A method and system for data fusion and perception collaborative processing

Abstract

The present invention discloses a method and system for data fusion and perception collaborative processing, belonging to the technical field of electric energy metering. The method of the present invention includes: calling the data of the information system, pre-cleaning the data, and using the cleaned data as the basic data for secondary identification of the cloud platform, and storing it in the secondary database of the cloud platform; performing multi-dimensional statistics on the basic data stored in the secondary database of the cloud platform, and performing multi-dimensional statistics on the secondary identification data stored in the secondary database of the cloud platform to obtain statistical information; obtaining the frozen power information of the electrical appliances within the preset time period of the information system, judging whether the electrical appliances are abnormal according to the statistical information, and completing the fusion and perception identification processing of the electrical appliance data. The present invention provides a data fusion application in a unified information collection system mode.

Description

Technical Field

[0001] The present invention relates to the field of electric energy metering technology, and more specifically, to a method and system for data fusion and perception collaborative processing. Background Art

[0002] With the continuous advancement of the application of non-intrusive load sensing technology and load sensing module products, the processing and effect presentation of existing network load sensing data, the auxiliary identification of cloud-based sensing algorithms, and the supervision of the industry market through certification and authentication specifications and the provision of certain intellectual property protection for module load sensing algorithms have become issues that need to be urgently addressed at this stage.

[0003] Load sensing data processing and effect presentation:

[0004] At present, only the original frozen data is uploaded, and there is a lack of processing for the rationality and integrity of the data information. There is no unified mechanism for on-demand refresh of electrical appliance types based on the original data, and the dimensions and evaluation methods of presenting the load perception effects based on the existing network have not yet been clarified.

[0005] Load sensing algorithm manufacturers can prove that their algorithms are practical and can meet the actual needs of the real environment of the existing network; understand the technical level of their own algorithms in the industry; and continuously improve and optimize their own algorithms through on-site load sensing data. Module manufacturers understand the matching degree and maturity level between their module hardware and corresponding algorithms. Users need to understand whether load sensing is accurate, stable and reliable in the existing network environment. Testing agencies also need to continuously optimize or adjust the test content and methods in the laboratory stage based on the current network situation. A bridge link for access to the fusion processing cloud system for multiple sensing algorithms is temporarily missing.

[0006] Assisted identification by cloud-based perception algorithms:

[0007] Through the algorithm research and load sensing effect analysis over the past year, we know that the load sensing module is limited by hardware resources and costs, and the algorithm's recognition ability has bottlenecks. It is necessary to supplement the load sensing algorithm on the module side with the help of the rich resources of the cloud platform. The non-intrusive load sensing module is restricted by various factors such as chips, storage, volume, power consumption, communication, and cost, making it difficult for the load sensing module to accurately identify in some cases. These situations include the need to make comprehensive judgments based on longer time cycle rules, seasons, time points, electrical appliance rules, actual power consumption, power and other information before accurate identification can be made. Due to the large differences in design principles, some electrical appliances of the same type have no commonality in physical properties during use, and it is very difficult to cluster and distinguish them on the module side. At the same time, the module cannot provide massive feature resources for machine learning methods. Therefore, the advancement of load sensing, especially the load sensing of such electrical appliances, will still need to rely on cloud resources in the future.

[0008] However, due to the limitation of the communication channel capacity, the information uploaded to the main station is very limited. The communication capacity of the current communication environment is not suitable for real-time cloud collaboration solutions with large amounts of information. It is necessary to plan the interface and mode of the information uploaded by cloud-assisted load sensing in advance based on the existing situation, so that the cloud-assisted load sensing algorithm can be better applied to the current application scenarios and have subsequent scalability. At present, the interface and mode are not clear, which affects the research and development direction and progress of the cloud-assisted load sensing algorithm.

[0009] Certification and authentication standard supervision industry market and algorithm intellectual property protection

[0010] Under the cloud-end collaborative architecture, it is impossible to distinguish the load sensing data of different algorithm manufacturers, different module manufacturers, and smart meter manufacturers; it is impossible to supervise the consistency between the load sensing algorithm version submitted for inspection and the existing network version; it is impossible to trace the source of the load sensing algorithm and protect the intellectual property rights of the original algorithm manufacturer. Summary of the invention

[0011] In view of the above problems, the present invention proposes a method for data fusion and perception collaborative processing, comprising:

[0012] Call the data of the information system, pre-clean the data, and use the cleaned data as the basic data for secondary identification of the cloud platform and store it in the secondary database of the cloud platform;

[0013] Perform multi-dimensional statistics on the basic data stored in the secondary database of the cloud platform, and perform multi-dimensional statistics on the secondary identified data stored in the secondary database of the cloud platform to obtain statistical information;

[0014] Obtain the frozen power information of electrical appliances within the preset time period of the information system, and use the cloud perception algorithm to assist in identifying the statistical information based on the statistical information, determine whether the electrical appliances have any abnormalities, and complete the fusion and perception recognition processing of the electrical appliance data.

[0015] Optionally, clean the data, including:

[0016] Complete the data and remove duplicate or abnormal data. At the same time, check for duplicates, omissions, and anomalies in the data of major electrical appliance categories and unknown electrical appliance categories and enter the additional feature information into the data;

[0017] The completion specifically includes checking for data omissions, triggering the pre-checking and freezing of data, and completing the data;

[0018] Duplicate data include: duplicate data of number, date or frozen period;

[0019] Abnormal data includes: negative power data, or data in which the comprehensive identified power exceeds the frozen power of the meter.

[0020] Optionally, cleaning includes: real-time cleaning, on-demand cleaning or scheduled cleaning.

[0021] Optional, multi-dimensional statistics, specifically: statistics on the data in terms of time dimension, appliance dimension, and geographical dimension of the region where the appliance is located;

[0022] The electrical appliance dimension includes an electrical appliance category dimension, an electrical appliance power dimension, and an additional feature information dimension of unknown electrical appliances and major electrical appliance categories.

[0023] Optionally, a cloud-based perception algorithm is used for auxiliary identification, including: Correcting statistical information, determining correct statistical information, and performing secondary identification on the correct statistical information, wherein the secondary identification includes:

[0024] Confirm the subdivision of electrical appliances in major categories and re-identify unknown electrical appliance types;

[0025] The additional cloud-based perception of unknown electrical appliances includes:

[0026] Report unknown appliance types and extract additional information including operating characteristics. Based on the additional information, perform secondary identification using machine learning or manual monitoring based on a massive feature library;

[0027] The secondary identification performs secondary labeling of the unknown electrical appliance type and the additional features, performs storage records in the module number dimension, and performs load identification of the unknown electrical appliance type corresponding to the additional features;

[0028] The confirmation of the subdivision of electrical appliances in the major category of electrical appliances includes:

[0029] Determine the additional characteristic information of the major categories of electrical appliances, report the additional characteristic information to the cloud platform, analyze the accompanying electrical appliances, power, short-term operating time and operating period information through the basic data in the secondary database of the cloud platform, find the longer-term operating rules of the electrical appliances through basic statistical data on the electrical appliances, power, short-term operating time and operating period information, and complete the subdivision confirmation of the major categories of electrical appliances;

[0030] The additional characteristic information includes: the number of times the electrical appliance starts and stops, the real 1-minute start and stop time information and the load event number information across the freezing period.

[0031] The present invention also proposes a system for data fusion and perception collaborative processing, comprising:

[0032] The cleaning module calls the data of the information system, pre-cleans the data, and uses the cleaned data as the basic data for secondary identification of the cloud platform and stores it in the secondary database of the cloud platform;

[0033] The statistical module performs multi-dimensional statistics on the basic data stored in the secondary database of the cloud platform, and performs multi-dimensional statistics on the secondary identified data stored in the secondary database of the cloud platform to obtain statistical information;

[0034] The identification module obtains the frozen power information of the electrical appliances within the preset time period of the information system, and uses the cloud perception algorithm to assist in identifying the statistical information based on the statistical information, determines whether the electrical appliances have any abnormalities, and completes the fusion and perception of the electrical appliance data.

[0035] Optionally, clean the data, including:

[0036] Complete the data and remove duplicate or abnormal data. At the same time, check for duplicates, omissions, and anomalies in the data of major electrical appliance categories and unknown electrical appliance categories and enter the additional feature information into the data;

[0037] The completion specifically includes checking for data omissions, triggering the pre-checking and freezing of data, and completing the data;

[0038] Duplicate data include: duplicate data of number, date or frozen period;

[0039] Abnormal data includes: negative power data, or data in which the comprehensive identified power exceeds the frozen power of the meter.

[0040] Optionally, cleaning includes: real-time cleaning, on-demand cleaning or scheduled cleaning.

[0041] Optional, multi-dimensional statistics, specifically: statistics on the data in terms of time dimension, appliance dimension, and geographical dimension of the region where the appliance is located;

[0042] The electrical appliance dimension includes the electrical appliance category dimension, the electrical appliance power dimension, and the additional characteristic information dimension of unknown electrical appliances and major categories of electrical appliances.

[0043] Optional,

[0044] Using cloud-based perception algorithms for auxiliary identification includes: Correcting statistical information, determining correct statistical information, and performing secondary identification on the correct statistical information, wherein the secondary identification includes:

[0045] Confirm the subdivision of electrical appliances in major categories and re-identify unknown electrical appliance types;

[0046] The additional cloud-based perception of unknown electrical appliances includes:

[0047] Report unknown appliance types and extract additional information including operating characteristics. Based on the additional information, perform secondary identification using machine learning or manual monitoring based on a massive feature library;

[0048] The secondary identification performs secondary labeling of the unknown electrical appliance type and the additional features, performs storage records in the module number dimension, and performs load identification of the unknown electrical appliance type corresponding to the additional features;

[0049] The confirmation of the subdivision of electrical appliances in the major category of electrical appliances includes:

[0050] Determine the additional characteristic information of the major categories of electrical appliances, report the additional characteristic information to the cloud platform, analyze the accompanying electrical appliances, power, short-term operating time and operating period information through the basic data in the secondary database of the cloud platform, find the longer-term operating rules of the electrical appliances through basic statistical data on the electrical appliances, power, short-term operating time and operating period information, and complete the subdivision confirmation of the major categories of electrical appliances;

[0051] The additional characteristic information includes: the number of times the electrical appliance starts and stops, the real 1-minute start and stop time information and the load event number information across the freezing period.

[0052] The present invention provides a unified information collection system mode, realizes the fusion application of data, unifies the interface and mode details of the acquisition and load perception modules, provides a basis for the fusion processing of different algorithms of multiple manufacturers on the same cloud platform, and provides a method for algorithm operation and maintenance, evaluation, directional analysis and upgrading. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 A flowchart of a method for data fusion and perception collaborative processing according to the present invention;

[0054] Figure 2 An architectural diagram of a cloud platform system for cloud collaboration used in a method for data fusion and perception collaborative processing of the present invention;

[0055] Figure 3 A flowchart of data cleaning of a method for data fusion and perception collaborative processing according to the present invention;

[0056] Figure 4 A flowchart of a method for data fusion and perception collaborative processing of the present invention for data recognition error;

[0057] Figure 5 A flowchart of cloud-based additional load perception of a method for data fusion and perception collaborative processing according to the present invention;

[0058] Figure 6 This is a system structure diagram for data fusion and perception collaborative processing in the present invention. DETAILED DESCRIPTION

[0059] Now, exemplary embodiments of the present invention are described with reference to the accompanying drawings. However, the present invention can be implemented in many different forms and is not limited to the embodiments described herein. These embodiments are provided to disclose the present invention in detail and completely and to fully convey the scope of the present invention to those skilled in the art. The terms used in the exemplary embodiments shown in the accompanying drawings are not intended to limit the present invention. In the accompanying drawings, the same units / elements are marked with the same reference numerals.

[0060] Unless otherwise specified, the terms (including technical terms) used herein have the commonly understood meanings to those skilled in the art. In addition, it is understood that the terms defined in commonly used dictionaries should be understood to have the same meanings as those in the context of the relevant fields, and should not be understood as idealized or overly formal meanings.

[0061] The present invention proposes a method for data fusion and perception collaborative processing, such as Figure 1 As shown, including:

[0062] Call the data of the information system, pre-clean the data, and use the cleaned data as the basic data for secondary identification of the cloud platform and store it in the secondary database of the cloud platform;

[0063] Perform multi-dimensional statistics on the basic data stored in the secondary database of the cloud platform, and perform multi-dimensional statistics on the secondary identified data stored in the secondary database of the cloud platform to obtain statistical information;

[0064] Obtain the frozen power information of electrical appliances within the preset time period of the information system, and use the cloud perception algorithm to assist in identifying the statistical information based on the statistical information, determine whether the electrical appliances have any abnormalities, and complete the fusion and perception recognition processing of the electrical appliance data.

[0065] Among them, data cleaning includes:

[0066] Complete the data and remove duplicate or abnormal data. At the same time, check for duplicates, omissions, and anomalies in the data of major electrical appliance categories and unknown electrical appliance categories and enter the additional feature information into the data;

[0067] The completion specifically includes checking for data omissions, triggering the pre-checking and freezing of data, and completing the data;

[0068] Duplicate data include: duplicate data of number, date or frozen period;

[0069] Abnormal data includes: negative power data, or data in which the comprehensive identified power exceeds the frozen power of the meter.

[0070] Among them, cleaning includes: real-time cleaning, on-demand cleaning or scheduled cleaning.

[0071] Among them, multi-dimensional statistics, specifically: statistics on data in terms of time dimension, appliance dimension, and geographical dimension of the region where the appliance is located;

[0072] The electrical appliance dimension includes the electrical appliance category dimension, the electrical appliance power dimension, and the additional characteristic information dimension of unknown electrical appliances and major categories of electrical appliances.

[0073] in,

[0074] Using cloud-based perception algorithms for auxiliary identification includes: Correcting statistical information, determining correct statistical information, and performing secondary identification on the correct statistical information, wherein the secondary identification includes:

[0075] Confirm the subdivision of electrical appliances in major categories and re-identify unknown electrical appliance types;

[0076] The additional cloud-based perception of unknown electrical appliances includes:

[0077] Report unknown appliance types and extract additional information including operating characteristics. Based on the additional information, perform secondary identification using machine learning or manual monitoring based on a massive feature library;

[0078] The secondary identification performs secondary labeling of the unknown electrical appliance type and the additional features, performs storage records in the module number dimension, and performs load identification of the unknown electrical appliance type corresponding to the additional features;

[0079] The confirmation of the subdivision of electrical appliances in the major category of electrical appliances includes:

[0080] Determine the additional characteristic information of the major categories of electrical appliances, report the additional characteristic information to the cloud platform, analyze the accompanying electrical appliances, power, short-term operating time and operating period information through the basic data in the secondary database of the cloud platform, find the longer-term operating rules of the electrical appliances through basic statistical data on the electrical appliances, power, short-term operating time and operating period information, and complete the subdivision confirmation of the major categories of electrical appliances;

[0081] The additional characteristic information includes: the number of times the electrical appliance starts and stops, the real 1-minute start and stop time information and the load event number information across the freezing period.

[0082] The present invention will be further described below in conjunction with embodiments:

[0083] First of all, the present invention is realized based on the cloud platform system of cloud collaboration. The system is integrated into the existing information system in the form of components, and relies on the existing information system to collect and store the identification data. After that, the data is analyzed and processed and stored again for the front-end effect display. The authentication mechanism is added. Only the load identification module that has passed the authorization authentication can obtain the SPI data sampled by the metering core, access the cloud platform system, and adapt the corresponding business layer business. The overall architecture of the system is as follows: Figure 2 As shown:

[0084] The specific process of implementing the present invention by the application system includes:

[0085] 1) Data cleaning, the process is as follows Figure 3 As shown:

[0086] Pre-process the data collected by the existing information system, including data completion and deletion of duplicate and abnormal data, and write the cleaned data into the cloud platform secondary database as the basic data for secondary identification of the cloud platform. Support real-time cleaning, historical scheduled cleaning, on-demand cleaning and other methods.

[0087] The specific process is as follows:

[0088] Collect load identification data in advance and store it in the original collection database;

[0089] The cleaning module performs preliminary processing on the original data by real-time cleaning or historical data cleaning. The processing includes: checking for data leaks, triggering pre-checking of frozen data; repeated data (number, date, frozen period); cleaning of abnormal data such as negative power, identifying that the total power far exceeds the frozen power information of the meter; refreshing the corresponding frozen electrical appliance types according to the cross-time load event number information and the last frozen electrical appliance type with the same continuous number;

[0090] After cleaning, the data will be entered into the cloud platform database as the basis for secondary identification. At the same time, the additional characteristic information attached to the electrical appliance category (non-detailed electrical appliance type data) and unknown electrical appliance category data will be checked for duplication, omissions, and anomalies and entered into the cloud platform secondary database;

[0091] Synchronize the corresponding frozen total power information in the usage data to the secondary database of the cloud platform;

[0092] Subsequent effects can be counted and displayed in multiple dimensions based on the processed data in the secondary database;

[0093] 2) Statistics:

[0094] It has two main functions. The first is to conduct statistics based on the basic data in the secondary database of the cloud platform, providing necessary statistical information for the secondary identification result correction and other functions in the cloud collaboration. The second is to conduct multi-dimensional statistics on the secondary identification data in the secondary data of the cloud platform, which is used as the data display source of the subsequent information system;

[0095] The specific process is as follows:

[0096] Statistics are conducted on the basic data in the secondary database of the cloud platform to provide dimensions such as year, month, and day, appliance category, power, unknown and additional feature information of major appliance categories, and statistical information on regional dimensions such as substations, communities, counties, provinces, and cities. This helps the cloud platform to analyze the load perception results based on regional electricity consumption patterns over a longer time span. For example, if a vacuum cleaner consumes 10 hours of electricity a day and consumes 12 kWh of electricity, this is obviously not in line with the electricity consumption characteristics of residential users.

[0097] Multi-dimensional statistics are performed on the data processed by all cloud load perception algorithms on the cloud platform, including statistical information on the year, month, and day time dimensions, appliance category dimensions, and substation, community, district, county, province, and city regional dimensions, for the final display of identification results.

[0098] 3) Data error identification, the process is as follows Figure 4 As shown:

[0099] Obtain the 15-minute frozen power information of the information system to determine the error situation where the total identified power exceeds the actual total power: For example, if the total identified power during the frozen period exceeds the total power during the frozen period (including the upper limit of the error), the identified data is unreliable and should be discarded;

[0100] Obtain secondary basic data from the cloud platform to determine if the electrical appliance and power are obviously inconsistent. For example, an electric water heater has an identified power of 80W, which is an obvious identification error.

[0101] Obtain secondary basic statistical data from the cloud platform to determine situations where the running time of electrical appliances is obviously abnormal. For example, a vacuum cleaner runs all night; or a large number of electrical appliances run very briefly in a short period of time (such as a freezing cycle);

[0102] Obtain secondary basic data from the cloud platform to determine situations where the number of electrical appliances is obviously abnormal, such as: during the first freezing cycle, three washing machines are running at the same time, or four microwave ovens are running at the same time.

[0103] Through the external auxiliary interface, seasonal temperature and other information are obtained and matched with secondary basic data and basic statistical data to identify obvious erroneous identification results, such as: the air conditioner is turned on for a long time when the temperature is around 18 to 20 degrees in spring and autumn.

[0104] At this point, the effects of the corresponding load identification meters, modules, and algorithms can be perceived based on the error identification situation and the identification data statistics before the additional load perception.

[0105] Assisted identification of cloud perception algorithms (optional application, providing an open cloud algorithm component interface):

[0106] After the data is corrected, the formal cloud algorithm process is carried out. The cloud platform cloud collaborative secondary identification mainly includes two parts of identification content, one is the confirmation of the electrical appliance subdivision of the major electrical appliance categories (non-detailed electrical appliances), and the other is the secondary identification of unknown electrical appliance types.

[0107] The cloud additional load perception process is as follows Figure 5 As shown;

[0108] Although some appliances are of the same type, their design principles are quite different, and their physical characteristics during use are quite different from other mainstream appliances. It is very difficult to cluster and distinguish them on the module side. At the same time, the module cannot provide massive feature resources for machine learning methods. Therefore, the load identification is advanced, and the identification of such appliances still needs to be supplemented by cloud resources.

[0109] For example, a hair dryer of a certain brand has many additional functions, and its actual operation characteristics are very different from those of ordinary mainstream hair dryers. The module side cannot identify it well through clustering. At this time, the module reports the unknown appliance type and needs to extract additional information including operation characteristics. Based on this additional auxiliary information, the cloud platform uses the high-end hardware conditions of the server to perform secondary identification based on machine learning methods or manual monitoring based on a massive feature library.

[0110] The cloud platform will re-identify the appliance and its features, record them in the database based on the module number, and push the appliance type corresponding to this feature to the corresponding module for subsequent load identification on the module side.

[0111] Confirmation of electrical appliance subdivisions of major electrical appliance categories (non-detailed electrical appliances):

[0112] For example, electric heating equipment has a single feature, and it is impossible to distinguish the content of the electrical appliances only from the physical feature quantity (reactive power, harmonics, etc.). It is necessary to make a comprehensive judgment based on more information (such as longer time cycle rules, seasons, time points, accompanying electrical appliance rules, power, etc.) before further analysis can be made. For example, the auxiliary heating of the washing machine is a purely resistive physical property. You cannot distinguish whether it is the auxiliary heating of the washing machine or another electric heating device is turned on when the washing machine is running just from the physical characteristics. For another example, a small kitchen treasure and a water dispenser, you cannot distinguish the obvious difference on the module side by just a few time periods.

[0113] At this time, the module needs to synchronously report additional characteristic information, such as the number of combined starts and stops of electrical appliances (description of the start and stop mechanism during the operation of electrical appliances), the actual 1-minute start and stop time information, and the load event number information across the freezing period, etc., and report to the cloud platform. The cloud-side collaboration analyzes the basic data in the secondary database of the cloud platform along with information such as electrical appliances, power, short-term operating time and operating time periods, and uses basic statistical data to find the longer-term operating rules of this appliance, as well as additional information such as the number of combined starts and stops (electrical appliance start and stop description information), 1-minute start and stop time, to judge the operating conditions of the appliance, and add other auxiliary information obtained from the external interface to further subdivide the actual detailed appliances of this large category of electric heating appliances.

[0114] The present invention also proposes a system 200 for data fusion and perception collaborative processing, such as Figure 6 As shown, including:

[0115] The cleaning module 201 calls the data of the information system, pre-cleans the data, and uses the cleaned data as the basic data for secondary identification of the cloud platform and stores it in the secondary database of the cloud platform;

[0116] The statistical module 202 performs multi-dimensional statistics on the basic data stored in the secondary database of the cloud platform, and performs multi-dimensional statistics on the secondary identified data stored in the secondary database of the cloud platform to obtain statistical information;

[0117] The identification module 203 obtains the frozen power information of the electrical appliances within a preset time period of the information system, and uses the cloud perception algorithm to assist in identifying the statistical information based on the statistical information, determines whether the electrical appliances are abnormal, and completes the fusion and perception of the electrical appliance data.

[0118] Among them, data cleaning includes:

[0119] Complete the data and remove duplicate or abnormal data. At the same time, check for duplicates, omissions, and anomalies in the data of major electrical appliance categories and unknown electrical appliance categories and enter the additional feature information into the data;

[0120] The completion specifically includes checking for data omissions, triggering the pre-checking and freezing of data, and completing the data;

[0121] Duplicate data include: duplicate data of number, date or frozen period;

[0122] Abnormal data includes: negative power data, or data in which the comprehensive identified power exceeds the frozen power of the meter.

[0123] Among them, cleaning includes: real-time cleaning, on-demand cleaning or scheduled cleaning.

[0124] Among them, multi-dimensional statistics, specifically: statistics on data in terms of time dimension, appliance dimension, and geographical dimension of the region where the appliance is located;

[0125] The electrical appliance dimension includes the electrical appliance category dimension, the electrical appliance power dimension, and the additional characteristic information dimension of unknown electrical appliances and major categories of electrical appliances.

[0126] in,

[0127] Using cloud-based perception algorithms for auxiliary identification includes: Correcting statistical information, determining correct statistical information, and performing secondary identification on the correct statistical information, wherein the secondary identification includes:

[0128] Confirm the subdivision of electrical appliances in major categories and re-identify unknown electrical appliance types;

[0129] The additional cloud-based perception of unknown electrical appliances includes:

[0130] Report unknown appliance types and extract additional information including operating characteristics. Based on the additional information, perform secondary identification using machine learning or manual monitoring based on a massive feature library;

[0131] The secondary identification performs secondary labeling of the unknown electrical appliance type and the additional features, performs storage records in the module number dimension, and performs load identification of the unknown electrical appliance type corresponding to the additional features;

[0132] The confirmation of the subdivision of electrical appliances in the major category of electrical appliances includes:

[0133] Determine the additional characteristic information of the major categories of electrical appliances, report the additional characteristic information to the cloud platform, analyze the accompanying electrical appliances, power, short-term operating time and operating period information through the basic data in the secondary database of the cloud platform, find the longer-term operating rules of the electrical appliances through basic statistical data on the electrical appliances, power, short-term operating time and operating period information, and complete the subdivision confirmation of the major categories of electrical appliances;

[0134] The additional characteristic information includes: the number of times the electrical appliance starts and stops, the real 1-minute start and stop time information and the load event number information across the freezing period.

[0135] The present invention provides a unified information collection system mode, realizes the fusion application of data, unifies the interface and mode details of the acquisition and load perception modules, provides a basis for the fusion processing of different algorithms of multiple manufacturers on the same cloud platform, and provides a method for algorithm operation and maintenance, evaluation, directional analysis and upgrading.

[0136] Those skilled in the art will appreciate that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of complete hardware embodiments, complete software embodiments, or embodiments in combination with software and hardware. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code. The scheme in the embodiments of the present application can be implemented in various computer languages, for example, object-oriented programming language Java and literal scripting language JavaScript, etc.

[0137] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0138] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0139] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process in the computer or other programmable device. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0140] Although the preferred embodiments of the present application have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0141] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. A method for data fusion and perception collaborative processing, the method comprising: Call the data of the information system, pre-clean the data, and use the cleaned data as the basic data for secondary identification of the cloud platform and store it in the secondary database of the cloud platform; Perform multi-dimensional statistics on the basic data stored in the secondary database of the cloud platform, and perform multi-dimensional statistics on the secondary identified data stored in the secondary database of the cloud platform to obtain statistical information; Obtain the frozen power information of the electrical appliances within the preset time period of the information system, and use the cloud perception algorithm to assist in identifying the statistical information based on the statistical information, determine whether the electrical appliances are abnormal, and complete the fusion and perception recognition processing of the electrical appliance data; The use of cloud-based perception algorithms for auxiliary identification includes: Correct the statistical information, determine the correct statistical information, and perform secondary identification on the correct statistical information, wherein the secondary identification includes: Confirm the subdivision of electrical appliances in major categories and perform additional cloud-based perception of unknown appliances; The additional cloud-based perception of unknown electrical appliances includes: Report unknown appliance types and extract additional information including operating characteristics. Based on the additional information, perform secondary identification using machine learning or manual monitoring based on a massive feature library; The secondary identification performs secondary labeling of the unknown electrical appliance type and the additional features, performs storage records in the module number dimension, and performs load identification of the unknown electrical appliance type corresponding to the additional features; The confirmation of the subdivision of electrical appliances in the major category of electrical appliances includes: Determine the additional characteristic information of the major categories of electrical appliances, report the additional characteristic information to the cloud platform, analyze the accompanying electrical appliances, power, short-term operating time and operating period information through the basic data in the secondary database of the cloud platform, find the longer-term operating rules of the electrical appliances through basic statistical data on the electrical appliances, power, short-term operating time and operating period information, and complete the subdivision confirmation of the major categories of electrical appliances; The additional characteristic information includes: the number of times the electrical appliance starts and stops, the real 1-minute start and stop time information and the load event number information across the freezing period.

2. According to the method of claim 1, the pre-cleaning of data specifically comprises: Complete the data and remove duplicate or abnormal data. At the same time, check for duplicates, omissions, and anomalies in the data of major electrical appliance categories and unknown electrical appliance categories and enter the additional feature information into the data; The completion specifically includes checking for leaks in the data, triggering a pre-check to freeze the data, and completing the data; The repeated data include: number, date or frozen period repeated data; The abnormal data includes: negative power data, or data in which the comprehensive identified power exceeds the frozen power of the meter.

3. The method according to claim 1, wherein the pre-cleaning comprises: Clean in real time, on demand or at a scheduled time.

4. According to the method of claim 1, the multi-dimensional statistics are specifically: statistics are performed on the data in terms of time dimension, appliance dimension, and region dimension of the region where the appliance is located; The electrical appliance dimension includes an electrical appliance category dimension, an electrical appliance power dimension, and an additional feature information dimension of unknown electrical appliances and major electrical appliance categories.

5. A system for data fusion and perception collaborative processing, the system comprising: The cleaning module calls the data of the information system, pre-cleans the data, and uses the cleaned data as the basic data for secondary identification of the cloud platform and stores it in the secondary database of the cloud platform; The statistical module performs multi-dimensional statistics on the basic data stored in the secondary database of the cloud platform, and performs multi-dimensional statistics on the secondary identified data stored in the secondary database of the cloud platform to obtain statistical information; The identification module obtains the frozen power information of the electrical appliances within the preset time period of the information system, and uses the cloud perception algorithm to assist in identifying the statistical information based on the statistical information, judge whether the electrical appliances are abnormal, and complete the fusion and perception of the electrical appliance data. The use of cloud-based perception algorithms for auxiliary identification includes: Correct the statistical information, determine the correct statistical information, and perform secondary identification on the correct statistical information, wherein the secondary identification includes: Confirm the subdivision of electrical appliances in major categories and perform additional cloud-based perception of unknown appliances; The additional cloud-based perception of unknown electrical appliances includes: Report unknown appliance types and extract additional information including operating characteristics. Based on the additional information, perform secondary identification using machine learning or manual monitoring based on a massive feature library; The secondary identification performs secondary labeling of the unknown electrical appliance type and the additional features, performs storage records in the module number dimension, and performs load identification of the unknown electrical appliance type corresponding to the additional features; The confirmation of the subdivision of electrical appliances in the major category of electrical appliances includes: Determine the additional characteristic information of the major categories of electrical appliances, report the additional characteristic information to the cloud platform, analyze the accompanying electrical appliances, power, short-term operating time and operating period information through the basic data in the secondary database of the cloud platform, find the longer-term operating rules of the electrical appliances through basic statistical data on the electrical appliances, power, short-term operating time and operating period information, and complete the subdivision confirmation of the major categories of electrical appliances; The additional characteristic information includes: the number of times the electrical appliance starts and stops, the real 1-minute start and stop time information and the load event number information across the freezing period.

6. According to the system of claim 5, the pre-cleaning of data specifically comprises: Complete the data and remove duplicate or abnormal data. At the same time, check for duplicates, omissions, and anomalies in the data of major electrical appliance categories and unknown electrical appliance categories and enter the additional feature information into the data; The completion specifically includes checking for leaks in the data, triggering a pre-check to freeze the data, and completing the data; The repeated data include: number, date or frozen period repeated data; The abnormal data includes: negative power data, or data in which the comprehensive identified power exceeds the frozen power of the meter.

7. The system of claim 5, wherein the pre-cleaning comprises: Clean in real time, on demand or at a scheduled time.

8. The system according to claim 5, wherein the multi-dimensional statistics are specifically: statistics are performed on the data in terms of time dimension, appliance dimension, and region dimension of the region where the appliance is located; The electrical appliance dimension includes an electrical appliance category dimension, an electrical appliance power dimension, and an additional feature information dimension of unknown electrical appliances and major electrical appliance categories.

Images