Multi-mode environment monitoring expansion adapter and power distribution room working condition early warning method

Abstract

The invention relates to a multi-mode environment monitoring expansion adapter and a power distribution room working condition early warning method, and the device comprises a core control board which is provided with a plurality of types of external interfaces and is used for being in communication connection with measurement and control edge equipment; the first JK control board is in communication connection with the core control board and is used for acquiring outgoing line voltage, outgoing line current and incoming line voltage; the second JK control board is in communication connection with the core control board and is used for collecting power supply voltage and power supply current; the upper computer is used for receiving the power distribution room working condition data and completing remote monitoring and early warning of the community power distribution room according to the power distribution room working condition data. Communication ports are expanded by utilizing the core control board, the problem of insufficient communication network ports of an upper computer is solved, the information monitoring types of the power distribution room are enriched, meanwhile, the safety state of the power distribution room can be monitored and automatically identified in real time through the working condition data of the power distribution room, electric power workers can find potential safety hazards in time, and the safety of the power distribution room is improved. Therefore, the risk of accidents is reduced.

Description

Technical Field

[0001] The present invention relates to the field of power distribution room safety technology, and in particular to a multi-modal environment monitoring expansion adapter and a power distribution room working condition early warning method. Background Art

[0002] The community power distribution room is the terminal link of the power grid. The level of its management directly affects the power supply capacity and quality of the community, and is related to the operation level and social image of the power grid company. However, for a long time, the management reliability of community power distribution rooms has always been a weak link. Traditional power distribution rooms do not support online monitoring and remote intelligent operation and maintenance. They are mainly based on pure manual inspections, which are large in quantity, scattered and inefficient. Usually, they can only be handled urgently after an accident occurs, causing great economic losses and customer complaints. In addition, since the power distribution room was connected to the host computer, the information collection capability of the power distribution room has been improved to a certain extent. However, due to the excessive access to monitoring information, the system faces a large number of protocol conversion tasks, high computing pressure, long-term high-power operation, severe heat generation, unstable operation, and frequent crashes. Summary of the invention

[0003] To solve the above problems, an embodiment of the present invention aims to provide a multi-modal environment monitoring expansion adapter and a distribution room working condition early warning method.

[0004] A multi-modal environment monitoring expansion adapter, comprising:

[0005] A core control board, wherein the core control board is provided with various types of external interfaces for communicating with the measurement and control edge device; the measurement and control edge device is used to collect the working condition data of the power distribution room;

[0006] A first JK control board is connected to the core control board for collecting outgoing line voltage, outgoing line current and incoming line voltage;

[0007] A second JK control board is connected to the core control board for collecting power supply voltage and power supply current;

[0008] The communication circuit board is respectively connected to the core control board and the host computer for communication; the host computer is used to receive the working condition data of the distribution room, and complete the remote monitoring and early warning of the community distribution room according to the working condition data of the distribution room.

[0009] Preferably, the core control board is used to uniformly convert the distribution room operating condition data collected by different measurement and control edge devices into IEC 104 protocol data using a protocol conversion model, and send it to the host computer.

[0010] Preferably, the external interface types of the core control board include: RS485, RS232, CAN, Ethernet and USB.

[0011] The present invention also provides a distribution room working condition early warning method, which is applied to a multi-modal environment monitoring expansion adapter, comprising:

[0012] Step 1: Obtain monitoring images of the working conditions of the community distribution room;

[0013] Step 2: Performing illumination balance on the monitoring image to obtain an illumination balanced monitoring image;

[0014] Step 3: Annotate the surveillance images to distinguish whether the community distribution room is safe;

[0015] Step 4: Input the labeled training samples into the deep learning network for training to obtain the community distribution room warning model;

[0016] Step 5: Use the community distribution room early warning model to complete the identification of the unsafe state of the target community distribution room and issue an early warning.

[0017] Preferably, the step 2: performing illumination balancing on the monitoring image to obtain the illumination-balanced monitoring image comprises:

[0018] Step 2.1: Convert the monitoring image into HSV space for contrast enhancement to form a contrast enhanced image;

[0019] Step 2.2: construct an illumination balance window with any point as the midpoint from the contrast-enhanced image;

[0020] Step 2.3: Calculate the mean value of the pixels in the light balance window;

[0021] Step 2.4: Construct the light balance coefficient based on the mean value of the pixels in the light balance window;

[0022] Step 2.5: Performing illumination balance on the monitoring image using the illumination balance coefficient to obtain an illumination-balanced monitoring image.

[0023] Preferably, in step 2.4, the light balance coefficient is:

[0024]

[0025] Among them, R(x,y) is the light balance coefficient, n×n is the size of the light balance coefficient W with point (i,j) as the midpoint, and v(i,j) is the brightness value of point (i,j). is the average brightness value under the light balance coefficient window W, s(i,j) is the saturation of point (i,j), is the average saturation value under the light balance coefficient window W, δ v (x,y) is the brightness value variance, δ s(x,y) is the saturation variance.

[0026] Preferably, the step 2.5: performing illumination balance on the monitoring image using the illumination balance coefficient to obtain the illumination-balanced monitoring image comprises:

[0027] A light balance formula is constructed based on the set coefficients, and the saturation of the monitoring image is processed using the light balance formula to obtain a standardized monitoring image; wherein the standardized processing formula is:

[0028] s′(x,y)=s(x,y)+t(v′(x,y)-v(x,y))×R(x,y)

[0029] Among them, s′(x,y) is the saturation of the standardized processing, and t is the balance coefficient;

[0030] The standardized monitoring image is subjected to denoising to obtain a monitoring image after light balance.

[0031] Preferably, the step 4: inputting the labeled training samples into the deep learning network for training to obtain the community distribution room early warning model includes:

[0032] The samples are input into the deep learning network for training, and the deep learning network is optimized using the loss function. When the value of the loss function is less than a preset value, the community distribution room early warning model is output.

[0033] Preferably, in step 4, the loss function is:

[0034]

[0035] Among them, Loss is the value of the loss function, n is the number of samples, y i is the actual label of the i-th sample, The labels predicted by the deep learning network.

[0036] The present invention also provides a computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps in the above-mentioned multimodal environment monitoring expansion adapter are implemented.

[0037] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0038] The present invention relates to a multimodal environment monitoring expansion adapter and a distribution room working condition early warning method. Compared with the prior art, the present invention utilizes a core control board to expand the communication port, solves the problem of insufficient communication network ports of a host computer, and enriches the types of information monitoring in the distribution room. At the same time, through the early warning model of the community distribution room, the safety status of the distribution room can be monitored and automatically identified in real time, so that power workers can discover potential safety hazards in time, thereby reducing the risk of accidents.

[0039] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0041] Figure 1 A schematic diagram of a multi-modal environment monitoring expansion adapter in an embodiment of the present invention;

[0042] Figure 2 A schematic diagram of the external interface design of the adapter in the embodiment provided by the present invention;

[0043] Figure 3 A schematic diagram of a protocol conversion model in an embodiment of the present invention;

[0044] Figure 4 This is a flow chart of the distribution room working condition early warning method in the embodiment provided by the present invention. DETAILED DESCRIPTION

[0045] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0046] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0047] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0048] See also Figure 1 , a multi-modal environment monitoring expansion adapter, comprising:

[0049] A core control board, wherein the core control board is provided with various types of external interfaces for communicating with the measurement and control edge device; the measurement and control edge device is used to collect the working condition data of the power distribution room;

[0050] A first JK control board is connected to the core control board for collecting outgoing line voltage, outgoing line current and incoming line voltage;

[0051] A second JK control board is connected to the core control board for collecting power supply voltage and power supply current;

[0052] The communication circuit board is respectively connected to the core control board and the host computer for communication; the host computer is used to receive the working condition data of the distribution room, and complete the remote monitoring and early warning of the community distribution room according to the working condition data of the distribution room.

[0053] Furthermore, the overall hardware panels of the present invention are divided into: LED board, ZK board (core control board), JK1 board (first JK control board), JK2 board (second JK control board), ZB board (communication circuit board). Among them, the LED board and ZK board are fixed on the upper casing, and the ZK board and JK1 board are connected through a flexible flat cable. The JK1 board is mounted on the JK2 board and fixed on the lower casing. The JK1 and JK2 boards are connected through pin headers and fixed with two 23.5mm copper columns. The LED board adopts a 2-layer board design to realize OLED display, buttons, indicator lights, and USB interface functions. The ZK board adopts a 4-layer board design, uses the Feiling OKMX6ULL-S core board, runs at 800MHz, is equipped with 256MB RAM and 256MB NAND FLASH (supports SD expansion), and realizes 1 232 maintenance, 232*1, 485*2, network port*4, CAN, GPS, Bluetooth (reserved), encryption, 4G / 5G interface, carrier interface, USB port, OLED display, LED indicator, button, remote control, and 4-way remote signaling functions. The JK1 board adopts a 4-layer board design and uses the STM32F103ZET6 microcontroller: realize outgoing line voltage and current sampling, realize protection function, incoming line voltage sampling (anti-islanding design); 3-way remote signaling acquisition, protection output control, and expansion of 4-way 485. The JK2 board adopts a 2-layer board design to realize remote signaling and remote control, power supply, voltage, and current sampling. The voltage sampling ports are divided into the incoming line UA, UB, UC (measurement), the outgoing line UA, UB, UC, U0 (protection), and the current sampling ports are divided into IA, IB, IC (measurement), IA, IB, IC, I0 (protection). The overall hardware solution block diagram is as follows Figure 2 The power module inputs a rated AC voltage of 220V, which is converted to DC 12V and 5V output by the ACDC module. The JK2 board converts DC 12V to DC 3.3V and 5V. The power module principle diagram is shown in the figure. Figure 1 shown.

[0054] Expand the external interface of the adapter design such as Figure 2 As shown in Table 1, the communication network ports such as RS485, RS232, CAN, and Ethernet are expanded to realize multi-channel information collection and solve the defect of insufficient communication network ports of the host computer.

[0055] Table 1 Terminal external interface design details

[0056]

[0057] The software architecture of the protocol conversion model of the present invention is based on APP deployment as the core, which can meet the IoT needs of the "cloud-pipe-edge-end" architecture, and can quickly deploy the master station software to achieve observable, measurable, controllable and adjustable operations on the equipment. The software is APP-based, supports APP deployment based on the data center, supports the power distribution Internet of Things interface, takes the MQTT bus as the base, supports the simultaneous deployment of IEC101, IEC103, IEC104, MQTT, and MODBUS master station protocols and completes the conversion between the above protocols. Among them, the e9361APP integrates the gateway function and various commonly used protocols in the distribution room to assist the host computer in completing the conversion between commonly used protocols such as IEC101, IEC103, IEC104, and MODBUS. Generally, the data stream is uniformly converted into 104 protocols and sent to the host computer, which effectively reduces the running memory of the host computer CPU, avoids the serious heating phenomenon caused by the large computing pressure of the host computer, and makes the host computer run more stably. The integrated gateway function can save routing equipment in the distribution room renovation project and save capital costs.

[0058] Furthermore, in an embodiment of the present invention, since the multimodal monitoring of the upper computer needs to access a variety of measurement and control edge devices, and the models of different types of devices are very different, it is necessary to study a unified protocol conversion model to convert various device models into a unified device model, so that various heterogeneous data are converted into homogeneous data, which is convenient for the forwarding module to perform unified processing. The communication management devices in the power communication system can generally be divided into telesignaling, telemetering, telepulse, remote control, environmental dynamics and other information. First, the information models of different devices are constructed, and these models are stored in the real-time database according to relevant protocols. The point number of the IEC103 protocol device is used to uniquely identify a point by the function type and information serial number. The IEC101 and IEC104 protocol devices distinguish different information points according to the address of the information object. The MODBUS protocol distinguishes different information points according to different register addresses. The unified protocol conversion model is such as Figure 3As shown. Subsequently, the message from the device is received and parsed, and the data frame of the message source protocol is parsed in the Ethernet environment to extract key information (such as device address, information sequence number, control code, etc.), and the parsing results are stored in the real-time database. Next, the data mapping operation is performed, that is, the result data is extracted and parsed from the database, and the specific path of data mapping is determined by querying the mapping file. According to the rules of the established target protocol, these data are mapped to realize the conversion of the source protocol data into the data format of the target protocol. The protocol conversion has multi-protocol support function. The strategy of the data mapping operation is to create and configure the mapping file, which acts as a link between the traditional device communication point table and the information model. Finally, the converted data is encapsulated according to the frame format of the target protocol to generate a complete data frame, and the encapsulated data frame is transmitted to the target device through the specified communication interface. During the transmission process, data verification is required to ensure the integrity and correctness of the data.

[0059] By configuring the mapping file, the present invention can describe how the data of traditional devices is mapped to the information model of the unified protocol conversion, thereby achieving the correct correspondence and mapping of the data. The configuration file is intended to guide the system to convert the data meaning in the traditional protocol frame structure into the smallest reference unit with hierarchical attributes in the unified protocol conversion model. This process is called "finding the mapping object". To ensure accurate search, a unique identifier must be used. Taking the 104 protocol as an example, its data model usually includes five elements: device address, information sequence number, value, quality and time stamp. The combination of the information body address (information sequence number) and the public information address (device address) is the key link in the conversion process. The mapping configuration file can be used to efficiently locate the main protocol data reference to be converted in the database and realize reverse query. After statistical analysis, an information body address model can be established in the protocol conversion model for each information point. Different types of protocol models can be converted by appropriate methods, so that they can be unified into the information model of the protocol converter. For example, the function type and information serial number of the information point of the IEC103 device each occupy one byte. By shifting the function type value left by 8 bits and adding it to the information serial number, the information body address can be formed, thus completing the mapping between the IEC103 model and the unified model library. When organizing the downlink message, the function type and information serial number are extracted from the information body address using the reverse method, and then it can be organized into a 103 message for transmission. Similarly, other protocols can be processed similarly to unify them into the information model of the protocol converter.

[0060] The host computer directly reads the intelligent distribution transformer terminal data through the RS232 / 485 interface of the intelligent gateway to calculate the three-phase unbalance parameters.

[0061] Daily three-phase unbalance calculation: When the three-phase unbalance of each T min section value of the distribution transformer with Yyn0 connection is above 15%, and the three-phase unbalance of each T min section value of the distribution transformer with Dyn11 connection is above 25%, and the number of occurrences is not less than 4 times (1h) on the same day, the three-phase load of the distribution transformer on that day is judged to be unbalanced (for example, if the ledger cannot distinguish between Yyn0 connection and Dyn11 connection, both are judged as 15%). The formula is as follows:

[0062] Three-phase unbalance = 1-minimum value of three-phase current / maximum value of three-phase current

[0063] Monthly three-phase unbalance calculation: Three-phase unbalance occurs for 5 days or more in a natural month and the monthly average load rate of the distribution transformer is above 20%. To calculate the monthly average three-phase unbalance, the daily average three-phase unbalance of the month is averaged, and the formula is as follows:

[0064] Monthly average three-phase imbalance of distribution transformer = ∑(daily average three-phase imbalance) / number of days

[0065] The present invention uses the core control board to expand the communication port, solves the problem of insufficient communication network ports of the host computer, and enriches the types of information monitoring in the power distribution room. In addition, the present invention assists the host computer in completing protocol conversion, shares the computing pressure, effectively reduces the CPU memory of the host computer, and reduces the heating phenomenon of the mainboard.

[0066] See also Figure 4 The present invention also provides a distribution room working condition early warning method, which is applied to a multi-modal environment monitoring expansion adapter, including:

[0067] Step 1: Obtain monitoring images of the working conditions of the community distribution room;

[0068] Step 2: Performing illumination balance on the monitoring image to obtain an illumination balanced monitoring image;

[0069] Furthermore, the step 2 comprises:

[0070] Step 2.1: Convert the monitoring image into HSV space for contrast enhancement to form a contrast enhanced image;

[0071] Since the vividness of color in HSV space is determined by the color and brightness, the change of brightness value (V component) will affect the color sense of the image, because there is a relative relationship between brightness and saturation (S component). When the brightness increases, if the saturation is not adjusted adaptively, it means that the same color is mixed with more brightness, and the richness of the color will be relatively reduced. Therefore, after the brightness V is enhanced, if the saturation S component is not further adjusted accordingly, the saturation of the image will be reduced to a certain extent. In order to maintain the color balance of the image and the naturalness of the visual effect, after the brightness value of the image V component is processed, in order to ensure the natural transition of the color, it is usually necessary to adaptively adjust the saturation (S component).

[0072] The process is as follows:

[0073] Step 2.2: construct an illumination balance window with any point as the midpoint from the contrast-enhanced image;

[0074] Step 2.3: Calculate the mean value of the pixels in the light balance window;

[0075] Step 2.4: Construct the light balance coefficient based on the mean value of the pixels in the light balance window;

[0076] In step 2.4, the light balance coefficient is:

[0077]

[0078] Among them, R(x,y) is the light balance coefficient, n×n is the size of the light balance coefficient W with point (i,j) as the midpoint, and v(i,j) is the brightness value of point (i,j). is the average brightness value under the light balance coefficient window W, s(i,j) is the saturation of point (i,j), is the average saturation value under the light balance coefficient window W, δ v (x,y) is the brightness value variance, δ s (x,y) is the saturation variance.

[0079] Step 2.5: Use the light balance coefficient to perform light balance on the monitoring image to obtain a light balanced monitoring image:

[0080] Furthermore, step 2.5 includes:

[0081] A light balance formula is constructed based on the set coefficients, and the saturation of the monitoring image is processed using the light balance formula to obtain a standardized monitoring image; wherein the standardized processing formula is:

[0082] s′(x,y)=s(x,y)+t(v′(x,y)-v(x,y))×R(x,y)

[0083] Among them, s′(x,y) is the saturation of the standardized processing, and t is the balance coefficient;

[0084] By adjusting the saturation appropriately, you can ensure that the color transitions between multiple brightness levels are more natural and avoid color distortion caused by lighting changes.

[0085] Some noise points on the enhanced monitoring image may be amplified. In order to improve the visual quality of the image, it is necessary to suppress the noise of the image. Therefore, the present invention needs to perform denoising on the standardized monitoring image to obtain a monitoring image after light balance.

[0086] Step 3: Annotate the surveillance images to distinguish whether the community distribution room is safe;

[0087] Step 4: Input the labeled training samples into the deep learning network for training to obtain the community distribution room warning model;

[0088] Furthermore, step 4 includes:

[0089] The samples are input into the deep learning network for training, and the deep learning network is optimized using the loss function. When the value of the loss function is less than the preset value, the early warning model of the community distribution room is output; wherein the loss function is:

[0090]

[0091] Among them, Loss is the value of the loss function, n is the number of samples, yi is the actual label of the i-th sample, The labels predicted by the deep learning network.

[0092] Step 5: Use the community distribution room early warning model to complete the identification of the unsafe state of the target community distribution room and issue an early warning.

[0093] The present invention can monitor and automatically identify the safety status of the distribution room in real time through the early warning model of the community distribution room, so that power workers can discover potential safety hazards in time, thereby reducing the risk of accidents.

[0094] The present invention also provides a computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps of the above-mentioned method for early warning of operating conditions in a distribution room are implemented. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by the present invention are the same as the beneficial effects of the method for early warning of operating conditions in a distribution room described in the above-mentioned technical scheme, which will not be repeated here.

[0095] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technical solution that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A multi-modal environment monitoring expansion adapter, characterized in that: include: A core control board, wherein the core control board is provided with various types of external interfaces for communicating with the measurement and control edge device; the measurement and control edge device is used to collect the working condition data of the power distribution room; A first JK control board is connected to the core control board for collecting outgoing line voltage, outgoing line current and incoming line voltage; A second JK control board is connected to the core control board for collecting power supply voltage and power supply current; The communication circuit board is respectively connected to the core control board and the host computer for communication; the host computer is used to receive the working condition data of the distribution room, and complete the remote monitoring and early warning of the community distribution room according to the working condition data of the distribution room.

2. The multimodal environment monitoring expansion adapter according to claim 1, characterized in that: The core control board is used to convert the distribution room operating data collected by different measurement and control edge devices into IEC 104 protocol data by using the protocol conversion model, and send it to the host computer.

3. A multi-modal environment monitoring expansion adapter according to claim 2, characterized in that: The external interface types of the core control board include: RS485, RS232, CAN, Ethernet and USB.

4. A distribution room working condition early warning method, characterized in that: The method is applied to a multimodal environment monitoring expansion adapter according to any one of claims 1 to 3, comprising: Step 1: Obtain monitoring images of the working conditions of the community distribution room; Step 2: Performing illumination balance on the monitoring image to obtain an illumination balanced monitoring image; Step 3: Annotate the surveillance images to distinguish whether the community distribution room is safe; Step 4: Input the labeled training samples into the deep learning network for training to obtain the community distribution room warning model; Step 5: Use the community distribution room early warning model to complete the identification of the unsafe state of the target community distribution room and issue an early warning.

5. A method for early warning of power distribution room working conditions according to claim 4, characterized in that: The step 2: performing illumination balancing on the monitoring image to obtain an illumination-balanced monitoring image comprises: Step 2.1: Convert the monitoring image into HSV space for contrast enhancement to form a contrast enhanced image; Step 2.2: construct an illumination balance window with any point as the midpoint from the contrast-enhanced image; Step 2.3: Calculate the mean value of the pixels in the light balance window; Step 2.4: Construct the light balance coefficient based on the mean value of the pixels in the light balance window; Step 2.5: Performing illumination balance on the monitoring image using the illumination balance coefficient to obtain an illumination-balanced monitoring image.

6. A method for early warning of power distribution room working conditions according to claim 5, characterized in that: In step 2.4, the light balance coefficient is: Among them, R(x,y) is the light balance coefficient, n×n is the size of the light balance coefficient W with point (i,j) as the midpoint, and v(i,j) is the brightness value of point (i,j). is the average brightness value under the light balance coefficient window W, s(i,j) is the saturation of point (i,j), is the average saturation value under the light balance coefficient window W, δ v (x,y) is the brightness value variance, δ s (x,y) is the saturation variance.

7. A method for early warning of power distribution room working conditions according to claim 6, characterized in that: The step 2.5: performing illumination balance on the monitoring image using the illumination balance coefficient to obtain the illumination-balanced monitoring image comprises: A light balance formula is constructed based on the set coefficients, and the saturation of the monitoring image is processed using the light balance formula to obtain a standardized monitoring image; wherein the standardized processing formula is: s′(x,y)=s(x,y)+t(v′(x,y)-v(x,y))×R(x,y) Among them, s′(x,y) is the saturation of the standardized processing, and t is the balance coefficient; The standardized monitoring image is subjected to denoising to obtain a monitoring image after light balance.

8. A method for early warning of power distribution room working conditions according to claim 7, characterized in that: Step 4: inputting the labeled training samples into the deep learning network for training to obtain the community distribution room early warning model, including: The samples are input into the deep learning network for training, and the deep learning network is optimized using the loss function. When the value of the loss function is less than a preset value, the community distribution room early warning model is output.

9. A method for early warning of power distribution room working conditions according to claim 8, characterized in that: in, In step 4, the loss function is: Among them, Loss is the value of the loss function, n is the number of samples, yi is the actual label of the i-th sample, and yi is the label predicted by the deep learning network.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, the steps in the power distribution room working condition early warning method as described in any one of claims 4-8 are implemented.

Images