Detection data generation method, electric equipment energy consumption processing method and device

By acquiring and reconstructing sensor detection data in the smart park, generating target detection data and performing energy consumption prediction, the problem of energy consumption management lag in the smart park is solved, and more accurate energy consumption management and equipment abnormal identification is achieved.

CN119669660APending Publication Date: 2025-03-21BEIJING JINGDONG YUANSHENG TECH CO LTD +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202311233375.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2023-09-21
Publication Date
2025-03-21

AI Technical Summary

Technical Problem

The energy consumption management of smart devices in smart parks relies on manual management, resulting in lag in abnormal detection and optimization.

Method used

By acquiring multiple sensor detection data that matches the target area and the target scenario, identifying and reconstructing missing data, generating target detection data, and inputting it into the energy consumption prediction model to identify abnormalities.

Benefits of technology

Improve the accuracy of sensor detection data, identify equipment abnormalities in advance, optimize energy consumption management, and reduce management and maintenance costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119669660A_ABST
    Figure CN119669660A_ABST
Patent Text Reader

Abstract

The invention provides a detection data generation method and an electric equipment energy consumption processing method and device, which can be applied to the technical fields of intelligent equipment, intelligent industry, artificial intelligence and Internet of Things. The detection data generation method comprises the steps that multiple pieces of sensor detection data matched with a target area and a target scene are obtained, the target area represents a functional area in a building object, the target scene represents a use scene of electric equipment, and the sensor detection data comprise detection data collected by at least one sensor; in response to detection that missing data exists in the multiple pieces of sensor detection data, reconstructing the missing data based on the data missing category to obtain sensor reconstruction data corresponding to the missing data; and generating target detection data according to complete data in the sensor reconstruction data and the sensor detection data.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present disclosure relates to the fields of intelligent equipment, smart industry, artificial intelligence, and the Internet of Things, and more specifically, to a detection data generation method, an electrical equipment energy consumption processing method, and a device. Background Art

[0002] With the rapid development of smart equipment, smart industry, artificial intelligence, and the Internet of Things, smart appliances are increasingly being used in daily life, such as smart air conditioners, smart refrigerators, and smart lighting equipment. In large-scale smart parks, a large number of complex types of smart equipment are generally deployed in multiple locations in the park.

[0003] In the process of realizing the concept disclosed in the present invention, the inventors found that there are at least the following problems in the related technology: the energy consumption management of smart equipment in the smart park relies on manual management, and there is a lag in discovering anomalies, handling anomalies, and optimizing equipment energy consumption through manual inspections. Summary of the invention

[0004] In view of this, the present disclosure provides a detection data generating method, an electrical equipment energy consumption processing method and a device.

[0005] One aspect of the present disclosure provides a method for generating detection data, comprising:

[0006] Acquire multiple sensor detection data matching the target area and the target scene, wherein the target area represents the functional area in the building object, the target scene represents the usage scene of the electrical equipment, and the sensor detection data includes detection data collected by at least one sensor;

[0007] In response to detecting that missing data exists in the plurality of sensor detection data, reconstructing the missing data based on the data missing category to obtain sensor reconstructed data corresponding to the missing data;

[0008] The target detection data is generated according to the complete data in the sensor reconstruction data and the sensor detection data.

[0009] According to an embodiment of the present disclosure, the plurality of sensor detection data includes detection data of M sensor types, where M is greater than or equal to 1; based on the data missing category, the missing data is reconstructed to obtain sensor reconstructed data corresponding to the missing data, including:

[0010] Based on the data missing category, the missing data is reconstructed using detection data of the same type as the sensor that collected the missing data to obtain sensor reconstructed data.

[0011] According to an embodiment of the present disclosure, the data missing categories include point missing and block missing, the point missing represents data missing at one moment, and the block missing represents data missing at T consecutive moments, where T is greater than or equal to 2; based on the data missing categories, the missing data is reconstructed, and the sensor reconstructed data corresponding to the missing data is obtained, including:

[0012] In response to determining that the data missing category is point missing, reconstructing the missing data using the first detection data to which the missing data belongs by linear interpolation to obtain sensor reconstructed data;

[0013] In response to determining that the data missing category is block missing, sensor reconstruction data is determined according to the first detection data and K second detection data of the same type as that of a sensor used to collect the first detection data, where K is greater than or equal to 1.

[0014] According to an embodiment of the present disclosure, determining the sensor reconstruction data according to the first detection data and K second detection data of the same type as that of the sensor that collected the first detection data includes:

[0015] Dividing the first detection data into a forward data set and a backward data set according to positions of the missing data in the first detection data; and

[0016] The forward data set, the backward data set and K second detection data are input into a bidirectional recurrent neural network, and the sensor reconstructed data corresponding to the missing data is output.

[0017] According to an embodiment of the present disclosure, the forward data set, the backward data set and K second detection data are input into a bidirectional recurrent neural network, and the sensor reconstruction data corresponding to the missing data is output, including:

[0018] Repeat the following steps until the sensor reconstruction data at T time points is obtained:

[0019] Determine a forward prediction value corresponding to a t-th round forward data set and a backward prediction value corresponding to a t-th round backward data set, wherein the t-th round forward data set includes the forward data set and sensor reconstructed data at t consecutive moments obtained after t rounds of processing, t is greater than or equal to 1, and t is less than or equal to T;

[0020] Based on the attention mechanism, the forward prediction value and the backward prediction value are processed to obtain the prediction value at the t+1th moment;

[0021] The predicted value at the t+1th moment and the fused data at the t+1th moment are fused to obtain the sensor reconstruction data at the t+1th moment, wherein the fused data is determined based on the K second detection data.

[0022] According to an embodiment of the present disclosure, the method further includes:

[0023] Determine first historical detection data of a first sensor that collects first detection data, and second historical detection data of a second sensor that collects each second detection data;

[0024] Calculating the similarity between the first historical detection data and each second historical detection data at the missing position; and

[0025] The second detection data corresponding to the second sensor with the highest similarity is determined as the fusion data.

[0026] According to an embodiment of the present disclosure, the method further includes: determining position information of a first sensor that collects first detection data;

[0027] Determine, according to the location information, K second sensors which are closest to the first sensor and have the same sensor type as the first sensor; and

[0028] The second detection data collected by each of the K second sensors is acquired to obtain K second detection data.

[0029] According to an embodiment of the present disclosure, acquiring a plurality of sensor detection data matching a target area and a target scene includes:

[0030] Determine at least one electrical device in a target area and a target scene;

[0031] Determining a sensor detection scheme corresponding to a target area and a target scene according to a sensor type and a sensor quantity of at least one electrical device; and

[0032] Based on the sensor detection scheme, multiple sensor detection data are obtained.

[0033] According to an embodiment of the present disclosure, in response to detecting that there is missing data in the plurality of sensor detection data, before reconstructing the missing data based on the data missing category to obtain sensor reconstructed data corresponding to the missing data, the method includes:

[0034] Dividing the detection data of the plurality of sensors according to a predetermined collection interval;

[0035] In response to detecting that the data lengths of the divided sensor detection data are different, filling the missing data with zeros to update the sensor detection data;

[0036] In response to detecting that the data lengths of the divided sensor detection data are the same, the same time index is added to the divided sensor detection data, and the sensor detection data is updated.

[0037] According to an embodiment of the present disclosure, in response to detecting that there is missing data in the plurality of sensor detection data, before reconstructing the missing data based on the data missing category to obtain sensor reconstructed data corresponding to the missing data, the method further includes:

[0038] Determine an abnormal moment in the sensor detection data according to detection data at two adjacent moments in the sensor detection data and an abnormality detection threshold; and

[0039] Clear the detection data at the abnormal time and update the sensor detection data.

[0040] Another aspect of the present disclosure provides a method for processing energy consumption of an electric device, comprising:

[0041] Acquire target detection data corresponding to the target area and the target scene, wherein the target detection data is determined according to the detection data generation method;

[0042] Determining third detection data for the target electrical device from the target detection data;

[0043] The third detection data is input into an energy consumption prediction model for the target electrical equipment, and target prediction data is output, wherein the target prediction data includes energy consumption prediction values ​​at S consecutive moments after the third detection data, S is greater than or equal to 1, and the energy consumption prediction model includes a recurrent neural network.

[0044] According to an embodiment of the present disclosure, it also includes:

[0045] Anomaly recognition is performed on the target prediction data to obtain anomaly recognition results.

[0046] According to an embodiment of the present disclosure, the target prediction data is subjected to abnormality identification to obtain an abnormality identification result, which includes:

[0047] Identifying energy consumption patterns that match target forecast data; and

[0048] Anomalies are identified on target prediction data based on the energy consumption pattern to obtain anomaly identification results.

[0049] According to an embodiment of the present disclosure, the target prediction data is subjected to abnormality identification based on the energy consumption mode, and the abnormality identification result is obtained, including:

[0050] Acquire third historical detection data of the target electrical device in the energy consumption mode for S' moments, wherein the S' moment and the S moment belong to the same time period of different periods, and S' is greater than or equal to S; and

[0051] An abnormality recognition result is determined based on the similarity between the third historical detection data and the target prediction data.

[0052] According to an embodiment of the present disclosure, determining the abnormality recognition result according to the similarity between the third historical detection data and the target prediction data includes:

[0053] Calculating the distance between the third historical detection data and the target prediction data using a dynamic time warping algorithm; and

[0054] The anomaly recognition result is determined based on the comparison relationship between the distance and the trend threshold.

[0055] According to an embodiment of the present disclosure, it also includes:

[0056] In response to detecting that missing data exists in the third detection data, issuing an alarm based on the first alarm mode;

[0057] In response to the abnormality identification result indicating an abnormality, an alarm is issued based on the second alarm mode.

[0058] According to an embodiment of the present disclosure, the method further includes: obtaining a utility interval for a target electrical device according to a target area and a target scene, wherein the utility interval represents an operating parameter interval of an optimal energy consumption state of the electrical device; and

[0059] An adjustment strategy for the target electrical equipment is generated according to the utility interval and the third detection data.

[0060] Another aspect of the present disclosure provides a detection data generating device, comprising:

[0061] A first acquisition module is used to acquire multiple sensor detection data matching a target area and a target scene, wherein the target area represents a functional area within a building object, the target scene represents a usage scene of an electrical device, and the sensor detection data includes detection data collected by at least one sensor;

[0062] A reconstruction module, configured to, in response to detecting that missing data exists in the detection data of the multiple sensors, reconstruct the missing data based on the data missing category to obtain sensor reconstructed data corresponding to the missing data;

[0063] The generation module is used to generate target detection data according to the complete data in the sensor reconstruction data and the sensor detection data.

[0064] Another aspect of the present disclosure provides an apparatus for processing energy consumption of an electric device, comprising:

[0065] A second acquisition module, used to acquire target detection data corresponding to the target area and the target scene, wherein the target detection data is determined according to any one of claims 1 to 10;

[0066] A determination module, configured to determine third detection data for a target electrical device from the target detection data;

[0067] A prediction module is used to input the third detection data into an energy consumption prediction model for a target electrical device and output target prediction data, wherein the target prediction data includes energy consumption prediction values ​​at S consecutive moments after the third detection data, S is greater than or equal to 1, and the energy consumption prediction model includes a recurrent neural network.

[0068] Another aspect of the present disclosure provides an electronic device, comprising: one or more processors; a memory for storing one or more programs, wherein when the one or more programs are executed by one or more processors, the one or more processors implement the above-mentioned detection data generation method and electrical equipment energy consumption processing method.

[0069] Another aspect of the present disclosure provides a computer-readable storage medium storing computer-executable instructions, which are used to implement the above method when executed.

[0070] Another aspect of the present disclosure provides a computer program product, which includes computer executable instructions, and the instructions are used to implement the above method when executed.

[0071] The embodiments of the present disclosure can obtain sensor detection data in a targeted manner based on the target area and the target usage scenario of the electrical equipment by acquiring multiple sensor detection data that match the target area and the target scenario, so as to process the sensor detection data of the electrical equipment with different operating rules. In response to detecting the presence of missing data in multiple sensor detection data, the missing data is reconstructed based on the data missing category to obtain sensor reconstruction data corresponding to the missing data; target detection data is generated based on the sensor reconstruction data and the complete data in the sensor detection data, and abnormal values ​​and missing values ​​in the sensor detection data can be reconstructed to improve the accuracy of the sensor detection data and more accurately obtain the actual operating conditions of the electrical equipment with different operating rules. In addition, the above embodiments are also helpful to improve the efficiency of abnormal detection of electrical equipment in smart parks and reduce abnormal detection costs and equipment maintenance costs based on more accurate sensor detection data. BRIEF DESCRIPTION OF THE DRAWINGS

[0072] The above and other objects, features and advantages of the present disclosure will become more apparent through the following description of the embodiments of the present disclosure with reference to the accompanying drawings, in which:

[0073] Figure 1 The application scenarios of the detection data generation method and the power consumption processing method of the present disclosure are schematically shown;

[0074] Figure 2 A flow chart of a method for generating detection data according to an embodiment of the present disclosure is schematically shown;

[0075] Figure 3 Schematically shows a flow chart for determining sensor reconstruction data according to an embodiment of the present disclosure;

[0076] Figure 4 A flowchart of determining sensor reconstruction data according to a forward data set, a backward data set and K second detection data according to an embodiment of the present disclosure is schematically shown;

[0077] Figure 5 A flowchart schematically shows a method for processing energy consumption of an electric device according to another embodiment of the present disclosure;

[0078] Figure 6 A flowchart schematically shows a method for calculating the distance between third historical detection data and target prediction data according to another embodiment of the present disclosure;

[0079] Figure 7 A flowchart of a method for generating an adjustment strategy for an electric device according to another embodiment of the present disclosure is schematically shown;

[0080] Figure 8 The energy consumption management system for electrical equipment in a smart park according to an embodiment of the present disclosure is schematically shown;

[0081] Fig. 9 A flowchart of a method for generating detection data of electrical equipment in a smart park and processing energy consumption according to a specific embodiment of the present disclosure is schematically shown;

[0082] Fig.10 The structure block diagram of the detection data generating method according to the embodiment of the present disclosure is schematically shown;

[0083] Fig.11 A structural block diagram schematically shows a method for processing energy consumption of an electric device according to an embodiment of the present disclosure; and

[0084] Fig.12 A block diagram of an electronic device suitable for implementing a detection data generating method and an electrical equipment energy consumption processing method according to an embodiment of the present disclosure is schematically shown. DETAILED DESCRIPTION

[0085] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present disclosure. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present disclosure. However, it is apparent that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present disclosure.

[0086] The terms used herein are only for describing specific embodiments and are not intended to limit the present disclosure. The terms "comprise", "include", etc. used herein indicate the existence of the features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0087] All terms (including technical and scientific terms) used herein have the meanings commonly understood by those skilled in the art, unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification, and should not be interpreted in an idealized or overly rigid manner.

[0088] When using expressions such as "at least one of A, B, and C, etc.", they should generally be interpreted according to the meaning of the expression commonly understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include but is not limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.).

[0089] In the embodiments of the present disclosure, the collection, updating, analysis, processing, use, transmission, provision, disclosure, storage, etc. of the data involved (for example, including but not limited to user personal information) are in compliance with the provisions of relevant laws and regulations, are used for legitimate purposes, and do not violate public order and good morals. In particular, necessary measures are taken for user personal information to prevent illegal access to user personal information data and maintain the security of user personal information, network security, and national security.

[0090] In the embodiments of the present disclosure, the user's authorization or consent is obtained before obtaining or collecting the user's personal information.

[0091] Smart parks are generally a collection of multiple types of buildings and multiple smart devices, such as industrial parks, schools, and communities. Smart parks generally include multiple functional areas, such as office areas, dormitory areas, production areas, and learning areas. In actual applications, the activities of people in different functional areas are different, and the operation rules of smart devices are also different. For example, the power consumption in the dormitory area is at its peak in the evening and at its low peak during the day; the production area equipped with automated industrial robots is at its peak all day long; the entertainment area is only at its peak at noon or in the evening, etc.

[0092] The existing smart park management system still uses traditional manual patrol and manual management methods. For example, if abnormal energy consumption of equipment is discovered through manual patrol, lighting equipment is turned on and off at different times, and the air conditioning is adjusted for heating and cooling. For large-scale industrial smart parks, since a variety of electrical equipment has its own specific usage preferences and operating rules, the use of a unified manual management method is inefficient, costly, and unable to accurately detect the actual operating conditions of electrical equipment. In addition, a unified manual management method cannot identify abnormalities in advance and maintain abnormal equipment.

[0093] The embodiments of the present disclosure provide a detection data generation method, an electrical equipment energy consumption processing method, and an apparatus. The detection data generation method includes acquiring multiple sensor detection data that match a target area and a target scene, wherein the target area represents a functional area within a building object, the target scene represents a usage scene of the electrical equipment, and the sensor detection data includes detection data collected by at least one sensor; in response to detecting that there is missing data in the multiple sensor detection data, based on the data missing category, the missing data is reconstructed to obtain sensor reconstruction data corresponding to the missing data; based on the sensor reconstruction data and the complete data in the sensor detection data, target detection data is generated, which can improve the accuracy of the sensor detection data, more accurately detect the actual operation of the electrical equipment, and help improve the management efficiency of electrical equipment in smart parks and reduce management costs.

[0094] The method for processing energy consumption of electrical equipment includes: obtaining target detection data corresponding to a target area and a target scene, wherein the target detection data is determined according to the above-mentioned detection data generation method; determining third detection data for the target electrical equipment from the target detection data; inputting the third detection data into an energy consumption prediction model for the target electrical equipment, and outputting target prediction data, wherein the target prediction data includes energy consumption prediction values ​​at S consecutive moments after the third detection data, S being greater than or equal to 1, and being able to identify anomalies before the equipment becomes abnormal and implement equipment operation and maintenance, thereby improving the efficiency of abnormality processing of electrical equipment in smart parks and reducing equipment maintenance costs.

[0095] Figure 1 The application scenario 100 of the detection data generation method and the power consumption processing method of the present disclosure is schematically shown. It should be noted that: Figure 1 What is shown are merely examples of application scenarios to which the embodiments of the present disclosure can be applied, in order to help those skilled in the art understand the technical content of the present disclosure, but it does not mean that the embodiments of the present disclosure cannot be used in other devices, systems, environments or scenarios.

[0096] like Figure 1As shown, the application scenario 100 of this embodiment may include multiple sensors and multiple electrical devices, such as a first sensor 101-1, a second sensor 101-2, a third sensor 101-3, a smart air conditioner 102-1, and a smart lighting device 102-2.

[0097] The network is used as a medium for providing a communication link between the first sensor 101-1, the second sensor 101-2, the third sensor 101-3, the smart air conditioner 102-1, the smart lighting device 102-2 and the server. The network may include various connection types, such as wired and / or wireless communication links, etc.

[0098] The first sensor 101-1, the second sensor 101-2, and the third sensor 101-3 may interact with the server to receive or send messages, etc. For example, the first sensor 101-1, the second sensor 101-2, and the third sensor 101-3 may collect sensor detection data of the environment around the electrical device, such as temperature parameters, humidity parameters, light parameters, heat parameters, etc., and send the collected sensor detection data to the server.

[0099] Electrical equipment may include various smart devices including various automatic lighting systems, air conditioning and heating systems, induction water outlet equipment, etc., such as smart air conditioners 102-1 and smart lighting equipment 102-2.

[0100] The first sensor 101 - 1 , the second sensor 101 - 2 , and the third sensor 101 - 3 can collect and manage the environmental information around the smart air conditioner 102 - 1 and the smart lighting device 102 - 2 in the form of the Internet of Things.

[0101] The server may be a server that provides various services, such as processing, managing, and analyzing the sensor detection data sent by the first sensor 101-1, the second sensor 101-2, and the third sensor 101-3. The server may also control electrical appliances such as the smart air conditioner 102-1 and the smart lighting device 102-2, and feed back control instructions to the electrical appliances such as the smart air conditioner 102-1 and the smart lighting device 102-2.

[0102] It should be noted that the detection data generation method and the power consumption processing method provided in the embodiments of the present disclosure can generally be executed by a server. Accordingly, the detection data generation device and the power consumption processing device provided in the embodiments of the present disclosure can generally be set in a server. The detection data generation method and the power consumption processing method provided in the embodiments of the present disclosure can also be executed by a server or server cluster that is different from the server and can communicate with the first sensor 101-1, the second sensor 101-2, the third sensor 101-3 and / or the smart air conditioner 102-1 and the smart lighting device 102-2. Correspondingly, the detection data generation device and the power consumption processing device provided in the embodiments of the present disclosure can also be set in a server or server cluster that is different from the server and can communicate with the first sensor 101-1, the second sensor 101-2, the third sensor 101-3 and / or the smart air conditioner 102-1 and the smart lighting device 102-2.

[0103] For example, the first sensor 101-1, the second sensor 101-2, and the third sensor 101-3 send the sensor detection data of the entire smart park to the server, and the server receives and stores the sensor detection data. The server obtains multiple sensor detection data that match the target area and the target scene; in response to detecting that there are missing data in the multiple sensor detection data, based on the data missing category, the missing data is reconstructed to obtain sensor reconstruction data corresponding to the missing data; and target detection data is generated based on the sensor reconstruction data and the complete data in the sensor detection data.

[0104] Alternatively, the server may also obtain target detection data corresponding to the target area and target scene; determine third detection data for the target electrical equipment from the target detection data; input the third detection data into an energy consumption prediction model for the target electrical equipment, and output target prediction data, wherein the target prediction data includes energy consumption prediction values ​​at S consecutive moments after the third detection data, S is greater than or equal to 1, and the energy consumption prediction model includes a recurrent neural network.

[0105] It should be understood that Figure 1 The number and type of smart devices and sensors in the embodiment are only for illustration purposes. Any number and type of smart devices and sensors may be used as required.

[0106] Figure 2 The flowchart of the detection data generating method according to the embodiment of the present disclosure is schematically shown.

[0107] like Figure 2 As shown, the method includes operations S210 to S230.

[0108] In operation S210 , a plurality of sensor detection data matching a target area and a target scene is acquired.

[0109] According to an embodiment of the present disclosure, the target area represents a functional area within a building object. The building object may be a large industrial park, a smart community, a smart office area, etc. The functional area may be understood as a plurality of areas divided according to the functions of the building object. For example, the functional area includes at least one of the following: an office area, an entertainment area, a rest area, a dormitory area, a robot work area, etc. According to actual needs, different functional areas may be divided for different building objects, and the functional areas are not specifically limited here.

[0110] According to an embodiment of the present disclosure, the target scenario represents the usage scenario of the electrical device. The usage scenario may include multiple usage factors, such as temperature, humidity, season, weather, regional brightness, holidays, etc. In usage scenarios composed of different usage factors, the power consumption or energy consumption of the same electrical device is different.

[0111] For example, when the indoor temperature is 20℃ and the humidity is 60%, the energy consumption of the air conditioner in cooling mode for one hour is 2kw; when the indoor temperature is 25℃ and the humidity is 60%, the energy consumption of the air conditioner in cooling mode for one hour is 2.5kw. For another example, the energy consumption of air conditioners on holidays is less than that on weekdays.

[0112] According to an embodiment of the present disclosure, both the target scene and the target area can be determined by a selection operation of the user on the visualization page.

[0113] According to an embodiment of the present disclosure, the sensor detection data includes detection data collected by at least one sensor. At least one sensor may be used to collect multiple sensor detection data of at least one electrical device in a target area and a target scene.

[0114] According to an embodiment of the present disclosure, the plurality of sensor detection data includes detection data collected by a plurality of sensors of the same sensor type and detection data collected by sensors of a plurality of sensor types.

[0115] For example, the sensor detection data around the air conditioner can be collected through the temperature sensor and the humidity sensor; the sensor detection data around the lighting device can be collected through the temperature sensor and the photosensitive sensor. In addition, the temperature detection data around the air conditioner and the lighting device can be collected through one or more temperature sensors.

[0116] According to an embodiment of the present disclosure, each sensor detection data includes sensor detection values ​​collected at multiple consecutive moments. The sensor detection data is formed by arranging the sensor detection values ​​according to the collection moments. It can be understood that the sensor detection data is time series data.

[0117] In operation S220 , in response to detecting that missing data exists in the plurality of sensor detection data, the missing data is reconstructed based on the data missing category to obtain sensor reconstructed data corresponding to the missing data.

[0118] According to an embodiment of the present disclosure, due to a collection failure, a transmission failure or a network failure, abnormal data or missing data may appear in the detection data of multiple sensors. Therefore, after acquiring the detection data of multiple sensors, if the missing data is stored, the missing data is reconstructed, and if there is no missing data, the complete data is not reconstructed.

[0119] According to an embodiment of the present disclosure, in response to detecting that there is missing data in multiple sensor detection data, the missing data can be reconstructed based on the data missing category. The data missing category is used to characterize the missing situation of the missing data, for example, the sensor detection value at a certain moment is missing, or the sensor detection values ​​at several consecutive moments are missing.

[0120] According to an embodiment of the present disclosure, multiple data missing categories may correspond to multiple abnormal factors. For multiple data missing categories, multiple data reconstruction operations may be performed to obtain sensor reconstructed data for the data missing categories.

[0121] In operation S230 , target detection data is generated according to the sensor reconstruction data and the complete data in the sensor detection data.

[0122] According to an embodiment of the present disclosure, after reconstructing missing data in a plurality of sensor detection data, the reconstructed sensor reconstruction data is combined with the complete data to obtain target detection data.

[0123] According to an embodiment of the present disclosure, since the sensor detection data may be time series data, part of the sensor detection data may be missing, and the remaining part is complete data.

[0124] For example, for the sensor detection data A{x 1 、x 2 、x 3 …x N}, where x 2 The missing data may be the missing data after missing value processing, wherein the missing value processing includes zero filling operation. 1 、x 2 、x 3 …x N} except x 2 All other data except x are sensor detection values ​​collected at the corresponding time, and the other data are complete data. 2 After reconstruction, we can get the missing data x2 The corresponding sensor reconstructed data x 2 '. will be combined with the missing data x 2 The corresponding sensor reconstruction data and {x 1 、x 3 …x N} combination can obtain target detection data {x 1 、x 2 ', x 3 …x N}.

[0125] The embodiments of the present disclosure can obtain sensor detection data in a targeted manner based on the target area and the target usage scenario of the electrical equipment by acquiring multiple sensor detection data that match the target area and the target scenario, so as to process the sensor detection data of the electrical equipment with different operating rules. In response to detecting the presence of missing data in multiple sensor detection data, the missing data is reconstructed based on the data missing category to obtain sensor reconstruction data corresponding to the missing data; target detection data is generated based on the sensor reconstruction data and the complete data in the sensor detection data, and abnormal values ​​and missing values ​​in the sensor detection data can be reconstructed to improve the accuracy of the sensor detection data and more accurately obtain the actual operating conditions of the electrical equipment with different operating rules. In addition, the above embodiments are also helpful to improve the efficiency of abnormal detection of electrical equipment in smart parks and reduce abnormal detection costs and equipment maintenance costs based on more accurate sensor detection data.

[0126] According to an embodiment of the present disclosure, the plurality of sensor detection data includes detection data of M sensor types, where M is greater than or equal to 1. Based on the data missing category, reconstructing the missing data to obtain sensor reconstructed data corresponding to the missing data includes: based on the data missing category, reconstructing the missing data using detection data of the same type as that of the sensor that collected the missing data to obtain the sensor reconstructed data.

[0127] According to an embodiment of the present disclosure, multiple sensor detection data may be acquired by at least one sensor of at least one sensor type. For example, three sensor detection data for temperature may be acquired by three temperature sensors, and one sensor detection data for humidity may be acquired by one humidity sensor.

[0128] According to an embodiment of the present disclosure, in response to detecting that there is missing data in the detection data of multiple sensors, a corresponding data reconstruction method is determined according to the data missing category to which the missing data belongs. Based on the data reconstruction method, the missing data is reconstructed using the detection data of the same type as the sensor that collected the missing data.

[0129] According to an embodiment of the present disclosure, the detection data of the same type as the sensor that collected the missing data may be the first detection data itself to which the missing data belongs, or may be one or more other sensor detection data collected by other sensors of the same type as the sensor that collected the missing data, such as the second detection data.

[0130] For example, the sensor detection data A{x 1 、x 2 、x 3 …x N}, where x 2 is missing data, and the detection data of the same type as the sensor that collected the missing data can be the complete data {x 1 、x 3 …x N The detection data of the same type as the sensor that collects the missing data can also be the sensor detection data B{y 1 ,y 2 ,y 3 …y N}.

[0131] According to an embodiment of the present disclosure, data missing categories include point missing and block missing, based on which point missing represents data missing at one moment, and block missing represents data missing at T consecutive moments, where T is greater than or equal to 2.

[0132] For example, the sensor detects data A{x 1 、x 2 、x 3 …x N}, where x 2 、x N-1 、x N All are missing data, among which x 2 For point missing, x N-1 、x N The data of two consecutive moments are missing, so x N-1 、x N A building block is missing.

[0133] The embodiments of the present disclosure reconstruct missing data by using detection data of the same sensor type, and can reconstruct detection data of the same dimension in the same target area and the same target scene, thereby improving the accuracy of sensor detection data.

[0134] According to an embodiment of the present disclosure, reconstructing missing data to obtain sensor reconstructed data corresponding to the missing data includes: in response to determining that the data missing category is point missing, reconstructing the missing data using first detection data to which the missing data belongs by linear interpolation to obtain sensor reconstructed data; in response to determining that the data missing category is block missing, determining the sensor reconstructed data based on the first detection data and K second detection data of the same type as the sensor that collected the first detection data, where K is greater than or equal to 1.

[0135] According to an embodiment of the present disclosure, for a missing point, two or more data located before and after the missing point can be obtained from the first detection data to which the missing data belongs, and the sensor reconstruction value of the point missing position can be calculated by a linear interpolation method, and the sensor reconstruction value can be used as the sensor reconstruction data of the missing data.

[0136] According to an embodiment of the present disclosure, for a missing block, K second detection data can be determined according to the sensor type of the first detection data to which the missing data belongs, and the sensor reconstruction data corresponding to the missing data can be determined using the complete data in the first detection data and the K second detection data.

[0137] Figure 3 The flowchart of determining sensor reconstruction data according to an embodiment of the present disclosure is schematically shown.

[0138] like Figure 3 As shown, the first detection data 301 includes the sensor detection values ​​x collected at T consecutive moments. 1 、x 2 、x 3 , …, x p 、x p+1 , …, x p+t 、x p+t+1 , …, x T , the first detection data 301{x 1 、x 2 、x 3 , …, x p 、x p+1 , …, x p+t 、x p+t+1 , …, x T The first detection data 301 includes first missing data 3011 belonging to point missing, second missing data 3012 belonging to block missing and other complete data. The first missing data 3011 can be x 2 , the second missing data 3012 may be {x p+1 , …, x p+t}.

[0139] For the first missing data 3011, the first sensor reconstructed data 3031 corresponding to the first missing data 3011 can be calculated based on the linear interpolation method using the first detection data 301 itself or other complete data in the first detection data 301, such as x 2 '.

[0140] For the second missing data 3012, a plurality of second detection data 302 of the same sensor type as the first detection data 301 can be determined according to the sensor type of the first detection data 301. Then, the second sensor reconstructed data 3032 corresponding to the second missing data 3012 is determined according to the plurality of second detection data 302 and the first detection data 301, such as {x p+1 ', ..., x p+t '}.

[0141] According to the first sensor reconstructed data 3031 , the second sensor reconstructed data 3032 and other complete data in the first detection data 301 , the target detection data 303 corresponding to the first detection data 301 may be obtained.

[0142] According to an embodiment of the present disclosure, each sensor detection data may include only one or more missing data of one data missing type, one or more missing data of two missing types, or may include no missing data.

[0143] In the case of point missing, since the surrounding environment of the electrical equipment will not change suddenly, the embodiment of the present disclosure can directly use a simple linear interpolation method to obtain sensor reconstruction data without performing other reconstruction operations to improve data processing efficiency. In the case of block missing, since the collected values ​​at two or more consecutive moments are missing, the sensor detection data may change significantly compared to the data before the block missing. Therefore, by using the first detection data and K second detection data of the same type as the sensor that collected the first detection data, accurate sensor detection data can be obtained.

[0144] The embodiments of the present disclosure aim at different data missing categories and reconstruct missing data using different data based on different data reconstruction methods, which can improve the accuracy of sensor reconstructed data while improving data processing efficiency.

[0145] According to an embodiment of the present disclosure, determining sensor reconstruction data based on first detection data and K second detection data of the same type as that of a sensor used to collect the first detection data includes: dividing the first detection data into a forward data set and a backward data set according to the position of the missing data in the first detection data; and inputting the forward data set, the backward data set and the K second detection data into a bidirectional recurrent neural network, and outputting sensor reconstruction data corresponding to the missing data.

[0146] According to an embodiment of the present disclosure, the sensor detection data used to detect the target area and the target scene has a specific operating law. For a block missing, according to the forward data set before the block missing, the backward data set after the block missing and the K second detection data, the data of the block missing part can be predicted based on the operating law.

[0147] According to an embodiment of the present disclosure, the forward data set, the backward data set and K second detection data can be respectively input into a bidirectional recurrent neural network as features of three dimensions, and sensor reconstruction data of the missing block part can be obtained based on the prediction of the bidirectional recurrent neural network.

[0148] According to the embodiments of the present disclosure, a bidirectional recurrent neural network (Bi-directional Recurrent Neural Network, Bi-RNN) and its deformations can be used to predict missing data, such as a long short-term memory network (Long Short-Term Memory, LSTM), a gated recurrent neural network (Gated Recurrent Neural Network, GRU), etc. Other neural networks that can perform bidirectional prediction based on time series data can also be used, and the embodiments of the present disclosure are not limited here.

[0149] According to an embodiment of the present disclosure, a forward data set, a backward data set and K second detection data are input into a bidirectional recurrent neural network, and outputting sensor reconstruction data corresponding to the missing data includes: repeatedly executing the following steps until the sensor reconstruction data of T moments are obtained: determining a forward prediction value corresponding to the t-th round forward data set and a backward prediction value corresponding to the t-th round backward data set, wherein the t-th round forward data set includes the forward data set and sensor reconstruction data of t consecutive moments obtained after t rounds of processing, t is greater than or equal to 1, and t is less than or equal to T; based on the attention mechanism, processing the forward prediction value and the backward prediction value to obtain a prediction value at the t+1th moment; fusing the prediction value at the t+1th moment with the fused data at the t+1th moment to obtain the sensor reconstruction data at the t+1th moment, wherein the fused data is determined based on the K second detection data.

[0150] According to an embodiment of the present disclosure, for missing data at T missing moments, sensor reconstruction data at one moment can be obtained in each round of prediction based on the forward data set, the backward data set and K second detection data; then the sensor reconstruction data obtained in this round of prediction is fused into the forward data set or the backward data set to predict the next moment of that moment.

[0151] For example, the sensor detects data {x 1 、x 2 、x 3 , …, xp 、x p+1 , …, x p+t 、x p+t+1 , …, x T The missing data in {x p+1 , …, x p+t}, the forward data set is {x 1 、x 2 、x 3 , …, x p}, the backward dataset is {x p+t+1 , …, x T}.

[0152] The missing data of the missing block can be predicted from front to back. For example, for the first missing moment, the fusion data of the forward data set, the backward data set and the first missing moment are used to determine the value of x p+1 The corresponding sensor reconstructed data x p+1 '. For the second missing moment, the first round forward data set is {x 1 、x 2 、x 3 , …, x p 、x p+1 '}, the first round of forward and backward data sets are {x p+t+1 , …, x T}.

[0153] Alternatively, the missing data of the missing block is predicted from back to front. For example, for the Tth missing moment, the data corresponding to x is determined based on the fusion data of the forward data set, the backward data set, and the tth missing moment. p+t The corresponding sensor reconstructed data x p+t '. For the t-1th missing moment, the first round forward data set is {x 1 、x 2 、x 3 , …, x p}, the first round of forward and backward data sets are {x p+t ', x p+t+1 , …, x T}.

[0154] According to the embodiments of the present disclosure, the forward prediction value of forward propagation is determined according to the forward data set of each round, and the backward prediction value of backward propagation is determined according to the backward data set of each round. Based on the attention mechanism, different weights are assigned to the forward prediction value and the backward prediction value, and the prediction value at the prediction moment is obtained by combining them.

[0155] Figure 4 The flowchart of determining sensor reconstruction data according to a forward data set, a backward data set and K second detection data according to an embodiment of the present disclosure is schematically shown.

[0156] like Figure 4 As shown, for the prediction process of the t+1th round, the prediction value 403 at the t+1th moment is determined according to the t-round forward data set 401 and the t-round backward data set 402. The prediction value 403 at the t+1th moment and the fusion data 404 at the t+1th moment are combined to obtain the sensor reconstruction data 405 at the t+1th moment. Afterwards, the sensor reconstruction data 405 at the t+1th moment and the t-round forward data set 401 are combined into the t+1th round forward data set 406, and the sensor reconstruction data 405 at the t+1th moment and the t-round backward data set 402 are combined into the t+1th round backward data set 407.

[0157] Similar to the calculation of the sensor reconstruction data at the t+1th time, the prediction value 408 at the t+2th time is determined based on the t+1th round forward data set 406 and the t+1th round backward data set 407. The prediction value 408 at the t+2th time and the fusion data 409 at the t+2th time are combined to obtain the sensor reconstruction data 410 at the t+2th time.

[0158] According to the embodiments of the present disclosure, although the missing data of the block can be predicted only by the first detection data to which the missing data belongs, the predicted value may be biased as a whole. Therefore, the K second detection data of the same sensor type as the first detection data can be combined to fuse the data of the same sensor type dimension to obtain the final sensor detection data.

[0159] The embodiments of the present disclosure can improve the accuracy of generating sensor reconstructed data and reduce data deviation caused by prediction based only on data features before and after block loss by fusing data features before block loss, data features after block loss, and features of second detection data located at the block loss.

[0160] According to an embodiment of the present disclosure, the process of determining fused data based on K second detection data may include: determining the first historical detection data of the first sensor that collects the first detection data, and the second historical detection data of the second sensor that collects each second detection data; calculating the similarity between the first historical detection data and each second historical detection data at the missing position; and determining the second detection data corresponding to the second sensor with the highest similarity as the fused data.

[0161] According to an embodiment of the present disclosure, the first sensor and the second sensor are of the same sensor type, the first sensor is used to collect the first detection data, and the second sensor is used to detect the second detection data. For example, the first sensor and the second sensor are both temperature sensors.

[0162] According to an embodiment of the present disclosure, the first historical detection data and the second historical detection data are respectively historical detection data collected by the first sensor and the second sensor in the target area, target scene, and the same time period. The first historical detection data and the second historical detection data are both time series data.

[0163] According to an embodiment of the present disclosure, the first historical detection data can be jointly determined according to the target area identifier, the target scene identifier and the first sensor identifier; and the second historical detection data can be jointly determined according to the target area identifier, the target scene identifier and the second sensor identifier.

[0164] According to the embodiment of the present disclosure, since both the first historical detection data and the second historical detection data are time series data, the similarity between the first historical detection data and the second historical detection data can be determined by calculating the distance between the two time series data, and the second detection data corresponding to the second sensor with the highest similarity is determined as the fusion data.

[0165] According to an embodiment of the present disclosure, the above method determines the second sensor that is most relevant to the first sensor by calculating the similarity between the second sensor of each of the K second detection data and the data collected by the first sensor, and determines the second detection data collected by the most relevant second sensor as the fusion data. It is capable of determining the detection data that is closest to the first detection data from the K second detection data, thereby further improving the accuracy of generating sensor reconstruction data.

[0166] As another specific embodiment of the present disclosure, the similarity between the data collected by the second sensor and the first sensor of each of the K second detection data can be used as a weight, and the K second detection data can be weighted averaged to obtain fused data, so as to improve the accuracy of generating sensor reconstruction data.

[0167] According to an embodiment of the present disclosure, the step of determining K second detection data includes: determining the position information of the first sensor that collects the first detection data; determining, based on the position information, K second sensors that are closest to the first sensor and have the same sensor type as the first sensor; and obtaining the second detection data collected by each of the K second sensors to obtain K second detection data.

[0168] According to the embodiments of the present disclosure, each functional area in the smart park covers a wide geographical range, and the sensor detection values ​​collected by the same sensor type at different locations may be different. For example, in the office area, the temperature at the air-conditioning outlet is lower than that in other areas, and the temperature under the lighting equipment is slightly higher than the temperature of the ground. Therefore, K second sensors can be determined based on the location information and sensor type of the first sensor that collects the first detection data to determine K second detection data.

[0169] According to an embodiment of the present disclosure, the location information of the first sensor can be obtained from the system according to the first sensor identifier, and then K second sensors that are closest to the first sensor and have the same sensor type as the first sensor are matched in the system. After the K second sensors are determined, the sensor detection data for the target area and the target scene collected by each of the K second sensors can be obtained, that is, the second detection data.

[0170] According to an embodiment of the present disclosure, multiple sensor detection data matching a target area and a target scene are obtained, including: determining at least one electrical device in the target area and the target scene; determining a sensor detection scheme corresponding to the target area and the target scene based on the sensor type and the number of sensors of each of the at least one electrical device; and obtaining multiple sensor detection data based on the sensor detection scheme.

[0171] According to the embodiments of the present disclosure, at least one electrical device can be configured in the same target area and target scene. For example, an air conditioner and a lighting device can be configured in an office area. Each electrical device can be configured with at least one sensor type and at least one sensor quantity. For example, for an air conditioner, three temperature sensors and one humidity sensor can be configured; for a lighting device, two temperature sensors and one light sensor can be configured.

[0172] According to an embodiment of the present disclosure, the sensor type and the number of sensors are related to the electrical device, and the same sensor can be used to detect the operating status of different electrical devices. Therefore, according to the sensor type and the number of sensors of at least one electrical device, a sensor detection scheme corresponding to the target area and the target scene can be determined.

[0173] According to an embodiment of the present disclosure, for a general sensor type, the maximum value of the number of sensors in at least one electrical device can be determined as the number of sensors in the sensor detection scheme; for a special sensor type, the number of sensors is not modified. The general sensor type and the special sensor type can be determined according to the type identifier in the pre-configured sensor type table.

[0174] For example, still taking the above air conditioner and lighting equipment as an example, the target scene of the office area is equipped with an air conditioner and a lighting equipment. The air conditioner is configured with 3 temperature sensors and 1 humidity sensor, and the lighting equipment is configured with 2 temperature sensors and 1 photosensitive sensor. The temperature sensor and humidity sensor are both general sensor types. The number of temperature sensors in the sensor detection scheme is 3, and the number of humidity sensors is 1. The photosensitive sensor is a special sensor type, and the number of photosensitive sensors in the sensor detection scheme is 1.

[0175] According to an embodiment of the present disclosure, the sensor detection scheme also includes the sensor acquisition frequency. The acquisition frequency can be determined based on the characteristics of the target area, target scene, and electrical equipment as well as the sensor monitoring indicator characteristics. For example, the acquisition frequency of the temperature sensor can be once every 5 minutes.

[0176] According to an embodiment of the present disclosure, in response to detecting that there are missing data in multiple sensor detection data, based on the data missing category, the missing data is reconstructed to obtain sensor reconstructed data corresponding to the missing data, including: dividing the multiple sensor detection data according to a predetermined acquisition interval; in response to detecting that the data lengths of the divided sensor detection data are different, filling the missing data with zeros and updating the sensor detection data; in response to detecting that the data lengths of the divided sensor detection data are the same, adding the same time index to the divided sensor detection data and updating the sensor detection data.

[0177] According to an embodiment of the present disclosure, the collection interval may be determined according to the collection frequency. For example, the collection interval may be equal to the interval duration of the collection frequency, or may be an integer multiple of the interval duration of the collection frequency.

[0178] According to an embodiment of the present disclosure, when storing sensor detection data, the sensor detection data needs to be mapped to a fixed collection time period for subsequent processing. In addition, by dividing a plurality of sensor detection data according to a predetermined collection interval, the sensor detection data can also be time-corrected to determine whether there is missing data.

[0179] According to an embodiment of the present disclosure, detecting the data length of the divided sensor detection data includes: for the same usage element, respectively calculating the data volume of multiple sensor detection data under the time length N according to the collection interval.

[0180] For example, if the collection interval is a and the amount of data in a month is N m / a, the amount of data N in days d / a, the amount of data in hours N h / a.

[0181] According to an embodiment of the present disclosure, if the data amounts of multiple sensor detection data are the same and the data lengths representing the usage elements are the same, there is no missing data in the multiple sensor detection data. If the data amounts of multiple sensor detection data are different and the data lengths representing the usage elements are different, there is missing data in the multiple sensor detection data.

[0182] According to an embodiment of the present disclosure, different data lengths of elements are used to characterize the presence of missing data in multiple sensor detection data. Taking the sensor detection data with the longest data length as a reference, the missing data in other sensor detection data are filled with zeros to update the sensor detection data.

[0183] According to an embodiment of the present disclosure, when the data lengths of the divided sensor detection data are the same, the divided plurality of sensor detection data may be stored according to the same time index.

[0184] According to an embodiment of the present disclosure, the data after the zero padding operation can be used as point missing or block missing data. When detecting whether there is missing data in the multiple sensor detection data, it can be determined whether there is data padded by the zero padding operation in the multiple sensor detection data.

[0185] In the embodiments of the present disclosure, by processing sensor detection data with the same or different data lengths, the sensor detection data can be processed into standard time series data for subsequent storage and data processing.

[0186] According to an embodiment of the present disclosure, in response to detecting the presence of missing data in multiple sensor detection data, based on the data missing category, the missing data is reconstructed to obtain sensor reconstructed data corresponding to the missing data, and it also includes: determining the abnormal moment in the sensor detection data based on the detection data of two adjacent moments in the sensor detection data and an abnormal detection threshold; and clearing the detection data at the abnormal moment and updating the sensor detection data.

[0187] According to an embodiment of the present disclosure, based on the detection data of two adjacent moments in the sensor detection data and the abnormal detection threshold, determining the abnormal moment in the sensor detection data includes: calculating the difference between the detection data of the two adjacent moments and the ratio of the earlier moment of the two adjacent moments, and determining the abnormal moment in the sensor detection data by comparing the above ratio and the abnormal detection threshold. Specifically, when the ratio is greater than or equal to the abnormal detection threshold, it is determined that the sensor detection value of the later moment of the two adjacent moments is an abnormal value, and the moment is an abnormal moment.

[0188] According to the detection data of two adjacent moments in the sensor detection data and the abnormal detection threshold, the abnormal moment in the sensor detection data can be determined by referring to the following formula:

[0189]

[0190] Among them, x j and x i are all sensor detection values ​​in the sensor detection data, and x j The collection time is at x iAfterwards, δ represents the anomaly detection threshold and can be determined according to the actual situation.

[0191] According to an embodiment of the present disclosure, after the abnormal moment is determined, the sensor detection value at the abnormal moment may be cleared or set to a special value, such as 0.

[0192] In the embodiments of the present disclosure, by comparing the detection data of two adjacent moments in the sensor detection data with the abnormal detection threshold, the abnormal moment in the sensor detection data can be determined from the data processing perspective; by clearing the detection data at the abnormal moment, the abnormal moment can be processed as a point missing, so that the point missing can be reconstructed later, thereby avoiding the abnormal point from affecting the accuracy of the sensor detection data.

[0193] According to an embodiment of the present disclosure, after generating the above-mentioned sensor reconstruction data, it also includes: processing the sensor reconstruction data by wavelet transform to separate the noise data from the information data in the sensor reconstruction data; and replacing the sensor reconstruction data with the information data to obtain updated sensor reconstruction data. According to the updated sensor reconstruction data and the complete data in the sensor detection data, the target detection data is determined.

[0194] According to an embodiment of the present disclosure, the information data may be understood as data obtained by smoothing the sensor reconstructed data.

[0195] According to the embodiments of the present disclosure, by performing data smoothing again on the sensor reconstructed data, sensor reconstructed data that is more consistent with the actual operating state can be acquired, so as to obtain more accurate target detection data.

[0196] The traditional energy consumption management of electrical equipment in smart parks adopts traditional passive service management modes such as manual table lookup and manual troubleshooting. Its inefficient, high-cost, and low-service management mode cannot adapt to the property management of smart parks. The embodiments of the present disclosure provide a method for generating detection data by combining electrical equipment with the Internet of Things, which is conducive to the operation management and energy consumption detection of electrical equipment. In addition, it is also possible to reconstruct abnormal values ​​and missing values ​​in sensor detection data, improve the accuracy of sensor detection data, and more accurately obtain the actual operation conditions of electrical equipment with different operating rules.

[0197] Figure 5 The flowchart of a method for processing energy consumption of an electric device according to another embodiment of the present disclosure is schematically shown.

[0198] like Figure 5 As shown, the method includes operations S510 to S530.

[0199] In operation S510, object detection data corresponding to a target area and a target scene is acquired.

[0200] According to an embodiment of the present disclosure, the target detection data is determined according to the above detection data generation method. After executing the detection data generation method, the system can store the target detection data, and when the target detection data needs to be called, it can be obtained according to the target area identifier and the target scene identifier.

[0201] According to an embodiment of the present disclosure, the target area and the target scene are the same as those described in the above detection data generating method, and are not described again here.

[0202] In operation S520 , third detection data for a target electric device is determined from the target detection data.

[0203] According to an embodiment of the present disclosure, the target detection data matches the target area and the target scene, and there may be at least one electrical device in the target area and the target scene. Therefore, when performing energy consumption processing on the electrical device, the third detection data for the target electrical device may be extracted from the target detection data according to the identification of the target electrical device.

[0204] According to an embodiment of the present disclosure, the third detection data may include data reconstructed based on the above detection data generating method.

[0205] In operation S530, the third detection data is input into an energy consumption prediction model for a target electric device, and target prediction data is output.

[0206] According to the embodiments of the present disclosure, the energy consumption prediction model includes a recurrent neural network (RNN), and variations of the recurrent neural network, such as a long short-term memory network (LSTM), a gated recurrent neural network (GRU), etc. Other neural networks that can make predictions based on time series data may also be used, and the embodiments of the present disclosure are not limited here.

[0207] According to the embodiments of the present disclosure, since multiple electrical devices in a smart park may have multiple operating rules, corresponding energy consumption prediction models may be constructed in advance for different types of electrical devices in different scenarios. According to the embodiments of the present disclosure, the corresponding energy consumption prediction model may be determined according to the type identification of the electrical device.

[0208] According to an embodiment of the present disclosure, the target prediction data includes energy consumption prediction values ​​at S consecutive moments after the third detection data, where S is greater than or equal to 1. The target prediction data may also be power consumption.

[0209] According to the embodiments of the present disclosure, energy consumption prediction models corresponding to different scenarios and different types of electrical equipment may have different network layers and prediction time steps. The prediction time step may be understood as the number of predicted moments. For example, if the prediction time step is 6, the energy consumption prediction model can output prediction data for 6 consecutive moments.

[0210] According to an embodiment of the present disclosure, based on the above energy consumption prediction model, the energy consumption pattern can be identified by adding a network layer.

[0211] In the embodiments of the present disclosure, by acquiring the third detection data for the target detection data in the target area and the target scenario, the third detection data can be obtained in a targeted manner based on the target usage scenario of the target area and the target electrical equipment. By processing the third detection data with respect to the energy consumption identification model of the target electrical equipment, it is possible to apply the target electrical equipment with different operating rules to achieve the prediction of energy consumption values. In the embodiments of the present disclosure, by predicting the energy consumption of electrical equipment in the smart park at the next S moments, it is helpful to adjust the operating parameters of the electrical equipment in advance, identify energy consumption anomalies of the electrical equipment, etc., solves the technical problems of high operation and maintenance costs and low maintenance efficiency caused by handling problems after they occur, improves the efficiency of abnormal detection of electrical equipment in the smart park, and reduces abnormal detection costs and equipment maintenance costs.

[0212] According to an embodiment of the present disclosure, after acquiring the target prediction data for the target electrical equipment, the method further includes: performing abnormality identification on the target prediction data to obtain an abnormality identification result.

[0213] According to an embodiment of the present disclosure, performing abnormality identification on target prediction data to obtain abnormality identification results includes: determining an energy consumption pattern that matches the target prediction data; and performing abnormality identification on the target prediction data based on the energy consumption pattern to obtain abnormality identification results.

[0214] According to the embodiments of the present disclosure, the same electrical device has different operating rules in different energy consumption modes. For example, the energy consumption value in a high load state is higher than the energy consumption value in a low load state. Therefore, different abnormality identification operations can be performed for different energy consumption modes.

[0215] According to an embodiment of the present disclosure, determining an energy consumption pattern that matches target prediction data includes: inputting the target prediction data into an energy consumption pattern recognition model, and outputting an energy consumption pattern recognition result.

[0216] According to an embodiment of the present disclosure, determining an energy consumption pattern that matches the target prediction data further includes: determining an energy consumption pattern recognition result according to an energy consumption value interval to which the target prediction data belongs or a trend of energy consumption values ​​at S consecutive moments.

[0217] According to an embodiment of the present disclosure, anomaly identification is performed on target prediction data based on the energy consumption mode to obtain the anomaly identification result, including: obtaining the third historical detection data of the target electrical equipment for S' moments in the energy consumption mode, wherein the S' moments and the S moments belong to the same time period of different periods, and S' is greater than or equal to S; and determining the anomaly identification result based on the similarity between the third historical detection data and the target prediction data.

[0218] According to the embodiments of the present disclosure, in the above energy consumption mode, the third historical detection data of the target electrical equipment in the same target scene, target area, and the same time period at different times can reflect the historical operation law of the target electrical equipment. Therefore, it is possible to determine whether the target prediction data is abnormal based on the third historical detection data.

[0219] For example, for air conditioners in cloudy days, afternoons, working days, and office areas, the energy consumption value from 1:00 to 2:00 on November 1 is predicted. September 30 is also a working day and the weather is cloudy, so the third historical detection data can be the energy consumption value from 1:00 to 2:00 in the afternoon on cloudy days, working days, and office areas.

[0220] In the embodiment of the present disclosure, since the third historical detection data can reflect the historical operating rules of the target electrical equipment, the accuracy of abnormality identification can be improved by calculating the similarity between the third historical detection data and the target prediction data.

[0221] According to an embodiment of the present disclosure, determining the anomaly recognition result based on the similarity between the third historical detection data and the target prediction data includes: calculating the distance between the third historical detection data and the target prediction data using a dynamic time warping algorithm; and determining the anomaly recognition result based on a comparison relationship between the distance and a trend threshold.

[0222] According to an embodiment of the present disclosure, both the third historical detection data and the target prediction data are time series data. Through the dynamic time warping algorithm, the distance between the third historical detection data and the target prediction data can be calculated while ensuring the consistency of the operating trends.

[0223] According to an embodiment of the present disclosure, calculating the distance between the third historical detection data and the target prediction data includes: calculating the distance between the detection data at each moment in the third historical detection data and the target prediction data at each moment to form a distance matrix. Starting from the upper left corner of the distance matrix, determining the minimum distance sum to reach the lower right corner of the distance matrix, and using the above minimum distance sum as the distance between the third historical detection data and the target prediction data.

[0224] According to an embodiment of the present disclosure, the process of determining the minimum distance sum is the process of making the target prediction data correspond to one or more moments in the third historical detection data according to the trend of the time series data.

[0225] Figure 6 A flowchart of a method for calculating the distance between third historical detection data and target prediction data according to another embodiment of the present disclosure is schematically shown.

[0226] like Figure 6 As shown, the upper curve is the target prediction data, and the lower curve is the third historical detection data. The target prediction data includes point A 1 and point B 1 , the third historical detection data includes point C 1 , point D 1 and point D 2 .

[0227] Starting from the upper left corner of the distance matrix, in the process of determining the minimum distance to the lower right corner of the distance matrix, point A on the target prediction data can be 1 and point C in the third historical detection data 1 Correspondingly, point B on the target prediction data 1 and point D in the third historical detection data 1 and point D 2 Make sure that from point A 1 To point B 1 The curve trend is the same as that from point C 1 , D 1 To point D 2 The curve trend is the same.

[0228] According to an embodiment of the present disclosure, determining the abnormality recognition result based on the comparison relationship between the distance and the trend threshold includes: generating an abnormality recognition success when the distance is greater than or equal to the trend threshold; generating an abnormality recognition failure when the distance is less than the trend threshold.

[0229] In an embodiment of the present disclosure, by calculating the distance between the third historical detection data and the target prediction data, and determining the anomaly recognition result based on the comparison relationship between the distance and the trend threshold, the accuracy of anomaly detection can be improved.

[0230] According to an embodiment of the present disclosure, the method for processing energy consumption of electrical equipment also includes: in response to detecting missing data in the third detection data, issuing an alarm based on the first alarm mode; in response to the abnormality identification result representing an abnormality, issuing an alarm based on the second alarm mode.

[0231] According to an embodiment of the present disclosure, the first alarm mode indicates that the device is abnormal, which can be reflected by generating a first alarm message. The second alarm mode indicates that the energy consumption is abnormal, which can be reflected by generating a second alarm message.

[0232] According to an embodiment of the present disclosure, for each electric device, the third detection data of the electric device in the target area and target scene can be displayed to the user through a visualization interface. After predicting the energy consumption value of the electric device at S moments in the future, it can also be displayed on the visualization page in the form of a trend chart.

[0233] According to the embodiments of the present disclosure, the system can visualize the data of the detection data generation method and the power consumption processing method of the electrical equipment and display them to the user.

[0234] Figure 7 The flowchart of the method for generating an adjustment strategy for an electric device according to another embodiment of the present disclosure is schematically shown.

[0235] like Figure 7 As shown, the method includes operations S710 to S720.

[0236] In operation S710, a utility interval for a target electrical device is acquired according to a target area and a target scenario, wherein the utility interval represents an operating parameter interval of an optimal energy consumption state of the electrical device.

[0237] In operation S720 , an adjustment strategy for the target electric device is generated according to the utility interval and the third detection data.

[0238] According to the embodiments of the present disclosure, the corresponding utility interval can be set in advance on the system according to the functional area and usage scenario to which the electrical equipment belongs. For example, the temperature interval in summer is 25°C to 27°C, and the humidity interval is 50% to 60%.

[0239] According to an embodiment of the present disclosure, an adjustment strategy for the target electrical equipment is generated according to whether the third detection data is in the utility interval.

[0240] For example, for the air conditioner, the detection value of the temperature sensor in the third detection data is 24°C, and the detection value of the humidity sensor is 50%. The temperature range in the utility interval is 25°C to 27°C, and the humidity range is 50% to 60%. Since the detection value of the temperature sensor is not within the temperature range, and the detection value of the humidity sensor is within the humidity range, an adjustment strategy can be generated: maintain the humidity and set the temperature to be reached in the air conditioner to 25°C.

[0241] In the embodiments of the present disclosure, by predicting the energy consumption of electrical equipment in a smart park at the next S moments, and actively and in advance adjusting the operating parameters of the electrical equipment based on the utility range, the efficiency of abnormality detection of electrical equipment in the smart park can be improved, and the cost of abnormality detection and equipment maintenance can be reduced.

[0242] Figure 8The diagram schematically shows a smart park electrical equipment energy consumption management system according to an embodiment of the present disclosure.

[0243] like Figure 8 As shown, the smart park power equipment energy consumption management system 800 includes a functional area management module 801, a scene management module 802, and a device management module 803. The functional area management module 801 is used to perform partition management on multiple functional areas in the smart park. There is data interaction between the functional area management module 801, the scene management module 802, the device management module 803 and the power equipment sensor data acquisition module 804, the power equipment operation real-time detection module 805, and the power equipment operation optimization control module 806.

[0244] The scene management module 802 is used to classify and manage multiple usage elements in the smart park, and is also used to combine usage scenes. The equipment management module 803 is used to classify and manage multiple power-consuming devices in the smart park.

[0245] Through the pre-set smart park zoning data, scene classification data, equipment type, installation location data, and sensors used to monitor the operating status of electrical equipment in various scenarios, multiple devices can be classified and zoned, and can be used for the generation of multi-scenario decisions for the electrical equipment sensor data acquisition module, the electrical equipment operation real-time detection module, and the electrical equipment operation optimization control module. Among them, the zoning data includes the building number in the functional area set, etc.; the scene classification data includes conference rooms, employee workstations, etc.; the equipment types include lighting equipment, air conditioning and heating equipment, printers in office equipment, etc.; the installation location data includes the building number data, the building to which the equipment belongs, the room to which it belongs, etc.; the sensors include photosensitive sensors, light sensors, electricity meters, temperature sensors, humidity sensors, etc.

[0246] The power equipment sensor data acquisition module 804 is used to generate a multi-scenario power equipment energy consumption monitoring scheme 8041 based on known areas, scenarios, and equipment classifications to meet the needs of real-time monitoring of power equipment and optimization of operation decisions. The generation of a multi-scenario power equipment energy consumption monitoring scheme includes the types and quantities of power equipment and operating sensors. According to the working scenarios of different types of power equipment and the data characteristics of sensor monitoring indicators, a multi-scenario power equipment sensor data acquisition scheme 8042 is generated, and optimized and adjusted according to the needs of subsequent decisions. In order to ensure the smooth progress of subsequent monitoring and decision optimization, the power equipment sensor data acquisition module can also perform pre-processing work such as cleaning and conversion on the data collected by the power equipment sensors, and realize multi-scenario power equipment sensor detection data processing and storage 8043.

[0247] Real-time detection module for power equipment operation 805: Based on the collection of different types of power equipment operation data in different scenarios, the module uses the technology of time series pattern recognition to realize multi-scenario power equipment energy consumption recognition and prediction 8051. It is also possible to realize multi-scenario power equipment energy consumption anomaly detection 8052 and multi-scenario power equipment fault anomaly detection 8053 in the scenario dimension and the device type dimension based on the prediction technology of time series. For example, for real-time energy consumption data, energy consumption anomaly detection is performed by calculating the difference between the energy consumption mode, so that property management personnel can find energy consumption anomalies in time and prevent the waste of electric energy.

[0248] The real-time detection module 805 for the operation of electrical equipment can also visualize equipment failure data, energy consumption data, etc., and realize the operation of electrical equipment in multiple scenarios. The equipment visualization 8054 makes it convenient for property management personnel and equipment users to intuitively observe the use of electrical equipment.

[0249] The power equipment operation optimization control module 806 can realize multi-scenario power equipment operation data analysis 8061, multi-scenario power equipment operation parameter optimization 8062, and multi-scenario power equipment load optimization 8063 based on the power equipment sensor data acquisition module and the power equipment operation real-time detection module.

[0250] For example, it can identify energy consumption peaks and valleys, energy consumption trends and abnormal situations to help users understand energy consumption situations and trends; provide energy consumption reports and trend analysis functions so that users can understand the historical energy consumption of electrical equipment and discover potential energy-saving optimization opportunities; and formulate appropriate operating parameter ranges and strategies based on scenarios and equipment characteristics, and optimize energy consumption by adjusting the operating parameters of electrical equipment, such as temperature, humidity, light brightness, etc.; at the same time, it can monitor the load situation of electrical equipment, identify load peaks and fluctuations, and adjust the operating status and time period of electrical equipment through load management to balance the load and improve energy utilization efficiency.

[0251] Fig. 9 The present invention schematically shows a flow chart of a method for generating detection data of electrical equipment in a smart park and processing energy consumption according to a specific embodiment of the present disclosure.

[0252] like Fig. 9 As shown, in a specific embodiment of the present disclosure, the process from generating detection data to processing energy consumption of electrical equipment is described.

[0253] In operation S901, a target scene and a target area are acquired.

[0254] In operation S902 , a sensor detection solution for a target scene and a target area is generated.

[0255] In operation S903 , a sensor acquisition frequency is set.

[0256] In operation S904, time correction is performed on the sensor detection data.

[0257] In operation S905, missing value processing is performed on the data outliers.

[0258] In operation S906, the missing data is reconstructed.

[0259] In operation S907 , an energy consumption prediction model and an energy consumption pattern recognition model are constructed.

[0260] In operation S908, energy consumption of the electrical equipment is identified and predicted.

[0261] In operation S909 , abnormal energy consumption of the electric device is detected.

[0262] In operation S910, device abnormality detection is performed on an electric device.

[0263] In operation S911, operation data of the electric device is visualized.

[0264] In operation S912, a utility range of the electric device is set according to the target area and the target scene.

[0265] In operation S913 , sensor detection data is selected according to the target area and the target scene.

[0266] In operation S914, the operating parameters of the electric device are adjusted.

[0267] In operation S915 , load optimization adjustment is performed.

[0268] The traditional energy consumption management of electrical equipment in smart parks adopts traditional passive service management modes such as manual table lookup and manual troubleshooting. Its inefficient, high-cost, and low-service management mode cannot adapt to the property management of smart parks. The embodiments of the present disclosure provide a method for generating detection data by combining electrical equipment with the Internet of Things, which is beneficial to the operation management and energy consumption detection of electrical equipment. In addition, it is also possible to reconstruct abnormal values ​​and missing values ​​in sensor detection data, improve the accuracy of sensor detection data, and more accurately obtain the actual operation conditions of electrical equipment with different operating rules. At the same time, the present disclosure provides an execution framework for an energy consumption processing method for electrical equipment, which is beneficial to solving property electricity energy management problems from a system perspective.

[0269] Fig.10 The structural block diagram of the detection data generating method according to an embodiment of the present disclosure is schematically shown.

[0270] like Fig.10As shown, the detection data generating device 1000 includes a first acquisition module 1010 , a reconstruction module 1020 and a generation module 1030 .

[0271] A first acquisition module 1010 is used to acquire multiple sensor detection data matching a target area and a target scene, wherein the target area represents a functional area within a building object, the target scene represents a usage scene of an electrical device, and the sensor detection data includes detection data collected by at least one sensor;

[0272] A reconstruction module 1020, configured to, in response to detecting that missing data exists in the plurality of sensor detection data, reconstruct the missing data based on the data missing category to obtain sensor reconstructed data corresponding to the missing data;

[0273] The generating module 1030 is used to generate target detection data according to the complete data in the sensor reconstruction data and the sensor detection data.

[0274] According to an embodiment of the present disclosure, the plurality of sensor detection data includes detection data of M sensor types, where M is greater than or equal to 1. The reconstruction module 1020 includes a first reconstruction submodule, which is used to reconstruct the missing data based on the data missing category using detection data of the same type as the sensor that collected the missing data to obtain sensor reconstructed data.

[0275] According to an embodiment of the present disclosure, data missing categories include point missing and block missing. Point missing represents data missing at one moment, and block missing represents data missing for T consecutive moments, where T is greater than or equal to 2.

[0276] According to an embodiment of the present disclosure, the reconstruction module 1020 further includes a second reconstruction submodule and a third reconstruction submodule.

[0277] The second reconstruction submodule is used for, in response to determining that the data missing category is point missing, reconstructing the missing data by using the first detection data to which the missing data belongs through a linear interpolation method to obtain sensor reconstructed data.

[0278] The third reconstruction submodule is used to determine sensor reconstructed data in response to determining that the data missing category is block missing, based on the first detection data and K second detection data of the same type as the sensor that collected the first detection data, where K is greater than or equal to 1.

[0279] According to an embodiment of the present disclosure, the third reconstruction submodule includes a first reconstruction unit and a second reconstruction unit.

[0280] The first reconstruction unit is used to divide the first detection data into a forward data set and a backward data set according to the position of the missing data in the first detection data.

[0281] The second reconstruction unit is used to input the forward data set, the backward data set and K second detection data into the bidirectional recurrent neural network, and output sensor reconstructed data corresponding to the missing data.

[0282] According to an embodiment of the present disclosure, the second reconstruction unit includes a first reconstruction subunit, a second reconstruction subunit, and a third reconstruction subunit. Repeat the following steps until the sensor reconstruction data at T moments are obtained:

[0283] The first reconstruction subunit is used to determine a forward prediction value corresponding to a t-th round forward data set and a backward prediction value corresponding to a t-th round backward data set, wherein the t-th round forward data set includes a forward data set and sensor reconstructed data at t consecutive moments obtained after t rounds of processing, t is greater than or equal to 1, and t is less than or equal to T.

[0284] The second reconstruction subunit is used to process the forward prediction value and the backward prediction value based on the attention mechanism to obtain the prediction value at the t+1th moment.

[0285] The third reconstruction subunit is used to fuse the predicted value at the t+1th moment and the fused data at the t+1th moment to obtain the sensor reconstructed data at the t+1th moment, wherein the fused data is determined based on the K second detection data.

[0286] According to an embodiment of the present disclosure, the detection data generating device 1000 includes a fusion module, and the fusion module includes a first fusion submodule, a second fusion submodule and a third fusion submodule.

[0287] The first fusion submodule is used to determine the first historical detection data of the first sensor that collects the first detection data and the second historical detection data of the second sensor that collects each second detection data.

[0288] The second fusion submodule is used to calculate the similarity between the first historical detection data and each second historical detection data at the missing position.

[0289] The third fusion submodule is used to determine the second detection data corresponding to the second sensor with the highest similarity as the fusion data.

[0290] According to an embodiment of the present disclosure, the detection data generating device 1000 further includes a third acquisition module, and the fusion module includes a first acquisition submodule, a second acquisition submodule and a third acquisition submodule.

[0291] The first acquisition submodule is used to determine the position information of the first sensor for collecting the first detection data.

[0292] The second acquisition submodule is used to determine, according to the position information, K second sensors which are closest to the first sensor and have the same sensor type as the first sensor.

[0293] The third acquisition submodule is used to acquire the second detection data collected by each of the K second sensors to obtain K second detection data.

[0294] According to an embodiment of the present disclosure, the first acquisition module includes a fourth acquisition submodule, a fifth acquisition submodule and a sixth acquisition submodule.

[0295] The fourth acquisition submodule is used to determine at least one electrical device in the target area and the target scene.

[0296] The fifth acquisition submodule is used to determine a sensor detection scheme corresponding to the target area and the target scene according to the sensor type and the number of sensors of at least one electrical device.

[0297] The sixth acquisition submodule is used to acquire multiple sensor detection data based on the sensor detection solution.

[0298] According to an embodiment of the present disclosure, the detection data generating apparatus 1000 further includes a first updating module, and the first updating module includes a first updating submodule, a second updating submodule and a third updating submodule.

[0299] The first updating submodule is used to divide the detection data of multiple sensors according to a predetermined collection interval.

[0300] The second updating submodule is used for filling the missing data with zeros in response to detecting that the data lengths of the divided sensor detection data are different, thereby updating the sensor detection data.

[0301] The third updating submodule is used for adding the same time index to the divided sensor detection data in response to detecting that the data lengths of the divided sensor detection data are the same, thereby updating the sensor detection data.

[0302] According to an embodiment of the present disclosure, the detection data generating apparatus 1000 further includes a second updating module, and the second updating module includes a fourth updating submodule, a fifth updating submodule and a sixth updating submodule.

[0303] The fourth updating submodule is used to determine the abnormal moment in the sensor detection data according to the detection data of two adjacent moments in the sensor detection data and the abnormal detection threshold.

[0304] The fifth updating submodule is used to clear the detection data at the abnormal moment and update the sensor detection data.

[0305] According to the embodiments of the present invention, any one or more of the modules, submodules, units, and subunits, or at least part of the functions of any one of them can be implemented in one module. According to the embodiments of the present invention, any one or more of the modules, submodules, units, and subunits can be split into multiple modules for implementation. According to the embodiments of the present invention, any one or more of the modules, submodules, units, and subunits can be at least partially implemented as hardware circuits, such as field programmable gate arrays (FPGAs), programmable logic arrays (PLAs), systems on chips, systems on substrates, systems on packages, application specific integrated circuits (ASICs), or can be implemented by hardware or firmware in any other reasonable way of integrating or packaging the circuit, or implemented in any one of the three implementation methods of software, hardware, and firmware, or in any appropriate combination of any of them. Alternatively, according to the embodiments of the present invention, one or more of the modules, submodules, units, and subunits can be at least partially implemented as computer program modules, and when the computer program modules are run, the corresponding functions can be performed.

[0306] For example, any multiple of the first acquisition module 1010, the reconstruction module 1020, and the generation module 1030 can be combined in one module / unit / sub-unit for implementation, or any one of the modules / units / sub-units can be split into multiple modules / units / sub-units. Alternatively, at least part of the functions of one or more of these modules / units / sub-units can be combined with at least part of the functions of other modules / units / sub-units and implemented in one module / unit / sub-unit. According to an embodiment of the present disclosure, at least one of the first acquisition module 1010, the reconstruction module 1020, and the generation module 1030 can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application specific integrated circuit (ASIC), or can be implemented by hardware or firmware such as any other reasonable way of integrating or packaging the circuit, or by any one of the three implementation methods of software, hardware, and firmware, or by a suitable combination of any of them. Alternatively, at least one of the first acquisition module 1010 , the reconstruction module 1020 , and the generation module 1030 may be at least partially implemented as a computer program module, and when the computer program module is executed, the corresponding function may be performed.

[0307] It should be noted that the detection data generating device part in the embodiment of the present disclosure corresponds to the detection data generating method part in the embodiment of the present disclosure. The description of the detection data generating device part specifically refers to the detection data generating method part, which will not be repeated here.

[0308] Fig.11The structural block diagram of the method for processing energy consumption of electric equipment according to an embodiment of the present disclosure is schematically shown.

[0309] like Fig.11 As shown, the electrical equipment energy consumption processing device 1100 includes a second acquisition module 1110 , a determination module 1120 and a prediction module 1130 .

[0310] A second acquisition module 1110, configured to acquire target detection data corresponding to the target area and the target scene, wherein the target detection data is determined according to any one of claims 1 to 10;

[0311] A determination module 1120, configured to determine third detection data for a target electrical device from the target detection data;

[0312] The prediction module 1130 is used to input the third detection data into the energy consumption prediction model for the target electrical equipment and output the target prediction data, wherein the target prediction data includes the energy consumption prediction values ​​at S consecutive moments after the third detection data, S is greater than or equal to 1, and the energy consumption prediction model includes a recurrent neural network.

[0313] According to an embodiment of the present disclosure, the electrical equipment energy consumption processing device 1100 further includes an energy consumption anomaly identification module for performing anomaly identification on target prediction data to obtain an anomaly identification result.

[0314] According to an embodiment of the present disclosure, the energy consumption anomaly identification module further includes a mode determination submodule and an anomaly identification submodule.

[0315] The mode determination submodule is used to determine the energy consumption mode that matches the target prediction data.

[0316] The anomaly recognition submodule is used to perform anomaly recognition on the target prediction data based on the energy consumption mode to obtain anomaly recognition results.

[0317] According to an embodiment of the present disclosure, the abnormality identification submodule includes a first identification unit and a second identification unit.

[0318] The first identification unit is used to obtain third historical detection data of the target electrical device for S' moments in the energy consumption mode, wherein the S' moments and the S moments belong to the same time period of different periods, and S' is greater than or equal to S.

[0319] The second recognition unit is used to determine the abnormality recognition result according to the similarity between the third historical detection data and the target prediction data.

[0320] According to an embodiment of the present disclosure, the second recognition unit includes a first recognition subunit and a second recognition subunit.

[0321] The first identification subunit is used to calculate the distance between the third historical detection data and the target prediction data using a dynamic time warping algorithm.

[0322] The second identification subunit is used to determine the abnormality identification result according to the comparison relationship between the distance and the trend threshold.

[0323] According to an embodiment of the present disclosure, the electrical equipment energy consumption processing device 1100 further includes a first alarm module and a second alarm module.

[0324] The first alarm module is used for issuing an alarm based on a first alarm mode in response to detecting that missing data exists in the third detection data.

[0325] The second alarm module is used for generating an alarm based on a second alarm mode in response to the abnormality characterized by the abnormality identification result.

[0326] According to an embodiment of the present disclosure, the apparatus 1100 for processing energy consumption of an electric device further includes a fourth acquisition module and a strategy generation module.

[0327] The fourth acquisition module is used to acquire a utility interval for a target electrical device according to a target area and a target scenario, wherein the utility interval represents an operating parameter interval of an optimal energy consumption state of the electrical device.

[0328] The strategy generation module is used to generate an adjustment strategy for the target electrical equipment according to the utility interval and the third detection data.

[0329] According to the embodiments of the present invention, any one or more of the modules, submodules, units, and subunits, or at least part of the functions of any one of them can be implemented in one module. According to the embodiments of the present invention, any one or more of the modules, submodules, units, and subunits can be split into multiple modules for implementation. According to the embodiments of the present invention, any one or more of the modules, submodules, units, and subunits can be at least partially implemented as hardware circuits, such as field programmable gate arrays (FPGAs), programmable logic arrays (PLAs), systems on chips, systems on substrates, systems on packages, application specific integrated circuits (ASICs), or can be implemented by hardware or firmware in any other reasonable way of integrating or packaging the circuit, or implemented in any one of the three implementation methods of software, hardware, and firmware, or in any appropriate combination of any of them. Alternatively, according to the embodiments of the present invention, one or more of the modules, submodules, units, and subunits can be at least partially implemented as computer program modules, and when the computer program modules are run, the corresponding functions can be performed.

[0330] For example, any multiple of the second acquisition module 1110, the determination module 1120, and the prediction module 1130 can be combined in one module / unit / subunit for implementation, or any one of the modules / units / subunits can be split into multiple modules / units / subunits. Alternatively, at least part of the functions of one or more of these modules / units / subunits can be combined with at least part of the functions of other modules / units / subunits and implemented in one module / unit / subunit. According to an embodiment of the present disclosure, at least one of the second acquisition module 1110, the determination module 1120, and the prediction module 1130 can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application specific integrated circuit (ASIC), or can be implemented by hardware or firmware such as any other reasonable way of integrating or packaging the circuit, or by any one of the three implementation methods of software, hardware, and firmware, or by a suitable combination of any of them. Alternatively, at least one of the second acquisition module 1110 , the determination module 1120 , and the prediction module 1130 may be at least partially implemented as a computer program module, and when the computer program module is executed, a corresponding function may be performed.

[0331] It should be noted that the energy consumption processing device part of the electrical equipment in the embodiment of the present disclosure corresponds to the energy consumption processing method part of the electrical equipment in the embodiment of the present disclosure. The description of the energy consumption processing device part of the electrical equipment specifically refers to the energy consumption processing method part of the electrical equipment, which will not be repeated here.

[0332] Fig.12 A block diagram of an electronic device suitable for implementing a detection data generating method and an electrical equipment energy consumption processing method according to an embodiment of the present disclosure is schematically shown. Fig.12 The electronic device shown is only an example and should not bring any limitation to the functions and scope of use of the embodiments of the present disclosure.

[0333] like Fig.12 As shown, the electronic device 1200 according to an embodiment of the present disclosure includes a processor 1201, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1202 or a program loaded from a storage part 1208 into a random access memory (RAM) 1203. The processor 1201 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or a related chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 1201 may also include an onboard memory for caching purposes. The processor 1201 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present disclosure.

[0334] In RAM 1203, various programs and data required for the operation of electronic device 1200 are stored. Processor 1201, ROM 1202 and RAM 1203 are connected to each other via bus 1204. Processor 1201 performs various operations of the method flow according to the embodiment of the present disclosure by executing the program in ROM 1202 and / or RAM 1203. It should be noted that the program can also be stored in one or more memories other than ROM 1202 and RAM 1203. Processor 1201 can also perform various operations of the method flow according to the embodiment of the present disclosure by executing the program stored in the one or more memories.

[0335] According to an embodiment of the present disclosure, the electronic device 1200 may further include an input / output (I / O) interface 1205, which is also connected to the bus 1204. The system 1200 may further include one or more of the following components connected to the I / O interface 1205: an input section 1206 including a keyboard, a mouse, etc.; an output section 1207 including a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 1208 including a hard disk, etc.; and a communication section 1209 including a network interface card such as a LAN card, a modem, etc. The communication section 1209 performs communication processing via a network such as the Internet. A drive 1210 is also connected to the I / O interface 1205 as needed. A removable medium 1211, such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc., is installed on the drive 1210 as needed, so that a computer program read therefrom is installed into the storage section 1208 as needed.

[0336] According to an embodiment of the present disclosure, the method flow according to an embodiment of the present disclosure can be implemented as a computer software program. For example, an embodiment of the present disclosure includes a computer program product, which includes a computer program carried on a computer-readable storage medium, and the computer program contains a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 1209, and / or installed from the removable medium 1211. When the computer program is executed by the processor 1201, the above-mentioned functions defined in the system of the embodiment of the present disclosure are executed. According to an embodiment of the present disclosure, the system, equipment, device, module, unit, etc. described above can be implemented by a computer program module.

[0337] The present disclosure also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments; or may exist independently without being assembled into the device / apparatus / system. The above computer-readable storage medium carries one or more programs, and when the above one or more programs are executed, the method according to the embodiment of the present disclosure is implemented.

[0338] According to an embodiment of the present disclosure, the computer-readable storage medium may be a non-volatile computer-readable storage medium. For example, it may include, but is not limited to: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in combination with an instruction execution system, apparatus, or device.

[0339] For example, according to an embodiment of the present disclosure, the computer-readable storage medium may include the ROM 1202 and / or the RAM 1203 described above and / or one or more memories other than the ROM 1202 and the RAM 1203 .

[0340] The embodiments of the present disclosure also include a computer program product, which includes a computer program, and the computer program contains program code for executing the method provided by the embodiments of the present disclosure. When the computer program product runs on an electronic device, the program code is used to enable the electronic device to implement the method provided by the embodiments of the present disclosure.

[0341] When the computer program is executed by the processor 1201, the above functions defined in the system / device of the embodiment of the present disclosure are executed. According to the embodiment of the present disclosure, the system, device, module, unit, etc. described above can be implemented by a computer program module.

[0342] In one embodiment, the computer program may rely on tangible storage media such as optical storage devices, magnetic storage devices, etc. In another embodiment, the computer program may also be transmitted and distributed in the form of signals on a network medium, and downloaded and installed through the communication part 1209, and / or installed from the removable medium 1211. The program code contained in the computer program may be transmitted using any appropriate network medium, including but not limited to: wireless, wired, etc., or any suitable combination of the above.

[0343] According to an embodiment of the present disclosure, the program code for executing the computer program provided by the embodiment of the present disclosure can be written in any combination of one or more programming languages. Specifically, these computing programs can be implemented using high-level process and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages ​​include, but are not limited to, Java, C++, python, "C" language or similar programming languages. The program code can be executed entirely on the user computing device, partially on the user device, partially on the remote computing device, or entirely on the remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (for example, using an Internet service provider to connect through the Internet).

[0344] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram may represent a module, a program segment, or a part of a code, and the above-mentioned module, program segment, or a part of the code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box may also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions. It can be understood by those skilled in the art that the features recorded in the various embodiments and / or claims of the present disclosure can be combined and / or combined in a variety of ways, even if such a combination or combination is not explicitly recorded in the present disclosure. In particular, without departing from the spirit and teaching of the present disclosure, the features described in the various embodiments and / or claims of the present disclosure may be combined and / or combined in a variety of ways. All of these combinations and / or combinations fall within the scope of the present disclosure.

[0345] The embodiments of the present disclosure are described above. However, these embodiments are only for illustrative purposes and are not intended to limit the scope of the present disclosure. Although the embodiments are described above separately, this does not mean that the measures in the various embodiments cannot be used in combination to advantage. The scope of the present disclosure is defined by the attached claims and their equivalents. Without departing from the scope of the present disclosure, those skilled in the art may make a variety of substitutions and modifications, which should all fall within the scope of the present disclosure.

Claims

1. A method for generating detection data, comprising: Acquire multiple sensor detection data matching a target area and a target scene, wherein the target area represents a functional area within a building object, the target scene represents a usage scene of an electrical device, and the sensor detection data includes detection data collected by at least one sensor; In response to detecting that missing data exists in the plurality of sensor detection data, reconstructing the missing data based on the data missing category to obtain sensor reconstructed data corresponding to the missing data; Target detection data is generated according to the sensor reconstruction data and the complete data in the sensor detection data.

2. The method according to claim 1, wherein: The plurality of sensor detection data includes detection data of M sensor types, where M is greater than or equal to 1; and the reconstructing the missing data based on the data missing category to obtain sensor reconstructed data corresponding to the missing data includes: Based on the data missing category, the missing data is reconstructed using detection data of the same type as that of a sensor that collects the missing data to obtain the sensor reconstructed data.

3. The method according to claim 1, wherein: The data missing categories include point missing and block missing. The point missing represents data missing at one moment, and the block missing represents data missing at T consecutive moments, where T is greater than or equal to 2. The reconstructing the missing data based on the data missing category to obtain sensor reconstructed data corresponding to the missing data includes: In response to determining that the data missing category is point missing, reconstructing the missing data using the first detection data to which the missing data belongs by linear interpolation to obtain the sensor reconstructed data; In response to determining that the data missing category is block missing, the sensor reconstruction data is determined according to the first detection data and K second detection data of the same type as that of a sensor used to collect the first detection data, where K is greater than or equal to 1.

4. The method according to claim 3, wherein: The determining the sensor reconstruction data according to the first detection data and K second detection data of the same type as that of a sensor that acquires the first detection data comprises: Dividing the first detection data into a forward data set and a backward data set according to the position of the missing data in the first detection data; and The forward data set, the backward data set and the K second detection data are input into a bidirectional recurrent neural network, and the sensor reconstructed data corresponding to the missing data is output.

5. The method according to claim 4, wherein: The step of inputting the forward data set, the backward data set and the K second detection data into a bidirectional recurrent neural network and outputting the sensor reconstructed data corresponding to the missing data comprises: Repeat the following steps until the sensor reconstruction data at T time points is obtained: Determine a forward prediction value corresponding to a t-th round forward data set and a backward prediction value corresponding to a t-th round backward data set, wherein the t-th round forward data set includes the forward data set and sensor reconstructed data at t consecutive moments obtained after t rounds of processing, where t is greater than or equal to 1 and t is less than or equal to T; Based on the attention mechanism, the forward prediction value and the backward prediction value are processed to obtain the prediction value at the t+1th moment; The predicted value at the t+1th moment and the fused data at the t+1th moment are fused to obtain sensor reconstruction data at the t+1th moment, wherein the fused data is determined based on the K second detection data.

6. The method according to claim 5, further comprising: Determine first historical detection data of a first sensor that collects the first detection data, and second historical detection data of a second sensor that collects each second detection data; Calculating the similarity between the first historical detection data and each second historical detection data at the missing position; as well as The second detection data corresponding to the second sensor with the highest similarity is determined as the fused data.

7. The method according to claim 3, further comprising: Determine position information of a first sensor for collecting the first detection data; Determine, according to the location information, K second sensors which are closest to the first sensor and of the same sensor type as the first sensor; as well as The second detection data collected by each of the K second sensors is acquired to obtain the K second detection data.

8. The method according to any one of claims 1 to 7, wherein: The step of acquiring a plurality of sensor detection data matching the target area and the target scene includes: Determine at least one electrical device in the target area and the target scene; Determining a sensor detection scheme corresponding to the target area and the target scene according to the sensor type and the number of sensors of each of the at least one electrical device; and Based on the sensor detection scheme, the plurality of sensor detection data are acquired.

9. The method according to any one of claims 1 to 7, wherein: In response to detecting that missing data exists in the plurality of sensor detection data, before reconstructing the missing data based on the data missing category to obtain sensor reconstructed data corresponding to the missing data, the method includes: dividing the plurality of sensor detection data according to a predetermined collection interval; In response to detecting that the data lengths of the divided sensor detection data are different, filling the missing data with zeros to update the sensor detection data; In response to detecting that the data lengths of the divided sensor detection data are the same, adding the same time index to the divided sensor detection data, and updating the sensor detection data.

10. The method according to any one of claims 1 to 7, wherein: In response to detecting that there is missing data in the plurality of sensor detection data, before reconstructing the missing data based on the data missing category to obtain sensor reconstructed data corresponding to the missing data, the method further includes: Determining an abnormal moment in the sensor detection data according to detection data at two adjacent moments in the sensor detection data and an abnormality detection threshold; and The detection data at the abnormal moment is cleared, and the sensor detection data is updated.

11. A method for processing energy consumption of electrical equipment, comprising: Acquire target detection data corresponding to the target area and the target scene, wherein the target detection data is determined according to any one of claims 1 to 10; Determining third detection data for a target electrical device from the target detection data; The third detection data is input into an energy consumption prediction model for the target electrical equipment, and target prediction data is output, wherein the target prediction data includes energy consumption prediction values ​​at S consecutive moments after the third detection data, S is greater than or equal to 1, and the energy consumption prediction model includes a recurrent neural network.

12. The method according to claim 11, further comprising: Anomaly identification is performed on the target prediction data to obtain anomaly identification results.

13. The method according to claim 12, wherein: The performing abnormality identification on the target prediction data to obtain the abnormality identification result comprises: determining an energy consumption pattern that matches the target prediction data; and Based on the energy consumption pattern, abnormality identification is performed on the target prediction data to obtain an abnormality identification result.

14. The method according to claim 13, wherein: The performing abnormality identification on the target prediction data based on the energy consumption mode to obtain the abnormality identification result includes: Acquire third historical detection data of the target electrical device for S' moments in the energy consumption mode, wherein the S' moments and the S moments belong to the same time period of different periods, and S' is greater than or equal to S; and The abnormality recognition result is determined according to the similarity between the third historical detection data and the target prediction data.

15. The method according to claim 14, wherein: The determining the abnormality recognition result according to the similarity between the third historical detection data and the target prediction data comprises: Calculating the distance between the third historical detection data and the target prediction data using a dynamic time warping algorithm; and The abnormality identification result is determined according to the comparison relationship between the distance and the trend threshold.

16. The method according to any one of claims 12 to 15, further comprising: In response to detecting that missing data exists in the third detection data, issuing an alarm based on the first alarm mode; In response to the abnormality identification result indicating an abnormality, an alarm is issued based on a second alarm mode.

17. The method according to claim 11, further comprising: According to the target area and the target scenario, obtaining a utility interval for the target electrical equipment, wherein the utility interval represents an operating parameter interval of an optimal energy consumption state of the electrical equipment; and An adjustment strategy for the target electrical equipment is generated according to the utility interval and the third detection data.

18. A detection data generating device, comprising: A first acquisition module is used to acquire a plurality of sensor detection data matching a target area and a target scene, wherein the target area represents a functional area within a building object, the target scene represents a usage scene of an electrical device, and the sensor detection data includes detection data collected by at least one sensor; A reconstruction module, configured to, in response to detecting that missing data exists in the plurality of sensor detection data, reconstruct the missing data based on the data missing category to obtain sensor reconstructed data corresponding to the missing data; A generating module is used to generate target detection data according to the sensor reconstruction data and the complete data in the sensor detection data.

19. An energy consumption processing device for electrical equipment, comprising: A second acquisition module, used to acquire target detection data corresponding to the target area and the target scene, wherein the target detection data is determined according to any one of claims 1 to 10; A determination module, configured to determine third detection data for a target electrical device from the target detection data; A prediction module is used to input the third detection data into an energy consumption prediction model for the target electrical equipment and output target prediction data, wherein the target prediction data includes energy consumption prediction values ​​at S consecutive moments after the third detection data, S is greater than or equal to 1, and the energy consumption prediction model includes a recurrent neural network.

20. An electronic device, comprising: one or more processors; a memory for storing one or more programs, Wherein, when the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 17.

21. A computer-readable storage medium having executable instructions stored thereon, which, when executed by a processor, enables the processor to implement the method according to any one of claims 1 to 17.

22. A computer program product comprising a computer program which, when executed by a processor, implements the method according to any one of claims 1 to 17.