as temperature and holidays in time node and power consumption characteristic information to each regression sub-model, and each regression sub-model calculates the basic power consumption k ₁ , k ₂ up to k _n according to the input information; Power consumption k = (k ₁ + k ₂ +… + k _n ) / n.

Step S30: Acquire the actual power consumption corresponding to the time node information, and compare the actual power consumption with the theoretical power consumption to determine whether the actual power consumption is abnormal.

The server obtains the actual power consumption corresponding to the time node information, where the actual power consumption refers to the actual power consumption of the time node obtained by the detection device installed in the server in real time; The theoretical power consumption is compared to obtain a comparison result between the actual power consumption and the theoretical power consumption, and whether the actual power consumption is abnormal according to the comparison result. Wherein, according to a comparison result between the actual power consumption and the theoretical power consumption, the server may adopt different implementation methods to determine whether the actual power consumption is abnormal, specifically:

For example: perform a ratio calculation between the actual power consumption and the theoretical power consumption, and perform an abnormality determination based on the ratio between the actual power consumption and the theoretical power consumption, that is,

Step a1: Obtain an actual power consumption corresponding to the time node information, and calculate a ratio between the actual power consumption and the theoretical power consumption;

Step a2, if the ratio of the actual power consumption to the theoretical power consumption exceeds a preset threshold, determine that the actual power consumption is abnormal;

In step a3, if the ratio between the actual power consumption and the theoretical power consumption does not exceed a preset threshold, it is determined that the actual power consumption is normal.

In this embodiment, the server acquires the actual power consumption corresponding to the time node information, and calculates a ratio between the actual power consumption and the theoretical power consumption; the actual power consumption and the theoretical power consumption If the ratio of power exceeds a preset threshold, the server determines that the actual power consumption is abnormal; when the ratio of the actual power to the theoretical power does not exceed a preset threshold, the server determines the actual power The quantity is normal; the calculation of the abnormality determination in this embodiment is relatively simple, and the obtained abnormality determination result is intuitive and accurate.

In step S40, if the actual power consumption is abnormal, a power consumption abnormality prompt message is generated.

After the server obtains a result of abnormal actual power consumption according to the foregoing power consumption abnormality evaluation step, the server generates a power consumption abnormality prompt message. The prompt information of the abnormal power consumption is sent to the terminal, so that the end user can view the prompt information of the abnormal power consumption.

In this embodiment, a power consumption calculation regression model is established in advance. When the server receives the power consumption abnormality evaluation request, the current actual power consumption value of the power consumption abnormality evaluation request is obtained, and a theoretical power consumption value is calculated based on the preset regression model. Then, the actual power consumption value is compared with the theoretical power consumption value to determine whether the power consumption is abnormal. In this embodiment, the evaluation of the power consumption abnormality does not rely on historical data, which can effectively exclude the historical power consumption abnormality value from evaluating the power abnormality Impact, improve the accuracy of abnormality assessment, at the same time, the server uses a preset regression model to perform power abnormality assessment, making the abnormality assessment steps for power consumption simpler and more efficient.

Referring to FIG. 3, this embodiment of the method for evaluating abnormality in power consumption of this application is proposed on the basis of the first embodiment of this application. This embodiment is a refinement of step S40 in the first embodiment. When the power consumption is abnormal, the corresponding historical information is acquired to determine the specific situation of the abnormal power consumption.

The method for evaluating abnormal power consumption includes:

In step S41, if the actual power consumption is abnormal, the historical synchronization power consumption corresponding to the time node information is obtained.

If the server determines that the actual power consumption is abnormal, the server obtains the historical power consumption corresponding to the time node information. For example, the time node corresponding to the actual power consumption abnormality is: at 13:00 on May 2, 2018, the server obtains 2017 The power consumption on the afternoon of May 2, 2013 is the historical power consumption corresponding to the node information at that time.

Step S42: Calculate the actual power consumption and the historical power consumption according to a preset synchronization formula to obtain a year-on-year growth rate.

The server calculates the actual power consumption and historical power consumption according to a preset period formula to obtain the year-on-year growth rate. The preset period formula is a preset formula for calculating the year-on-year growth rate. Synchronous number) ÷ Synchronous number * 100%, that is, the server calculates the actual power consumption and the historical power consumption according to a preset synchronization formula to obtain a year-on-year growth rate.

Step S43: Generate power consumption abnormality prompt information based on the year-on-year growth rate and the actual power consumption for users to view.

The server generates the power consumption abnormality prompt information based on the calculated year-on-year growth rate and actual power consumption. For example, the server inputs the year-on-year growth rate and actual power consumption into the abnormality prompt template, and generates power consumption abnormality prompt information for users. Check it out.

In this embodiment, the user can check the abnormal power consumption according to the prompt information, reduce the occurrence of abnormal power consumption, and avoid the adverse effects caused by power failure.

Further, in the third embodiment of the method for evaluating abnormality of power consumption in this application, this embodiment is proposed on the basis of the first embodiment. In this embodiment, if there is no abnormality in power consumption, the power consumption is Anomaly assessment methods include:

Step S50: if the actual power consumption is normal, use the time node information and the power consumption characteristic information related to the time node information as an updated power sample.

If the server determines that the actual power consumption is normal, the server obtains time node information corresponding to the actual power consumption, and obtains power consumption characteristic information related to the time node information as an updated power sample, and updates the first implementation according to the updated power sample The preset regression model in the example.

Step S60: Save the updated power sample to the preset power sample set to obtain an updated power sample set.

The server saves the updated power sample to the preset power sample set to obtain the updated power sample set. After the updated power sample is added to the preset power sample set, the server adds tag information to the updated power sample to obtain the updated power sample. .

Step S70: When receiving a preset regression model update request, obtain the updated power sample in the updated power sample set.

The user triggers an update request based on a preset regression model, and when the server receives the preset regression model update request, obtains an updated power sample with tag information in the updated power sample set.

Step S80: Adjust each regression sub-model in the preset regression model based on the updated power sample to obtain an updated regression sub-model, package each of the updated regression sub-models, and generate an updated regression model.

The server obtains the power consumption characteristic information related to the time node information corresponding to the updated power sample, and adjusts the relevant parameters in each regression sub-model in the preset regression model according to the power consumption characteristic information related to the obtained time node information. An update regression sub-model, which encapsulates each of the update regression sub-models to generate an update regression model.

In this embodiment, the preset regression model may be updated, so that the preset regression model has real-time performance, so as to ensure the accuracy of the abnormal evaluation of power consumption.

In addition, referring to FIG. 4, an embodiment of the present application further provides an embodiment of an abnormality evaluation device for power consumption. The abnormality evaluation device for power consumption includes:

The receiving and acquiring module 10 is configured to receive a power consumption abnormality evaluation request, acquire time node information in the power consumption abnormality evaluation request, and acquire power consumption characteristic information related to the time node information;

An input calculation module 20, configured to input the power consumption characteristic information and the time node information into a preset regression model to obtain a theoretical power consumption corresponding to the time node information;

The obtaining and judging module 30 is configured to obtain the actual power consumption corresponding to the time node information, and compare the actual power consumption with the theoretical power consumption to determine whether the actual power consumption is abnormal;

The generating prompting module 40 is configured to generate prompting information of abnormal power consumption if the actual power consumption is abnormal.

Optionally, the apparatus for evaluating abnormality in power consumption includes:

A sample classification module, configured to obtain a power sample from a preset power sample set, and classify each of the power samples according to a preset classification rule to obtain n subsets of power samples;

The following steps are performed for each of the power sample subsets:

A sub-model generating module, configured to use the power sample subset as a target power sample subset, generate an initial regression model based on the target power sample subset, and obtain the n power sample subsets to remove the target power sample subset Other power sample subsets of the set, using the other power sample subsets to iteratively train the initial regression model to obtain a regression submodel corresponding to the target power sample subset;

A model packaging module is configured to package the regression sub-model corresponding to each of the target power sample subsets to generate a preset regression model.

Optionally, the input calculation module 20 includes:

An information input unit is configured to input the power consumption characteristic information and the time node information into each of the regression sub-models of the preset regression model to obtain a basic power consumption corresponding to each of the regression sub-models. the amount;

The theoretical value determining unit is configured to sum each of the basic power consumptions and calculate an average value to obtain a theoretical power consumption corresponding to the time node information.

Optionally, the receiving and acquiring module 10 includes:

A collection unit, configured to receive a power consumption abnormality evaluation request, acquire time node information in the power consumption abnormality evaluation request, and collect temperature information and holiday information corresponding to the time node information;

A characteristic data determining unit, configured to use the temperature information and the holiday information as power consumption characteristic information related to the time node information.

Optionally, the acquisition judgment module 30 includes:

An obtaining comparison unit, configured to obtain the actual power consumption corresponding to the time node information, and calculate a ratio between the actual power consumption and the theoretical power consumption;

A first determining unit, configured to determine that the actual power consumption is abnormal if a ratio between the actual power consumption and the theoretical power consumption exceeds a preset threshold;

A second determination unit is configured to determine that the actual power consumption is normal if a ratio between the actual power consumption and the theoretical power consumption does not exceed a preset threshold.

Optionally, the generating prompt module 40 includes:

A data acquisition unit, configured to acquire historical power consumption in the corresponding period corresponding to the time node information if the actual power consumption is abnormal;

A change rate calculation unit, configured to calculate the actual power consumption and the historical power consumption according to a preset synchronization formula to obtain a year-on-year growth rate;

A generating unit is configured to generate power consumption abnormality prompt information based on the year-on-year growth rate and the actual power consumption for a user to view.

Optionally, the apparatus for evaluating abnormality in power consumption further includes:

A sample processing module, configured to use the time node information and the power consumption characteristic information related to the time node information as an updated power sample if the actual power consumption is normal;

A sample saving module, configured to save the updated power sample to the preset power sample set to obtain an updated power sample set;

A receiving update module, configured to obtain the updated power sample in the updated power sample set when a preset regression model update request is received;

A model update module is configured to adjust each regression sub-model in the preset regression model based on the updated power sample, obtain an updated regression sub-model, and package each of the updated regression sub-models to generate an updated regression model.

For the steps implemented by the functional modules of the power consumption abnormality evaluation device, reference may be made to various embodiments of the power consumption abnormality evaluation method of the present application, and details are not described herein again.

In addition, an embodiment of the present application also provides a computer storage medium.

Computer-readable instructions are stored on the computer storage medium, and when the computer-readable instructions are executed by a processor, the operations in the method for evaluating abnormality in power consumption provided by the foregoing embodiments are implemented.

It should be noted that in this article, relational terms such as first and second are used only to distinguish one entity / operation / object from another entity / operation / object, and do not necessarily require or imply these. There is any such actual relationship or order between entities / operations / objects; the terms "including", "including" or any other variation thereof are intended to cover non-exclusive inclusion, thereby making the process, method, An article or system includes not only those elements, but also other elements not explicitly listed, or elements inherent to such a process, method, article, or system. Without more restrictions, an element limited by the sentence "including a ..." does not exclude the existence of other identical elements in the process, method, article, or system that includes the element.

The above-mentioned serial numbers of the embodiments of the present application are merely for description, and do not represent the superiority or inferiority of the embodiments.

The above are only preferred embodiments of the present application, and thus do not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made by using the contents of the specification and drawings of the present application, or directly or indirectly used in other related technical fields Are included in the scope of patent protection of this application.

Claims

A method for evaluating abnormality in power consumption, characterized in that the method for evaluating abnormality in power consumption includes the following steps:

Receiving a power consumption abnormality evaluation request, acquiring time node information in the power consumption abnormality evaluation request, and acquiring power consumption characteristic information related to the time node information;

Inputting the power consumption characteristic information and the time node information into a preset regression model to obtain a theoretical power consumption corresponding to the time node information;

Acquiring the actual power consumption corresponding to the time node information, and comparing the actual power consumption with the theoretical power consumption to determine whether the actual power consumption is abnormal;

If the actual power consumption is abnormal, the power consumption abnormality prompt information is generated.
The method for evaluating power consumption abnormality according to claim 1, wherein the receiving the power consumption abnormality evaluation request, obtaining time node information in the power consumption abnormality evaluation request, and acquiring the time node information Prior to the steps related to power usage information, include:

Obtaining power samples from a preset power sample set, classifying each of the power samples according to a preset classification rule, and obtaining n power sample subsets;

The following steps are performed for each of the power sample subsets:

Using the power sample subset as a target power sample subset, generating an initial regression model based on the target power sample subset, and obtaining other power sample subsets in the n power sample subsets excluding the target power sample subset Using the other power sample subsets to iteratively train the initial regression model to obtain a regression submodel corresponding to the target power sample subset;

Encapsulating the regression sub-model corresponding to each of the target power sample subsets to generate a preset regression model.
The method for evaluating abnormal power consumption according to claim 2, wherein the power consumption characteristic information and the time node information are input into a preset regression model to obtain a theory corresponding to the time node information Steps to use electricity, including:

Inputting the power consumption characteristic information and the time node information into each of the regression sub-models of the preset regression model to obtain a basic power consumption corresponding to each of the regression sub-models;

Sum the respective basic power consumptions and calculate an average value to obtain a theoretical power consumption corresponding to the time node information.
The method for evaluating abnormality of power consumption according to claim 2, wherein the actual power consumption corresponding to the time node information is obtained, and the actual power consumption and the theoretical power consumption are compared After the step of determining whether the actual power consumption is abnormal, the method includes:

If the actual power consumption is normal, using the time node information and power characteristic information related to the time node information as an updated power sample;

Saving the updated power sample to the preset power sample set to obtain an updated power sample set;

When receiving a preset regression model update request, obtaining the updated power sample in the updated power sample set;

Each regression sub-model in the preset regression model is adjusted based on the updated power sample to obtain an updated regression sub-model, and each of the updated regression sub-models is packaged to generate an updated regression model.
The method for evaluating power consumption abnormality according to claim 1, wherein the receiving the power consumption abnormality evaluation request, obtaining time node information in the power consumption abnormality evaluation request, and acquiring the time node information The steps of related power consumption characteristic information include:

Receiving a power consumption abnormality evaluation request, acquiring time node information in the power consumption abnormality evaluation request, and collecting temperature information and holiday information corresponding to the time node information;

Use the temperature information and the holiday information as the power consumption characteristic information related to the time node information.
The method for evaluating abnormal power consumption according to claim 1, wherein the actual power consumption corresponding to the time node information is obtained, and the actual power consumption and the theoretical power consumption are compared The steps to determine whether the actual power consumption is abnormal include:

Acquiring the actual power consumption corresponding to the time node information, and calculating a ratio between the actual power consumption and the theoretical power consumption;

If the ratio between the actual power consumption and the theoretical power consumption exceeds a preset threshold, determining that the actual power consumption is abnormal;

If the ratio between the actual power consumption and the theoretical power consumption does not exceed a preset threshold, it is determined that the actual power consumption is normal.
The method for evaluating abnormality of power consumption according to claim 1, wherein the step of generating the prompt information of abnormality of the power consumption if the actual power consumption is abnormal, comprises:

If the actual power consumption is abnormal, obtaining historical power consumption corresponding to the time node information;

Calculating the actual power consumption and the historical power consumption according to a preset synchronization formula to obtain a year-on-year growth rate;

Generate power consumption abnormality prompt information based on the year-on-year growth rate and the actual power consumption for users to view.
A power consumption abnormality evaluation device, characterized in that the power consumption abnormality evaluation device includes:

A receiving and acquiring module, configured to receive a request for abnormal evaluation of power consumption, acquire time node information in the request for abnormal evaluation of power consumption, and obtain power consumption characteristic information related to the time node information;

An input calculation module, configured to input the power consumption characteristic information and the time node information into a preset regression model to obtain a theoretical power consumption corresponding to the time node information;

An acquisition judgment module, configured to acquire the actual power consumption corresponding to the time node information, and compare the actual power consumption with the theoretical power consumption to determine whether the actual power consumption is abnormal;

A prompting module is configured to generate prompting information of abnormal power consumption if the actual power consumption is abnormal.
The apparatus for evaluating abnormality in power consumption according to claim 8, further comprising:

A sample classification module, configured to obtain a power sample from a preset power sample set, and classify each of the power samples according to a preset classification rule to obtain n subsets of power samples;

A sub-model generating module is configured to perform the following steps for each of the power sample subsets: use the power sample subset as a target power sample subset, generate an initial regression model based on the target power sample subset, and obtain the Remove the other power sample subsets of the target power sample subset from the n power sample subsets, and use the other power sample subsets to iteratively train the initial regression model to obtain the regression corresponding to the target power sample subset. Sub-model

A model packaging module is configured to package the regression sub-model corresponding to each of the target power sample subsets to generate a preset regression model.
The apparatus according to claim 9, wherein the input calculation module comprises:

An information input unit is configured to input the power consumption characteristic information and the time node information into each of the regression sub-models of the preset regression model to obtain a basic power consumption corresponding to each of the regression sub-models. the amount;

The theoretical value determining unit is configured to sum each of the basic power consumptions and calculate an average value to obtain a theoretical power consumption corresponding to the time node information.
The apparatus for evaluating abnormality in power consumption according to claim 9, wherein the apparatus for evaluating abnormality in power consumption further comprises:

A sample processing module, configured to use the time node information and the power consumption characteristic information related to the time node information as an updated power sample if the actual power consumption is normal;

A sample saving module, configured to save the updated power sample to the preset power sample set to obtain an updated power sample set;

A receiving update module, configured to obtain the updated power sample in the updated power sample set when a preset regression model update request is received;

A model update module is configured to adjust each regression sub-model in the preset regression model based on the updated power sample, obtain an updated regression sub-model, and package each of the updated regression sub-models to generate an updated regression model.
The apparatus for evaluating abnormal power consumption according to claim 8, wherein the receiving and acquiring module comprises:

A collection unit, configured to receive a power consumption abnormality evaluation request, acquire time node information in the power consumption abnormality evaluation request, and collect temperature information and holiday information corresponding to the time node information;

A characteristic data determining unit, configured to use the temperature information and the holiday information as power consumption characteristic information related to the time node information.
The apparatus for evaluating abnormality of power consumption according to claim 8, wherein the acquisition judgment module comprises:

An obtaining comparison unit, configured to obtain the actual power consumption corresponding to the time node information, and calculate a ratio between the actual power consumption and the theoretical power consumption;

A first determining unit, configured to determine that the actual power consumption is abnormal if a ratio between the actual power consumption and the theoretical power consumption exceeds a preset threshold;

A second determination unit is configured to determine that the actual power consumption is normal if a ratio between the actual power consumption and the theoretical power consumption does not exceed a preset threshold.
The apparatus for evaluating abnormality of power consumption according to claim 8, wherein the generating prompt module comprises:

A data acquisition unit, configured to acquire historical power consumption in the corresponding period corresponding to the time node information if the actual power consumption is abnormal;

A change rate calculation unit, configured to calculate the actual power consumption and the historical power consumption according to a preset synchronization formula to obtain a year-on-year growth rate;

A generating unit is configured to generate power consumption abnormality prompt information based on the year-on-year growth rate and the actual power consumption for a user to view.
A power consumption abnormality evaluation device, characterized in that the power consumption abnormality evaluation device includes a memory, a processor, and computer-readable instructions stored on the memory and executable on the processor, wherein : The computer-readable instructions, when executed by the processor, implement the following steps:

Receiving a power consumption abnormality evaluation request, acquiring time node information in the power consumption abnormality evaluation request, and acquiring power consumption characteristic information related to the time node information;

Inputting the power consumption characteristic information and the time node information into a preset regression model to obtain a theoretical power consumption corresponding to the time node information;

Acquiring the actual power consumption corresponding to the time node information, and comparing the actual power consumption with the theoretical power consumption to determine whether the actual power consumption is abnormal;

If the actual power consumption is abnormal, the power consumption abnormality prompt information is generated.
The apparatus for evaluating abnormality in power consumption according to claim 15, wherein when the computer-readable instructions are executed by the processor, the following steps are further implemented:

Obtaining power samples from a preset power sample set, classifying each of the power samples according to a preset classification rule, and obtaining n power sample subsets;

The following steps are performed for each of the power sample subsets:

Using the power sample subset as a target power sample subset, generating an initial regression model based on the target power sample subset, and obtaining other power sample subsets in the n power sample subsets excluding the target power sample subset Using the other power sample subsets to iteratively train the initial regression model to obtain a regression submodel corresponding to the target power sample subset;

Encapsulating the regression sub-model corresponding to each of the target power sample subsets to generate a preset regression model.
The power consumption abnormality evaluation device according to claim 16, wherein the power consumption characteristic information and the time node information are input into a preset regression model to obtain a theory corresponding to the time node information Steps to use electricity, including:

Inputting the power consumption characteristic information and the time node information into each of the regression sub-models of the preset regression model to obtain a basic power consumption corresponding to each of the regression sub-models;

Sum the respective basic power consumptions and calculate an average value to obtain a theoretical power consumption corresponding to the time node information.
The power consumption abnormality evaluation device according to claim 15, wherein the receiving the power consumption abnormality evaluation request, obtaining time node information in the power consumption abnormality evaluation request, and acquiring the time node information The steps of related power consumption characteristic information include:

Receiving a power consumption abnormality evaluation request, acquiring time node information in the power consumption abnormality evaluation request, and collecting temperature information and holiday information corresponding to the time node information;

Use the temperature information and the holiday information as the power consumption characteristic information related to the time node information.
The apparatus for evaluating abnormality of power consumption according to claim 15, wherein said acquiring the actual power consumption corresponding to said time node information, and comparing said actual power consumption with said theoretical power consumption The steps to determine whether the actual power consumption is abnormal include:

Acquiring the actual power consumption corresponding to the time node information, and calculating a ratio between the actual power consumption and the theoretical power consumption;

If the ratio between the actual power consumption and the theoretical power consumption exceeds a preset threshold, determining that the actual power consumption is abnormal;

If the ratio between the actual power consumption and the theoretical power consumption does not exceed a preset threshold, it is determined that the actual power consumption is normal.
A computer storage medium is characterized in that computer readable instructions are stored on the computer storage medium, and when the computer readable instructions are executed by a processor, the following steps are implemented:

Receiving a power consumption abnormality evaluation request, acquiring time node information in the power consumption abnormality evaluation request, and acquiring power consumption characteristic information related to the time node information;

Inputting the power consumption characteristic information and the time node information into a preset regression model to obtain a theoretical power consumption corresponding to the time node information;

Acquiring the actual power consumption corresponding to the time node information, and comparing the actual power consumption with the theoretical power consumption to determine whether the actual power consumption is abnormal;

If the actual power consumption is abnormal, the power consumption abnormality prompt information is generated.