Power distribution district equipment identification method and device based on unsupervised deep learning

Abstract

The application provides a power distribution area equipment identification method and device based on unsupervised deep learning, and relates to the technical field of power distribution detection. The method comprises the following steps: collecting power consumption time sequence data in a preset time period at the secondary side of a main transformer in a power distribution area; extracting change feature data in the power consumption time sequence data based on a preset rule; inputting the change feature data into a preset convolutional neural network to extract feature maps of different scales; and transmitting the feature maps to a preset detection module and matching each power consumption equipment in the power distribution area to identify a target power consumption equipment associated with the power consumption time sequence data. The application realizes equipment identification in the power distribution area by making full use of existing voltage and current sensors without adding additional equipment identification hardware.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power distribution detection, and particularly relates to a power distribution area equipment identification method and device based on unsupervised deep learning. BACKGROUND

[0002] The smart grid system is composed of a master station, a concentrator, a collector and an electric energy meter. The power company often measures at the main transformer as the basis for collecting electric meter power information.

[0003] In recent years, in response to the call of the state to vigorously promote artificial intelligence, to improve the management level of power grid marketing, low-voltage transformer area centralized meter reading reconstruction projects have been gradually carried out in various places. However, due to the lack of strict control of construction quality, line connection faults occur from time to time, resulting in design errors of power distribution area. In addition, the line in some old urban areas is complex, and the transformer area maintenance is not perfect, and the meter replacement information is not complete, which also makes it difficult to identify the equipment in the power distribution area. SUMMARY

[0004] In view of the above problems, the present application provides a power distribution area equipment identification method and device based on unsupervised deep learning.

[0005] The first aspect of the present application provides a power distribution area equipment identification method based on unsupervised deep learning, comprising:

[0006] Collecting power consumption time series data in a preset time period at the secondary side of the main transformer in the power distribution area;

[0007] Extracting change feature data in the power consumption time series data based on a preset rule;

[0008] Inputting the change feature data into a preset convolutional neural network to extract feature maps of different scales;

[0009] Passing the feature maps to a preset detection module and matching each power consumption equipment in the power distribution area to identify the target power consumption equipment associated with the power consumption time series data.

[0010] The second aspect of the present application provides a power distribution area equipment identification device based on unsupervised deep learning, comprising:

[0011] A data acquisition module for collecting power consumption time series data in a preset time period at the secondary side of the main transformer in the power distribution area;

[0012] A change feature extraction module for extracting change feature data in the power consumption time series data based on a preset rule;

[0013] A feature map extraction module for inputting the change feature data into a preset convolutional neural network to extract feature maps of different scales;

[0014] An equipment identification module is configured to pass the feature map to a preset detection module to match each electrical equipment in the power distribution area, and identify the target electrical equipment associated with the electrical time sequence data.

[0015] The third aspect of the present application 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 the one or more processors, the one or more processors perform the above-mentioned power distribution area equipment identification method based on unsupervised deep learning.

[0016] Compared with the prior art, the power distribution area equipment identification method and device based on unsupervised deep learning provided by the present application have at least the following beneficial effects:

[0017] (1) Without increasing additional equipment identification hardware, the existing voltage and current sensors are used as much as possible to realize the equipment identification of the power distribution area;

[0018] (2) Through statistical analysis of the user's power consumption data, the relationship between the user load equipment and the main transformer in the power distribution area can be obtained, which has important theoretical research significance and practical application value;

[0019] (3) The important features of the current power distribution area equipment operation are obtained through unsupervised learning, and the interference of complex line conditions on the result is reduced;

[0020] (4) A large number of residual structure blocks are used for interlayer jumping to reduce storage space occupation, realize multi-layer feature fusion, and introduce multi-scale prediction to improve accuracy. BRIEF DESCRIPTION OF DRAWINGS

[0021] The above and other objects, features and advantages of the present application will become more apparent from the following description of embodiments of the present application, taken in conjunction with the accompanying drawings, in which:

[0022] Figure 1 The application scenario diagram of the power distribution area equipment identification method based on unsupervised deep learning according to the embodiment of the present application is schematically shown;

[0023] Figure 2 The flowchart of the power distribution area equipment identification method based on unsupervised deep learning according to the embodiment of the present application is schematically shown;

[0024] Figure 3 The lightweight CenterNet network structure diagram according to the embodiment of the present application is schematically shown;

[0025] Figure 4A structural block diagram of the power distribution district equipment identification device based on unsupervised deep learning according to an embodiment of the present application is shown schematically.

[0026] Figure 5 A block diagram of an electronic device suitable for implementing the power distribution district equipment identification method based on unsupervised deep learning according to an embodiment of the present application is shown schematically. DETAILED DESCRIPTION

[0027] To make the objectives, technical solutions and advantages of the present application clearer, further detailed description will be made to the present application with reference to the embodiments and the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work fall within the protection scope of the present application.

[0028] The terms used herein are merely used to describe specific embodiments, and are not intended to limit the present application. The terms "comprise", "contain" and the like used herein indicate the existence of the described features, steps, operations and / or components, but do not exclude the existence or addition of one or more other features, steps, operations or components.

[0029] All terms used herein (including technical and scientific terms) have 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 meanings consistent with the context of the present specification, and should not be interpreted in an idealized or overly formal manner.

[0030] Figure 1 An application scenario diagram of the power distribution district equipment identification method based on unsupervised deep learning according to an embodiment of the present application is shown schematically. It should be noted that, Figure 1 The shown is only an example of an application scenario to which the embodiments of the present application can be applied, to help those skilled in the art understand the technical content of the present application, but does not mean that the embodiments of the present application cannot be used in other devices, systems, environments or scenarios.

[0031] As Figure 1 shown, the application scenario according to the embodiment can be a power distribution district 100, which specifically includes a main transformer 101, a branch control cabinet 102, a user load device 103 and a voltage and / or current sensor 104.

[0032] The secondary side of the main transformer 101 is connected to a plurality of different user load devices 103 through a plurality of branch control cabinets 102. The branch control cabinet 102 is used to control the power load demand of each user load device 103, and can also adjust some electrical parameters of each user load device 103. The voltage and / or current sensor 104 is installed on the secondary side of the main transformer 101 to collect time series data.

[0033] It should be understood that Figure 1 The number of main transformers, branch control cabinets, user load devices, and voltage and / or current sensors in the above-mentioned application scenario is only illustrative. Any number of main transformers, branch control cabinets, user load devices, and voltage and / or current sensors can be provided according to the needs of implementation.

[0034] The method of the embodiment of the application will be described in detail below based on Figure 1 the application scenario described above. Figures 2-3

[0035] Figure 2 An illustrative flowchart of the power distribution substation device identification method based on unsupervised deep learning according to the embodiment of the application is shown.

[0036] As shown in Figure 2 , the power distribution substation device identification method based on unsupervised deep learning according to the embodiment can include operation S210 to operation S240.

[0037] In operation S210, power consumption time series data within a preset time period is collected at the secondary side of the main transformer of the power distribution substation.

[0038] In operation S220, change feature data in the power consumption time series data is extracted based on a preset rule.

[0039] In operation S230, the change feature data is input into a preset convolutional neural network to extract feature maps of different scales.

[0040] In operation S240, the feature maps are transmitted to a preset detection module and matched with each power consumption device in the power distribution substation to identify a target power consumption device associated with the power consumption time series data.

[0041] The main transformer can be, for example, the main transformer 101 shown in Figure 1 , and the power consumption device can be, for example, the user load device 103 shown in Figure 1 .

[0042] ​Through the above embodiment, the power distribution area equipment identification method based on unsupervised deep learning provided by the application can realize the equipment identification of the power distribution area without increasing additional equipment identification hardware, and try to use the existing voltage and current sensors. The application can obtain the relationship between the user load equipment and the main transformer in the power distribution area through statistical analysis of the user's power consumption data, and has important theoretical research significance and practical application value.

[0043] In the embodiment of the application, a plurality of power consumption equipment in the power distribution area is connected to the main transformer through at least one branch control cabinet; and the power consumption time series data in a preset time period is collected through the voltage and / or current sensor installed on the secondary side of the main transformer in the power distribution area.

[0044] In the embodiment of the application, the power consumption time series data includes voltage time series data and current time series data, and the change characteristic data includes voltage change characteristic data and current change characteristic data. On this basis, the change characteristic data in the power consumption time series data is extracted based on the preset rule in the operation S220, which can specifically include:

[0045] The voltage time series data is segmented in time series according to the time point of the sudden change;

[0046] The maximum difference of the voltage values in the adjacent two time series segments is taken as the voltage change characteristic data;

[0047] The two points where the voltage maximum difference is obtained are taken as the starting point and the ending point to intercept a segment of time series data, and the maximum and minimum values of the current values in the segment of time series data are calculated, and the current difference is taken as the current change characteristic data.

[0048] Specifically, the application segments and divides the collected power consumption time series data according to the voltage change amount, takes the maximum difference of the voltage values in the adjacent two time series segments as the voltage change characteristic data, and takes the two points where the voltage maximum difference is obtained as the starting point and the ending point to intercept a segment of time series data, and calculates the maximum and minimum values of the current values in the segment of time series data, and takes the current difference as the current change characteristic data. Therefore, the voltage change characteristic data and the current change characteristic data jointly constitute the change characteristic data.

[0049] Figure 3 The light CenterNet network structure diagram according to the embodiment of the application is schematically shown.

[0050] As shown in Figure 3 In the embodiment of the application, before the change characteristic data is input into the preset convolutional neural network, it further includes:

[0051] The power distribution area equipment identification model is constructed by using a lightweight CenterNet network structure, wherein the power distribution area equipment identification model comprises a preset convolutional neural network and a preset detection module.

[0052] It should be noted that the CenterNet network structure has the following characteristics: the target is regarded as a point, i.e., the center of the target bounding box. Thus, the target detection problem can be converted into other target attributes, and the direction and pose are both parameterized by estimating the center. The position of the target is determined by estimating the left upper corner and the right lower corner of the target pose. The CenterNet is similar to the anchor-based first-order method, in which the center point can be regarded as an anchor point without size. The important difference lies in that: (1) the CenterNet network structure is only placed at the position without size, and there is no need to manually set the threshold for distinguishing the foreground and the background; (2) there is only one positive anchor point for each target, and the key points are obtained from the local peak value of the feature map, so that the non-maximum suppression (NMS) algorithm is not needed later.

[0053] Further, based on the lightweight CenterNet network structure, the preset convolutional neural network in the power distribution area equipment identification model comprises a CBL component, a Res Unit component and a Res_N component.

[0054] The CBL component is the smallest component therein, which is composed of a convolution layer (Conv), a batch normalization layer (Batch Norm) and an activation function layer (Leak ReLU).

[0055] The Res Unit component is a residual structure composed of 2 CBL components and an add operation. Specifically, the initial input value of the unit is first copied to a temporary variable, then the input value is processed through 2 CBL components, and finally the output result is added (add) to the value of the temporary variable to obtain the final output value of the unit, realizing residual jump.

[0056] The Res_N structure block is composed of 1 CBL component and N Res Unit components, and comprises a convolution layer. N is a natural number, and the first CBL component in the Res_N structure block is set to have a convolution step of 2, thereby changing the feature map size and making it halved.

[0057] Based on the structure design of the above convolutional neural network, please refer to Figure 3 In the embodiment of the present application, the feature maps of different scales include a first scale feature map, a second scale feature map and a third scale feature map. In the operation S230, the changed feature data is input into the preset convolutional neural network to extract feature maps of different scales, which can specifically include the following steps:

[0058] The change feature data is processed by 1 CBL component, and then sequentially processed by a Res_1 structure block, a Res_2 structure block and a Res_8 structure block, at this time, a copy of the output value is copied and named as a first output value;

[0059] The first output value is processed by the Res_8 structure block, at this time, a copy of the output value is copied and named as a second output value;

[0060] The second output value is processed by the Res_4 structure block, at this time, a copy of the output value is copied and named as a third output value;

[0061] The third output value is processed by 5 CBL components, at this time, a copy of the output value is copied and named as a fourth output value, and the fourth output value is processed by 1 CBL component and 1 convolutional layer to obtain a first scale feature map;

[0062] The fourth output value is processed by 1 CBL component and up-sampling, and then concatenated with the second output value to obtain a first concatenation value, the first concatenation value is processed by 5 CBL components, at this time, a copy of the output value is copied and named as a fifth output value, and the fifth output value is processed by 1 CBL component and 1 convolutional layer to obtain a second scale feature map;

[0063] The fifth output value is processed by 1 CBL component and up-sampling, and then concatenated with the first output value to obtain a second concatenation value, and the second concatenation value is sequentially processed by 5 CBL components, 1 CBL component and 1 convolutional layer to obtain a third scale feature map.

[0064] The acquisition processes of the first output value, the second output value and the third output value together constitute a traditional Darknet-53 network, which can extract 8 times, 16 times and 32 times down-sampling features of the input feature map, respectively.

[0065] Through the above embodiment, the voltage and current time sequence change feature data is extracted by using the convolutional neural network, and the extracted feature map is transmitted to the detection module. Moreover, the storage space occupation is reduced by layer jumping through a large number of residual structure blocks, multi-layer feature fusion is realized, and the accuracy is improved by introducing multi-scale prediction.

[0066] Then, in the embodiment of the present application, the feature map is transmitted to the preset detection module and matched with each power utilization equipment in the power distribution area in the operation S240, which can specifically include:

[0067] The first scale feature map, the second scale feature map and the third scale feature map are input into the detection module, 3 prior boxes of different sizes are generated for each scale feature map, and a detection box is obtained by decoding each prior box;

[0068] According to the preset division parameter, the feature map of each scale is divided into a plurality of grid cells;

[0069] For each grid cell, according to a preset evaluation index threshold, the detection frame is detected to obtain a detection result, and the detection result includes a detection frame position, a detection confidence and a category.

[0070] The detection frame position may include, for example, the center point coordinates, width and height of the target bounding box.

[0071] Specifically, after the time-varying feature data passes through the convolutional neural network, three feature maps of different scales are obtained, 3 prior boxes of different sizes are generated for each feature map, and the detection frame is obtained by decoding the prior box. In the inference process, the lightweight CenterNet network structure divides the input data into SxS grid cells for detection according to the size of the last feature map. The detection module uses the convolutional neural network to obtain the result through unsupervised learning. A detection confidence function is set at the output result of the detection module, and a threshold is set for the evaluation index. Considering the lightweight processing of the network, all output results below the evaluation index threshold are filtered out, and then the one with the highest evaluation index is selected as the final output. The center point coordinates, width and height of the target bounding box, and the confidence score are calculated.

[0072] Through the embodiment of the present application, the important features of the current power distribution area equipment operation are obtained through unsupervised learning, and the interference of complex line conditions on the result is reduced.

[0073] Based on the above disclosed method, the present application further provides a power distribution area equipment identification device based on unsupervised deep learning, which will be described in detail below Figure 4 The device is described in detail.

[0074] Figure 4 The structure block diagram of the power distribution area equipment identification device based on unsupervised deep learning according to the embodiment of the present application is schematically shown.

[0075] As Figure 4 shown, the power distribution area equipment identification device based on unsupervised deep learning 400 according to the embodiment includes a data acquisition module 410, a change feature extraction module 420, a feature map extraction module 430 and an equipment identification module 440.

[0076] The data acquisition module 410 is used to acquire power time series data in a preset time period at the secondary side of the main transformer of the power distribution area.

[0077] The change feature extraction module 420 is used to extract change feature data in the power time series data based on a preset rule.

[0078] The feature map extraction module 430 is configured to input the change feature data into a preset convolutional neural network, and extract feature maps of different scales.

[0079] The device identification module 440 is configured to transmit the feature maps to a preset detection module, and match the feature maps with each electrical device in the power distribution area, so as to identify the target electrical device associated with the electrical time sequence data.

[0080] It should be noted that the embodiments of the device part are similar to the embodiments of the method part, and the technical effects achieved are also similar. For specific details, please refer to the above method embodiment part, which will not be repeated here.

[0081] According to the embodiments of the present application, any multiple of the data acquisition module 410, the change feature extraction module 420, the feature map extraction module 430 and the device identification module 440 can be combined in one module, or any one of the modules can be split into multiple modules. Alternatively, at least part of the function of one or more of these modules can be combined with at least part of the function of other modules, and implemented in one module. According to the embodiments of the present application, at least one of the data acquisition module 410, the change feature extraction module 420, the feature map extraction module 430 and the device identification module 440 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 chip, a system on board, a system on package, an application specific integrated circuit (ASIC), or any other reasonable way of integrating or packaging a circuit, etc. hardware or firmware, or any one of software, hardware and firmware or any appropriate combination of several of them. Alternatively, at least one of the data acquisition module 410, the change feature extraction module 420, the feature map extraction module 430 and the device identification module 440 can be at least partially implemented as a computer program module which can perform corresponding functions when running.

[0082] Figure 5 The block diagram of the electronic device suitable for implementing the power distribution area device identification method based on unsupervised deep learning according to the embodiments of the present application is schematically shown.

[0083] As Figure 5As shown, the electronic device 500 according to an embodiment of the present application includes a processor 501 which can perform various appropriate actions and processes in accordance with a program stored in a read only memory (ROM) 502 or a program loaded into a random access memory (RAM) 503 from a storage section 508. The processor 501 can 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)), and so on. The processor 501 can also include an on-board memory for cache use. The processor 501 can include a single processing unit or multiple processing units to perform the various actions of the method processes according to embodiments of the present application.

[0084] In the RAM 503, various programs and data required for the operation of the electronic device 500 are stored. The processor 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. The processor 501 performs various operations of the method processes according to embodiments of the present application by executing the programs in the ROM 502 and / or the RAM 503. Note that the programs can also be stored in one or more memories other than the ROM 502 and the RAM 503. The processor 501 can also perform various operations of the method processes according to embodiments of the present application by executing the programs stored in the one or more memories.

[0085] According to an embodiment of the present application, the electronic device 500 can also include an input / output (I / O) interface 505 which is also connected to the bus 504. The electronic device 500 can also include one or more of the following components connected to the I / O interface 505: an input section 506 including a keyboard, a mouse, etc.; an output section 507 including a display such as a cathode ray tube (CRT), a liquid crystal display (LCD), etc., and a speaker, etc.; a storage section 508 including a hard disk, etc.; and a communication section 509 including a network interface card such as a LAN card, a modem, etc. The communication section 509 performs communication processing via a network such as the Internet. A drive 510 is also connected to the I / O interface 505 as necessary. A removable recording medium 511 such as a magnetic disk, an optical disk, a magneto-optical disk, a semiconductor memory, etc. is attached to the drive 510 as necessary, so that a computer program read out therefrom is installed in the storage section 508 as necessary.

[0086] Some of the blocks and / or flowcharts in the drawings 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, or other programmable data processing apparatus, so that the instructions can create means for implementing the functions / operations specified in the block diagrams and / or flowcharts.

[0087] In addition, the terms "first", "second", etc. are used only for the purpose of description, and should not be understood as indicating or implying relative importance or implying the number of the technical features indicated. Therefore, the features defined with "first", "second" can explicitly or implicitly include one or more of the features. In the description of the present application, the meaning of "a plurality of" is at least two, such as two, three, etc., unless otherwise specifically limited. In addition, the word "one" or "an" before an element does not exclude the presence of multiple such elements.

[0088] The above-described specific embodiments further illustrate the purpose, technical solutions and beneficial effects of the present application. It should be understood that the above-described is only a specific embodiment of the present application and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

Claims

1. A power distribution feeder identification method based on unsupervised deep learning, characterized in that, The application relates to a power distribution area change feature data extraction method and device. Collecting power consumption time sequence data in a preset time period at the secondary side of a main transformer in a power distribution area; Extracting change feature data in the power consumption time sequence data based on a preset rule; Inputting the change feature data into a preset convolutional neural network to extract feature maps of different scales; Transferring the feature maps to a preset detection module and matching with each power consumption equipment in the power distribution area to identify target power consumption equipment associated with the power consumption time sequence data; The power consumption time sequence data includes voltage time sequence data and current time sequence data, and the change feature data includes voltage change feature data and current change feature data; The method comprises the following steps: According to the time point of the voltage time sequence data mutation, the time sequence is segmented; Taking the maximum difference of voltage values in adjacent two time sequence segments as voltage change feature data; Taking the two points with the maximum voltage difference as the starting point and the ending point to intercept a segment of time sequence data, and taking the maximum and minimum values of current values in the segment of time sequence data as current change feature data; The method comprises the following steps: After the change feature data is processed by a CBL component once, the change feature data is processed by a Res_1 structural block, a Res_2 structural block and a Res_8 structural block in sequence, at this time, a copy of the output value is copied and named as a first output value; The first output value is processed by a Res_8 structural block, at this time, a copy of the output value is copied and named as a second output value; The second output value is processed by a Res_4 structural block, at this time, a copy of the output value is copied and named as a third output value; The third output value is processed by a CBL component for five times, at this time, a copy of the output value is copied and named as a fourth output value, and the fourth output value is processed by a CBL component once and a convolutional layer once to obtain a first scale feature map; The fourth output value is processed by a CBL component once and up-sampling, and then spliced with the second output value to obtain a first splicing value, and the first splicing value is processed by a CBL component for five times, at this time, a copy of the output value is copied and named as a fifth output value, and the fifth output value is processed by a CBL component once and a convolutional layer once to obtain a second scale feature map; The fifth output value is processed by a CBL component once and up-sampling, and then spliced with the first output value to obtain a second splicing value, and the second splicing value is processed by a CBL component for five times, a CBL component once and a convolutional layer once in sequence to obtain a third scale feature map.

2. The power distribution feeder segment asset identification method based on unsupervised deep learning according to claim 1, wherein, The plurality of power consumption equipments in the power distribution area are connected to the main transformer through at least one branch control cabinet; The power consumption time sequence data in a preset time period is collected through a voltage and / or current sensor installed at the secondary side of the main transformer in the power distribution area.

3. The power distribution feeder segment asset identification method based on unsupervised deep learning of claim 1, wherein, Before the change feature data is input into the preset convolutional neural network, the method further comprises the following steps: The power distribution area equipment identification model is constructed by using a lightweight CenterNet network structure, wherein the power distribution area equipment identification model comprises the preset convolutional neural network and a preset detection module.

4. The power distribution feeder segment asset identification method based on unsupervised deep learning according to claim 3, wherein, The preset convolutional neural network comprises a CBL component, a Res Unit component and a Res_N component. The CBL component is composed of a convolutional layer, a batch normalization layer and an activation function layer. The Res Unit component is a residual structure composed of two CBL components and an add operation. The Res_N structure block is composed of one CBL component and N Res Unit components, N is a natural number, and the first CBL component in the Res_N structure block is set to have a convolution step of 2.

5. The unsupervised deep learning based distribution feeder asset identification method of claim 1, wherein, The feature map is transmitted to the preset detection module and matched with each electrical equipment in the power distribution area, specifically including: The first scale feature map, the second scale feature map and the third scale feature map are input into the detection module, 3 prior boxes of different sizes are generated for each scale feature map, and a detection box is obtained by decoding each prior box. According to a preset division parameter, each scale feature map is divided into a plurality of grid units. For each grid unit, according to a preset evaluation index threshold, the detection box is detected to obtain a detection result, and the detection result comprises a detection box position, a detection confidence and a category.

6. The power distribution feeder segment asset identification method based on unsupervised deep learning according to claim 5, wherein, The detection box position comprises a center point coordinate, a width and a height of the target bounding box.

7. An apparatus for power distribution feeder identification based on unsupervised deep learning, applied to the method of any one of claims 1-6, characterized in that, The data acquisition module is configured to acquire power consumption time series data in a preset time period at a secondary side of a main transformer in a power distribution area. The change feature extraction module is configured to extract change feature data in the power consumption time series data based on a preset rule. The feature map extraction module is configured to input the change feature data into a preset convolutional neural network to extract feature maps of different scales. The equipment identification module is configured to transmit the feature maps to a preset detection module and match each electrical equipment in the power distribution area to identify a target electrical equipment associated with the power consumption time series data. The one or more processors are configured to execute the one or more programs to perform the method according to any one of claims 1-6.

8. An electronic device, comprising: The one or more processors are configured to execute the one or more programs to perform the method according to any one of claims 1-6. ​ ​ ​

Images