A temperature intelligent early warning method, device and equipment of a control cabinet and a medium

Abstract

The application discloses a kind of temperature intelligent early warning method, device, equipment and medium of control cabinet, is related to control cabinet temperature early warning field, the method first determines the initial three-dimensional temperature field distribution nephogram according to the three-dimensional coordinates of N preset positions in target control cabinet and corresponding measured temperature value, it is fused with the two-dimensional infrared temperature distribution image of each surface of cabinet body, the fusion temperature of N preset positions is obtained by the three-dimensional temperature field distribution nephogram after fusion.Through the real-time load current of target control cabinet, fan state and ambient temperature are input to the position temperature prediction model trained, obtain the predicted temperature of N preset positions, superimpose preset temperature increment, obtain dynamic early warning temperature threshold, when fusion temperature exceeds dynamic early warning threshold, intelligent alarm is realized.The application adjusts early warning temperature threshold dynamically by real-time working condition, effectively solves the problem that false alarm is frequent or alarm is lagging.

Description

Technical Field

[0001] This invention relates to the field of temperature early warning for control cabinets, and more particularly to a method, device, equipment, and medium for intelligent temperature early warning of control cabinets. Background Technology

[0002] Existing technology uses a detection and early warning structure in the control cabinet's safety monitoring and early warning device to monitor and detect the ambient temperature inside the control cabinet in real time, obtaining a specific value for the overall temperature inside the control cabinet. However, because there are many heat-generating components inside the cabinet, and the heat generation of each component varies greatly, the effect of using the overall temperature value for temperature early warning is poor.

[0003] In existing technologies, when the overall ambient temperature inside the control cabinet reaches a set, fixed warning threshold, abnormal detection information is output to trigger a safety alarm. However, when the temperature change inside the control cabinet is small, the fixed warning threshold is not reached. The temperature detection structure inside the control cabinet is not sensitive enough, resulting in frequent false alarms or delayed alarms. The alarm is not intelligent enough, and the warning effect is poor. Summary of the Invention

[0004] This invention provides a method, device, equipment, and medium for intelligent temperature early warning of control cabinets, in order to solve the problems in the prior art where, when the temperature change in the control cabinet is not significant, the fixed early warning threshold is not reached, and the temperature detection structure in the control cabinet is not sensitive enough, resulting in frequent false alarms or delayed alarms and poor early warning effect.

[0005] In a first aspect, the present invention provides a method for intelligent temperature early warning of a control cabinet, the method comprising:

[0006] Step 100: Obtain the real-time load current, fan status, and ambient temperature of the target control cabinet, as well as the three-dimensional coordinates of N preset positions within the target control cabinet and the corresponding measured temperature values, where N is a positive integer greater than 1.

[0007] Step 200: Based on the three-dimensional coordinates of N preset positions inside the target control cabinet and the corresponding measured temperature values, determine the initial three-dimensional temperature field distribution cloud map covering the entire internal space of the target control cabinet.

[0008] Step 300: Determine the two-dimensional infrared temperature distribution images of each surface of the target control cabinet;

[0009] Step 400: The initial three-dimensional temperature field distribution cloud map and the two-dimensional infrared temperature distribution image are fused to obtain a fused three-dimensional temperature field distribution cloud map of N preset locations inside the target control cabinet.

[0010] Step 500: Input the real-time load current, fan status and ambient temperature of the target control cabinet into the trained control cabinet location temperature prediction model to obtain the predicted temperature of N preset locations inside the target control cabinet.

[0011] Step 600: Summing the predicted temperature and preset temperature increment at N preset locations within the target control cabinet to obtain the dynamic early warning temperature threshold at the N preset locations within the target control cabinet.

[0012] Step 700: Based on the fused three-dimensional temperature field distribution cloud map of N preset locations within the target control cabinet, and combined with the dynamic early warning temperature threshold of the N preset locations within the target control cabinet, a temperature early warning result is obtained.

[0013] Secondly, the present invention provides a temperature intelligent early warning device for a control cabinet, the temperature intelligent early warning device for the control cabinet comprising:

[0014] The acquisition module is used to acquire the real-time load current, fan status and ambient temperature of the target control cabinet, the three-dimensional coordinates of N preset positions inside the target control cabinet and the corresponding measured temperature values, where N is a positive integer greater than 1.

[0015] The three-dimensional temperature field determination module is used to determine an initial three-dimensional temperature field distribution cloud map covering the entire internal space of the target control cabinet based on the three-dimensional coordinates of N preset positions inside the target control cabinet and the corresponding measured temperature values.

[0016] A two-dimensional temperature distribution determination module is used to determine the two-dimensional infrared temperature distribution images of each surface of the target control cabinet;

[0017] The temperature fusion module is used to fuse the initial three-dimensional temperature field distribution cloud map and the two-dimensional infrared temperature distribution image to obtain a fused three-dimensional temperature field distribution cloud map of N preset locations inside the target control cabinet.

[0018] The predicted temperature calculation module is used to input the real-time load current, fan status and ambient temperature of the target control cabinet into the trained control cabinet location temperature prediction model to obtain the predicted temperature of N preset locations inside the target control cabinet.

[0019] The warning temperature threshold calculation module is used to sum the predicted temperature and preset temperature increment of N preset locations in the target control cabinet to obtain the dynamic warning temperature threshold of N preset locations in the target control cabinet.

[0020] The temperature warning module is used to obtain a temperature warning result based on the fused three-dimensional temperature field distribution cloud map of N preset locations within the target control cabinet, combined with the dynamic warning temperature threshold of the N preset locations within the target control cabinet.

[0021] Thirdly, the present invention provides a computer device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the intelligent temperature early warning method for the control cabinet as described in the first aspect.

[0022] Fourthly, the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the intelligent temperature early warning method for the control cabinet as described in the first aspect.

[0023] The aforementioned intelligent temperature early warning method for control cabinets determines an initial three-dimensional temperature field distribution cloud map covering the entire internal space of the target control cabinet based on the three-dimensional coordinates and corresponding measured temperature values ​​of N preset locations within the target control cabinet. This cloud map is then fused with two-dimensional infrared temperature distribution images of each surface of the target control cabinet to obtain a fused three-dimensional temperature field distribution cloud map of the N preset locations within the target control cabinet, thus enabling the acquisition of the fused temperature at these locations. The real-time load current, fan status, and ambient temperature of the target control cabinet are input into a trained control cabinet location temperature prediction model to obtain the predicted temperature at the N preset locations within the target control cabinet. The predicted temperature and the preset temperature increment are summed to obtain the dynamic early warning temperature threshold for the N preset locations within the target control cabinet. When the fused temperature exceeds the dynamic early warning temperature threshold, an intelligent alarm is triggered. Compared to existing technologies, this invention dynamically adjusts the early warning temperature threshold based on real-time operating conditions, rather than using a fixed threshold, thus solving the problems of frequent false alarms or delayed alarms, improving the intelligence level of the alarm device, and optimizing the alarm effect to a certain extent. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a schematic diagram of an application environment for the intelligent temperature early warning method for the control cabinet in Embodiment 1 of the present invention;

[0026] Figure 2 This is a flowchart of the intelligent temperature early warning method for the control cabinet in Embodiment 1 of the present invention;

[0027] Figure 3 This is a flowchart of the intelligent temperature early warning method for the control cabinet in Embodiment 2 of the present invention;

[0028] Figure 4 This is a flowchart of the intelligent temperature early warning method for the control cabinet in Embodiment 3 of the present invention;

[0029] Figure 5 This is a flowchart of the intelligent temperature early warning method for the control cabinet in Embodiment 4 of the present invention;

[0030] Figure 6 This is a flowchart of the intelligent temperature early warning method for the control cabinet in Embodiment 5 of the present invention;

[0031] Figure 7 This is a flowchart of the intelligent temperature early warning method for the control cabinet in Embodiment Six of the present invention;

[0032] Figure 8 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0033] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0034] The intelligent temperature early warning method for control cabinets provided in this embodiment of the invention can be applied to, for example... Figure 1 The application environment is shown. Specifically, the intelligent temperature early warning method for this control cabinet is applied in a temperature early warning system, which includes, for example, […]. Figure 1 The diagram illustrates a client and server. The client and server communicate over a network to provide real-time temperature alerts. The client, also known as the user terminal, is the program that provides local services to the user, corresponding to the server. The client can be installed on, but is not limited to, various personal computers, laptops, smartphones, tablets, and portable wearable devices. The server can be a standalone server or a server cluster consisting of multiple servers.

[0035] In Example 1, as Figure 2 As shown, this embodiment provides a method for intelligent temperature early warning of a control cabinet, which is applied to... Figure 1 Taking the client as an example, the intelligent temperature early warning method for the control cabinet includes:

[0036] Step 100: Obtain the real-time load current, fan status, and ambient temperature of the target control cabinet, as well as the three-dimensional coordinates of N preset positions within the target control cabinet and the corresponding measured temperature values, where N is a positive integer greater than 1.

[0037] The target control cabinet refers to the designated object of this data acquisition. It is a cabinet containing electrical control components such as circuit breakers, contactors, frequency converters, and PLCs, and is a core device in industrial control and power distribution scenarios. Real-time load current refers to the actual current consumed by the circuits or controlled equipment within the control cabinet at the current moment, usually measured in amperes (A). Fan status refers to the operating status of the cooling fans inside the control cabinet used for forced ventilation and heat dissipation. In this embodiment, it mainly refers to the fan's operating speed. In other implementations, it can also be other parameters, such as the fan's operating frequency, the fan's blowing angle, etc. Ambient temperature refers to the air temperature surrounding the external space where the control cabinet is located and the installation site.

[0038] The N preset locations refer to N monitoring points pre-set inside the control cabinet, where N is a positive integer greater than 1. These points are typically located near heat-generating components, heat dissipation dead zones, or other areas requiring focused monitoring. The three-dimensional coordinates refer to the spatial position parameters of each preset location within the control cabinet space, composed of values ​​on the X, Y, and Z axes, used to precisely locate the physical position of the monitoring point. The corresponding measured temperature value refers to the actual temperature data detected at each preset three-dimensional coordinate location. This data allows for understanding the temperature distribution at different locations within the cabinet and timely detection of localized overheating issues.

[0039] In this embodiment, images acquired by specific infrared sensors are input into thermal simulation software. The software (such as ANSYS Icepak) is used to perform a preliminary analysis of the target electrical control cabinet, identifying the location distribution of the main components within the cabinet. These main components include, but are not limited to, critical heat-generating components (such as heat sinks for power devices and transformer core surfaces), traditional hot spots (such as cable terminal blocks and contactor contacts), critical airflow paths of cooling fans (such as fan inlets, outlets, and duct throats), and temperature sensors. Based on the location distribution of the main components within the target control cabinet, the real-time load current (read from the power distribution system or PLC) and fan status (fan speed) are obtained. The ambient temperature is obtained using an integrated temperature and humidity sensor located near the air inlet of the target control cabinet. Finally, the measured temperature values ​​at N preset locations are obtained using temperature sensors within the target control cabinet.

[0040] Step 200: Based on the three-dimensional coordinates of N preset positions inside the target control cabinet and the corresponding measured temperature values, determine the initial three-dimensional temperature field distribution cloud map covering the entire internal space of the target control cabinet.

[0041] The phrase "covering the entire internal space of the target control cabinet" refers to the generated temperature field data not being limited to N preset locations, but rather using interpolation algorithms to calculate the temperature values ​​at all spatial locations between the N preset locations, ultimately forming a temperature data set encompassing the entire three-dimensional space inside the control cabinet. The initial three-dimensional temperature field refers to the set of temperature values ​​at all locations within the three-dimensional space inside the control cabinet. It reflects the temperature distribution pattern in the X, Y, and Z dimensions of the cabinet, including both measured temperatures at the N preset locations and calculated temperatures in non-measuring areas. This temperature field is an initial version generated based on the current data from the N preset locations and has not undergone subsequent dynamic updates, iterative optimizations, or verification corrections.

[0042] A temperature field distribution cloud map is a visual representation of temperature field data. It uses different colors to represent different temperature ranges (e.g., red for high-temperature areas and blue for low-temperature areas) based on the calculated continuous temperature field data. The color gradients and distribution visually display the temperature differences and spatial distribution patterns within the cabinet, facilitating the rapid identification of localized overheating areas. In this embodiment, the estimated temperature values ​​for all spatial locations between N preset positions are obtained using Kriging interpolation.

[0043] Step 300: Determine the two-dimensional infrared temperature distribution images of each surface of the target control cabinet;

[0044] In this context, each surface of the target control cabinet refers to all external shell components (all outer surfaces of the control cabinet) observable by an infrared thermal imager. These typically include: the front door / cabinet door, usually equipped with indicator lights, buttons, and a display screen; the left and right side panels, often with ventilation louvers or vents; the top panel, which may house an exhaust fan; the back panel, the cable inlet side; and the bottom panel. A two-dimensional infrared temperature distribution image is generated by detecting the infrared energy radiated from various points on the control cabinet surface and converting it into corresponding temperature values. This results in a complete set of temperature data for each surface, used to generate a two-dimensional infrared temperature distribution image for each surface. The final image is a planar image with both length and width dimensions, and each infrared image corresponds to a specific observation plane of the control cabinet. In this embodiment, the two-dimensional infrared temperature distribution images of each surface of the target control cabinet are acquired by an infrared thermal imager.

[0045] Step 400: The initial three-dimensional temperature field distribution cloud map and the two-dimensional infrared temperature distribution image are fused to obtain a fused three-dimensional temperature field distribution cloud map of N preset locations inside the target control cabinet.

[0046] Among them, the fused three-dimensional temperature field distribution cloud map refers to the corrected version of the three-dimensional temperature visualization cloud map obtained after data fusion. It retains the ability of the initial three-dimensional temperature field distribution cloud map to present the temperature distribution in the cabinet space, and also makes corrections by combining the surface temperature data of the two-dimensional infrared temperature distribution image, which is closer to the actual temperature state of the control cabinet than the initial version.

[0047] In this embodiment, the generated initial three-dimensional temperature field distribution cloud map and the two-dimensional infrared temperature distribution image are weighted and fused. During fusion, infrared data is given higher weight in surface areas clearly covered by the infrared image. That is, in infrared-visible areas, the infrared measured temperature and the spatially interpolated temperature are weighted and fused based on the infrared measurement confidence and spatial interpolation confidence of that point to obtain the fused temperature of that point. In areas that cannot be directly measured by infrared, such as the back of components and obstructed areas, the result of Kriging interpolation is completely relied upon. That is, in infrared-invisible areas, the spatially interpolated temperature of that point is adopted as the fused temperature. A fused three-dimensional temperature field distribution cloud map is obtained based on the fused temperatures of N preset locations.

[0048] Step 500: Input the real-time load current, fan status and ambient temperature of the target control cabinet into the trained control cabinet location temperature prediction model to obtain the predicted temperature of N preset locations inside the target control cabinet.

[0049] The trained control cabinet position temperature prediction model refers to an algorithm model trained using machine learning algorithms (such as multiple linear regression and random forest) based on historical data (such as past load current, fan status, ambient temperature, and the actual measured temperature at the corresponding preset positions inside the cabinet). This model has the ability to obtain output temperature prediction values ​​from input operating parameters (such as past load current, fan status, and ambient temperature), and the prediction results have been verified, with accuracy meeting requirements. Predicted temperature refers to the value calculated by the control cabinet position temperature prediction model based on the input real-time operating parameters, estimating the future state of N preset positions under current conditions.

[0050] Step 600: Summing the predicted temperature and preset temperature increment at N preset locations within the target control cabinet to obtain the dynamic early warning temperature threshold at the N preset locations within the target control cabinet.

[0051] The preset temperature increment refers to the temperature difference determined in advance based on the control cabinet's equipment safety standards, operating conditions, and historical fault data. The unit is usually degrees Celsius (°C). It is a fixed safety buffer value, typically 5°C-10°C, used to avoid overly strict or lenient warning thresholds. For example, a larger increment can be set for areas with heating elements, while a smaller increment can be set for ordinary areas. The dynamic warning temperature threshold refers to a threshold that changes with the predicted temperature (rather than being a fixed value). It is the critical value for determining whether there is an overheating risk at that location (i.e., triggering an alarm). In this embodiment, the preset temperature increment is 5°C-10°C.

[0052] Step 700: Based on the fused three-dimensional temperature field distribution cloud map of N preset locations within the target control cabinet, and combined with the dynamic early warning temperature threshold of the N preset locations within the target control cabinet, a temperature early warning result is obtained.

[0053] The temperature warning result refers to the final judgment result obtained by comparing the fused temperature values ​​of N preset locations in the fused 3D temperature field distribution cloud map with the corresponding dynamic warning temperature threshold. Common judgment results include normal (threshold not reached), warning (threshold exceeded but not dangerous), alarm (threshold reached or safety limit exceeded), and emergency (severe overheating and may cause an accident). Corresponding operation and maintenance actions (such as prompting to check the fan, reduce the load, etc.) can be triggered according to the warning level.

[0054] The intelligent temperature early warning method for control cabinets in this embodiment determines an initial three-dimensional temperature field distribution cloud map covering the entire internal space of the target control cabinet based on the three-dimensional coordinates of N preset locations within the target control cabinet and the corresponding measured temperature values. This cloud map is then fused with two-dimensional infrared temperature distribution images of each surface of the target control cabinet to obtain a fused three-dimensional temperature field distribution cloud map of the N preset locations within the target control cabinet, thus acquiring the fused temperature of these N preset locations. The real-time load current, fan status, and ambient temperature of the target control cabinet are input into a trained control cabinet location temperature prediction model to obtain the predicted temperature of the N preset locations within the target control cabinet. The predicted temperature and the preset temperature increment are summed to obtain the dynamic early warning temperature threshold for the N preset locations within the target control cabinet. Comparing the fused temperature with the dynamic early warning temperature threshold yields the temperature early warning result. By dynamically adjusting the early warning temperature threshold in real-time according to the operating conditions, the problem of frequent false alarms or alarm lag is solved, improving the intelligence level of the alarm device and optimizing the alarm effect to a certain extent.

[0055] In Example 2, as Figure 3 As shown, step 200 includes:

[0056] Step 201: Obtain the average distance between the i-th target position of the target control cabinet and N preset positions, where i = 1, ..., n, and n is a positive integer greater than 1; and obtain the first mapping relationship between the average distance between each target position of the target control cabinet and M preset positions in the corresponding target area and the mean temperature variance between the M preset positions, where M is a positive integer greater than 1.

[0057] Here, the i-th target position refers to any position other than the N preset positions within the target control cabinet. The temperature of this target position cannot be directly obtained and can only be calculated through interpolation. The value of i is a positive integer from 1 to n. The average distance between the i-th target position and the N preset positions is the arithmetic mean of the distances between the i-th target position and the N preset positions. The average distance between each target position and the M preset positions within the corresponding target area is the arithmetic mean of the distances between each target position and the M preset positions within the target area (a preset range). The mean temperature variance between each target position and the M preset positions is the arithmetic mean of the temperature variance between each target position and the M preset positions within the preset target area. The first mapping relationship means that the average distance between each target position and the M preset positions within the corresponding target area has a certain correlation with the mean temperature variance between the M preset positions, satisfying a certain functional relationship.

[0058] Step 202: Based on the average distance between the i-th target position of the target control cabinet and N preset positions and the first mapping relationship, calculate the estimated temperature value of the i-th target position of the target control cabinet;

[0059] The estimated temperature value of the i-th target location refers to the temperature value that can be calculated by substituting the average distance between the i-th target location and N preset locations into the first mapping relationship.

[0060] Step 203: Based on the estimated temperature values ​​of the n target locations and the measured temperature values ​​of the N preset locations inside the target control cabinet, determine the initial three-dimensional temperature field distribution cloud map covering the entire internal space of the target control cabinet.

[0061] Among them, the initial three-dimensional temperature field distribution cloud map covering the entire internal space of the target control cabinet refers to the temperature field distribution cloud map obtained by combining the measured temperature values ​​of N preset positions obtained by direct reading with the estimated temperature values ​​calculated by the above interpolation method.

[0062] The intelligent temperature early warning method for the control cabinet in this embodiment, through the refined operations of steps 201 to 203, calculates and determines the temperature data that cannot be measured by temperature sensors located at N preset positions using the Kriging interpolation method, so as to form an initial three-dimensional temperature field distribution cloud map that completely covers the internal space of the target control cabinet. This facilitates fusion with the two-dimensional infrared temperature distribution image to obtain the fused temperature, which is then compared with the subsequently obtained dynamic early warning temperature threshold to obtain the temperature early warning result.

[0063] In Example 3, as Figure 4 As shown, in step 201, obtaining the first mapping relationship between the average distance between each target location of the target control cabinet and M preset locations within the corresponding target area and the mean temperature variance among the M preset locations includes:

[0064] Step 2011: Obtain the three-dimensional coordinates and corresponding measured temperature values ​​of N preset positions inside the target control cabinet, and normalize the three-dimensional coordinates of the N preset positions inside the target control cabinet to obtain normalized three-dimensional coordinates.

[0065] Normalization refers to the preprocessing of 3D coordinate data. It involves mathematically transforming 3D coordinates (length, width, and height dimensions of the digital model inside the cabinet) with different dimensions and numerical ranges into a unified, unitless numerical interval (common intervals are [0,1] or [-1,1]). This aims to eliminate the influence of differences in cabinet dimensions on the coordinate data and prevent excessively large or small coordinate values ​​from interfering with the accuracy of subsequent temperature field modeling and data fusion. Normalized 3D coordinates refer to the 3D coordinates of N preset positions after normalization processing.

[0066] Step 2012: Based on the normalized three-dimensional coordinates of N preset positions in the target control cabinet and the corresponding measured temperature values, calculate the distance between every two preset positions and the mean temperature variance of the corresponding N preset positions in the target control cabinet.

[0067] The distance between any two preset positions out of the N preset positions refers to the Euclidean distance in three-dimensional space calculated based on the normalized three-dimensional coordinates of any two preset positions. All distances are summed, and then divided to obtain the average distance between any two preset positions. The corresponding mean temperature variance is calculated by squaring the temperature difference between any two preset positions, summing the squares, and then dividing the sums to obtain the arithmetic mean.

[0068] Step 2013: Fit the distance between any two preset positions and the corresponding mean temperature variance of the N preset positions in the target control cabinet to obtain the first mapping relationship.

[0069] The first mapping relationship refers to fitting a continuous theoretical curve to the distance between any two preset locations and the corresponding mean temperature variance among N preset locations, resulting in a temperature spatial correlation function model. Commonly used models include the spherical model, the exponential model, and the Gaussian model. The model used in this embodiment is the spherical model, which has a clear physical meaning and conforms to the spatial variation law of the temperature field inside the cabinet. By fitting historical data using the least squares method or the maximum likelihood method, three key parameters of the theoretical model are obtained: range (the maximum effective distance of temperature spatial correlation; beyond this distance, there is no temperature correlation between points), sill value (the semivariance value when the distance approaches infinity, representing the overall variance of the data), and nugget value (the semivariance value when the distance is 0, representing random fluctuations caused by measurement errors or small-scale changes).

[0070] The intelligent temperature early warning method for the control cabinet in this embodiment obtains a first mapping relationship between the average distance between each target location of the target control cabinet and M preset locations within the corresponding target area and the mean temperature variance between the M preset locations. Substituting the average distance between each target location and N preset locations into the first mapping relationship, the estimated temperature value of each target location is obtained. Together with the measured temperature values ​​of the N preset locations, an initial three-dimensional temperature field distribution cloud map covering the entire internal space of the target control cabinet is formed. This cloud map is then fused with a two-dimensional infrared temperature distribution image to obtain a fused temperature. Finally, the fused temperature is compared with a subsequently obtained dynamic early warning temperature threshold to obtain a temperature early warning result.

[0071] In Example 4, as Figure 5 As shown, before step 201, the following steps are also included:

[0072] Determine the digital model inside the target control cabinet, as well as the preset high-temperature zone and preset low-temperature zone of the digital model;

[0073] The preset high-temperature zone refers to key high-temperature areas exceeding a preset temperature (e.g., 60℃), such as core components or areas within the control cabinet that have high heat generation power, poor heat dissipation conditions, and are sensitive to temperature thresholds (e.g., inverter modules, busbar connections). The preset low-temperature zone refers to non-critical areas below the preset temperature (e.g., 60℃), such as auxiliary areas within the cabinet that generate little heat, have good heat dissipation, and are not sensitive to temperature (e.g., cabinet side walls, circuit breaker installation areas). These areas experience gradual temperature changes and have a smaller impact on overall thermal management decisions.

[0074] The digital model inside the target control cabinet is divided into n1 first voxel grids and n2 second voxel grids, wherein the size of the first voxel grid is smaller than the size of the second voxel grid, the first voxel grid corresponds to the preset high temperature region, and the second voxel grid corresponds to the preset low temperature region, and n1 and n2 are both positive integers greater than 1.

[0075] Here, a voxel grid refers to a regularly shaped cubic region (e.g., a 1cm×1cm×1cm cube), and the voxel grid is divided based on temperature. In this embodiment, the size of the first voxel grid can be a fixed value (e.g., a 1cm×1cm×1cm cube) or a data set containing multiple sizes; the size of the second voxel grid is also determined in the same way, and can be either a specific numerical value or a data set.

[0076] Based on n1 first voxel grids, determine n1 target locations; based on n2 second voxel grids, determine n2 target locations, where n1 + n2 = n.

[0077] Here, n1 target positions refer to the positions determined by the center points of n1 first voxel grids, and n2 target positions refer to the positions determined by the center points of n2 second voxel grids. n1 and n2 together constitute the n determined target positions.

[0078] The intelligent temperature early warning method for the control cabinet in this embodiment improves computational efficiency while ensuring overall analysis accuracy by dividing the target location into n1 target locations corresponding to the first voxel grid and n2 target locations corresponding to the second voxel grid, based on temperature. This lays the foundation for the formation of the initial three-dimensional temperature field distribution cloud map, facilitating subsequent fusion with the two-dimensional infrared image. The fused temperature is obtained from the fused three-dimensional temperature field distribution cloud map and compared with the subsequently obtained dynamic early warning temperature threshold to obtain the temperature early warning result.

[0079] In Example 5, as Figure 6 As shown, step 400 includes:

[0080] Step 401: Obtain the first weighting coefficient based on the temperature at each location on the two-dimensional infrared temperature distribution image, and obtain the second weighting coefficient based on the temperature at each location on the initial three-dimensional temperature field distribution cloud map;

[0081] The first weighting coefficient is the temperature confidence weighting factor of the two-dimensional infrared temperature distribution image. The second weighting coefficient is the temperature confidence weighting factor of the initial three-dimensional temperature field distribution cloud map.

[0082] Step 402: When the two-dimensional infrared temperature distribution image and the initial three-dimensional temperature field distribution cloud map have the same position, the temperatures at each position on the two-dimensional infrared temperature distribution image and the temperatures at each position on the initial three-dimensional temperature field distribution cloud map are weighted and summed to obtain the fusion temperature of N1 preset positions, where N1 is a positive integer greater than 1, the sum of the first weighting coefficient and the second weighting coefficient is 1, and the first weighting coefficient is greater than the second weighting coefficient; when the two-dimensional infrared temperature distribution image and the initial three-dimensional temperature field distribution cloud map have different positions, the temperatures at each position on the initial three-dimensional temperature field distribution cloud map are used as the fusion temperature of N2 preset positions, where N2 is a positive integer greater than 1, and N1 + N2 = N;

[0083] The fused temperature of N1 preset locations is obtained by weighting and summing the temperatures at each location in the two-dimensional infrared temperature distribution image and the initial three-dimensional temperature field distribution cloud map according to the first and second weighting coefficients. One or more two-dimensional infrared temperature distribution images with the most common locations are selected and fused with the initial three-dimensional temperature field distribution cloud map. The N1 preset locations are typically located on the surface of the control cabinet. The fused temperature of N2 preset locations is obtained entirely from the temperatures at each location in the initial three-dimensional temperature field distribution cloud map. The N2 preset locations are typically located on surfaces other than the control cabinet.

[0084] Step 403: Based on the fusion temperature of N preset locations, determine the fusion three-dimensional temperature field distribution cloud map of N preset locations within the target control cabinet.

[0085] Among them, the fused three-dimensional temperature field distribution cloud map of N preset positions refers to the fused three-dimensional temperature field distribution cloud map obtained based on the fusion temperature of N1 preset positions and the fusion temperature of N2 preset positions.

[0086] The intelligent temperature early warning method for the control cabinet in this embodiment calculates the fused temperature at preset locations on the surface of the control cabinet and the fused temperature at preset locations on the remaining non-control cabinet surfaces, respectively, to determine the fused three-dimensional temperature field distribution cloud map of N preset locations within the target control cabinet. This facilitates comparison with the subsequently obtained dynamic early warning temperature threshold to obtain the temperature early warning result.

[0087] In Example 6, as Figure 7 As shown, before step 500, the following steps are also included:

[0088] Step 5001: Obtain the historical operating data of the target control cabinet. The historical operating data includes the load current, fan status, ambient temperature at each time, and the measured temperature values ​​at N preset locations inside the target control cabinet at each time.

[0089] The acquired historical operational data refers to a qualified dataset obtained by performing outlier detection, missing value completion, and normalization on all collected historical data. All data are historical operational data from various points in time within the same time dimension.

[0090] Step 5002: Divide the dataset consisting of the historical running data into a training set and a test set according to a preset ratio;

[0091] The preset ratio can be any ratio, and the sum of the preset ratio values ​​is 100%. The training set refers to the dataset used as input to the initial location temperature prediction model. The test set refers to the dataset used to evaluate the model's performance and determine whether the initial location temperature prediction model needs iterative training.

[0092] Step 5003: Construct an initial prediction model for outputting the predicted temperature at N preset locations inside the target control cabinet. The input of the initial prediction model is the load current, fan status and ambient temperature data in the training set, and the output is the predicted temperature at N preset locations inside the target control cabinet.

[0093] The initial prediction model can be either a multiple linear regression model or a random forest model.

[0094] Taking a multiple linear regression model as an example, the initial prediction model satisfies the following constraint relationship: y = T0 + a1x1 + a2x2 + a3x3 + b, where T0 is the initial temperature, a1 is the influence of load current on the predicted temperature (in °C), x1 is the normalized load current (an arbitrary constant between 0 and 1), a1x1 is the predicted temperature value obtained from the load current; a2 is the influence of fan status on the predicted temperature (in °C), x2 is the normalized fan status (an arbitrary constant between 0 and 1), a2x2 is the predicted temperature value obtained from the fan status; a3 is the influence of ambient temperature on the predicted temperature (in °C), x3 is the normalized ambient temperature (an arbitrary constant between 0 and 1), a3x3 is the predicted temperature value obtained from the ambient temperature; b is the error temperature, and y is the predicted temperature (in °C). By inputting the load current, fan status, and ambient temperature data corresponding to each of the N preset positions within the target control cabinet into the multiple linear regression model, the predicted temperatures of the N preset positions within the target control cabinet are obtained.

[0095] The random forest model consists of many decision trees. Each layer of the decision tree has many leaf nodes. Each leaf node in each layer is binary and represents a feature dimension.

[0096] Taking the random forest model as an example, the initial prediction model satisfies the following constraints: The first implementation includes the following steps: Step 1: First, the top leaf node of each of the n decision trees judges the input load current. For the input load current greater than or equal to the preset load current, the output is 1; for the input load current less than the preset load current, the output is 0. The output data (0 or 1) is then input into the next layer of leaf nodes. Step 2: For the leaf node with an output of 1, each of the n decision trees judges the fan status. For the input fan status greater than or equal to the preset fan status, the output is 1; for the input fan status less than the preset fan status, the output is 0. The output data (0 or 1) is then input into the next layer of leaf nodes. Leaf nodes with an output of 0 are discarded. Step 3: For the leaf node with an output of 1, each of the n decision trees judges the ambient temperature. For the input ambient temperature greater than or equal to the preset ambient temperature, the output is 1; for the input ambient temperature less than the preset ambient temperature, the output is 0. The output data (0 or 1) is then input into the last layer of leaf nodes. Leaf nodes with an output of 0 are discarded. Step 4: Each of the n decision trees outputs the pre-defined predicted temperature value corresponding to the leaf node with an output of 1. Step 5: Sum the predicted temperatures output by each of the n decision trees and then calculate the average (i.e., a voting mechanism) to obtain the predicted temperatures for the N pre-defined locations within the target control cabinet.

[0097] The second implementation method includes the following steps: Step 1: Divide the n decision trees into three equal parts to process the input load current, fan status, and ambient temperature data, respectively. Step 2: The top leaf node of each decision tree in the first part judges the input load current. For an input load current greater than or equal to a first preset load current, the output is 1; for a load current less than the first preset load current, the output is 0. The output data (0 or 1) is then input to the next layer of leaf nodes. Step 3: For each leaf node in the first part of the decision tree with an output of 1, if the load current is greater than or equal to a second preset load current, the output is 1; if the load current is less than the second preset load current, the output is 0. The output data (0 or 1) is then input to the last layer of leaf nodes. Leaf nodes with an output of 0 are discarded. Step 4: Each decision tree in the first part of the decision tree outputs the preset predicted temperature value corresponding to the leaf node with an output of 1. Step 5: Sum the predicted temperatures output by each decision tree in the first part of the decision tree and then calculate the average (i.e., a voting mechanism) to obtain the average predicted temperature affected by the load current at N preset locations within the target control cabinet. Step 6: Similarly, following steps two through five, calculate the average predicted temperature affected by the fan status and the average predicted temperature affected by the environment at N preset locations within the target control cabinet, respectively, using each decision tree in the second and third parts of the decision tree. Step 7: Summate the three average predicted temperatures and then perform an arithmetic average to obtain the predicted temperature at the N preset locations within the target control cabinet.

[0098] Step 5004: Using the data from the training set as input, the measured temperature values ​​at N preset locations as supervision targets, setting the parameters of the initial prediction model, and iteratively training the initial prediction model;

[0099] The initial prediction model is iteratively trained, the model parameters are continuously adjusted, and the model fitting effect is monitored in real time using test set data to avoid overfitting.

[0100] Step 5005: Input the data of the test set into the trained prediction model, and evaluate the prediction accuracy of the trained prediction model using a preset accuracy threshold. If the evaluation result meets the preset accuracy threshold, the trained prediction model is determined as the trained control cabinet position temperature prediction model. If the evaluation result does not meet the preset accuracy threshold, adjust the model structure or model parameters, return to step 5004 and re-iterate the training until the preset accuracy threshold is met.

[0101] The preset accuracy threshold refers to a pre-set predicted temperature value used to evaluate the prediction accuracy of the trained prediction model. The test set data is input into the trained prediction model. If the result meets the preset accuracy threshold, the trained prediction model is identified as the trained control cabinet location temperature prediction model. If the result does not meet the preset accuracy threshold, the model structure or training hyperparameters are adjusted, and the process returns to step 5004 for iterative training until the preset accuracy threshold is met.

[0102] The intelligent temperature early warning method for the control cabinet in this embodiment uses machine learning algorithms (such as multiple linear regression and random forest) to process historical data (load current, ambient temperature, and fan status) and establish a temperature prediction model for the control cabinet location. It can dynamically calculate the predicted temperature based on real-time operating conditions, dynamically adjust the early warning temperature threshold, and compare it with the fused temperature obtained by fusing an initial three-dimensional temperature field distribution cloud map and a two-dimensional infrared temperature distribution image to obtain the temperature early warning result.

[0103] In Embodiment Seven, prior to step 600, the method further includes:

[0104] When the predicted temperature is greater than the preset temperature, a first preset temperature increment is determined; when the predicted temperature is not greater than the preset temperature, a second preset temperature increment is determined, wherein the second preset temperature increment is less than the first preset temperature increment, and the value range of the preset temperature increment is 5℃-10℃.

[0105] The predicted temperature refers to inputting the real-time load current, fan status, and ambient temperature of the target control cabinet into a trained control cabinet location temperature prediction model to obtain the predicted temperature for N preset locations within the target control cabinet. The predicted temperature increases with increasing load current, fan status, and ambient temperature, and decreases with decreasing load current, fan status, and ambient temperature. The preset temperature refers to a temperature set in advance.

[0106] For example, if the preset temperature is 60℃, and the fan is running at a high speed under full load and high temperature conditions, the predicted temperature of a certain component may reach 65℃. In this case, the predicted temperature is higher than the preset temperature, and the first preset temperature increment is 10℃ (the warning threshold is automatically raised to 70℃). However, under light load and low temperature conditions, and when the fan is running at a slow speed, the predicted temperature is only 45℃. In this case, the predicted temperature is not higher than the preset temperature, and the second preset temperature increment is 5℃ (the warning threshold is automatically lowered to 55℃).

[0107] The intelligent temperature early warning method for the control cabinet in this embodiment calculates the dynamically predicted temperature based on real-time operating conditions such as load current, fan status, and changes in ambient temperature. This predicted temperature is then compared with a preset temperature. The early warning temperature threshold is dynamically adjusted based on the preset temperature increment. Finally, the predicted temperature is compared with the fused temperature obtained by fusing an initial three-dimensional temperature field distribution cloud map and a two-dimensional infrared temperature distribution image to obtain the temperature early warning result.

[0108] The temperature intelligent early warning device based on the control cabinet includes:

[0109] The acquisition module is used to acquire the real-time load current, fan status and ambient temperature of the target control cabinet, the three-dimensional coordinates of N preset positions inside the target control cabinet and the corresponding measured temperature values, where N is a positive integer greater than 1.

[0110] The three-dimensional temperature field determination module is used to determine an initial three-dimensional temperature field distribution cloud map covering the entire internal space of the target control cabinet based on the three-dimensional coordinates of N preset positions inside the target control cabinet and the corresponding measured temperature values.

[0111] A two-dimensional temperature distribution determination module is used to determine the two-dimensional infrared temperature distribution images of each surface of the target control cabinet;

[0112] The temperature fusion module is used to fuse the initial three-dimensional temperature field distribution cloud map and the two-dimensional infrared temperature distribution image to obtain a fused three-dimensional temperature field distribution cloud map of N preset locations inside the target control cabinet.

[0113] The predicted temperature calculation module is used to input the real-time load current, fan status and ambient temperature of the target control cabinet into the trained control cabinet location temperature prediction model to obtain the predicted temperature of N preset locations inside the target control cabinet.

[0114] The warning temperature threshold calculation module is used to sum the predicted temperature and preset temperature increment of N preset locations in the target control cabinet to obtain the dynamic warning temperature threshold of N preset locations in the target control cabinet.

[0115] The temperature warning module is used to obtain a temperature warning result based on the fused three-dimensional temperature field distribution cloud map of N preset locations within the target control cabinet, combined with the dynamic warning temperature threshold of the N preset locations within the target control cabinet.

[0116] It should be noted that the information interaction and execution process between the above modules are based on the same concept as the method embodiments of the present invention. For details on their specific functions and technical effects, please refer to the method embodiments section, which will not be repeated here.

[0117] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0118] In one embodiment, a computer device is provided, such as Figure 8 As shown, it includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the intelligent temperature early warning method for the control cabinet described in the above embodiments, for example... Figure 2 Steps 100 to 700 shown are omitted here to avoid repetition.

[0119] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When executed by a processor, the computer program implements the intelligent temperature early warning method for the control cabinet described above, for example... Figure 2 Steps 100 to 700 shown are omitted here to avoid repetition.

[0120] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. This computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0121] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0122] While specific embodiments of the invention have been described in detail by way of example, those skilled in the art should understand that the examples are for illustrative purposes only and not intended to limit the scope of the invention. Those skilled in the art should also understand that various modifications can be made to the embodiments without departing from the scope and spirit of the invention. The scope of this invention is defined by the appended claims.

Claims

1. A method for intelligent temperature early warning of a control cabinet, characterized in that, The intelligent temperature early warning method for the control cabinet includes: Step 100: Obtain the real-time load current, fan status, and ambient temperature of the target control cabinet, as well as the three-dimensional coordinates of N preset positions within the target control cabinet and the corresponding measured temperature values, where N is a positive integer greater than 1. Step 200: Based on the three-dimensional coordinates of N preset positions inside the target control cabinet and the corresponding measured temperature values, determine the initial three-dimensional temperature field distribution cloud map covering the entire internal space of the target control cabinet. Step 300: Determine the two-dimensional infrared temperature distribution images of each surface of the target control cabinet; Step 400: The initial three-dimensional temperature field distribution cloud map and the two-dimensional infrared temperature distribution image are fused to obtain a fused three-dimensional temperature field distribution cloud map of N preset locations inside the target control cabinet. Step 400 includes: Step 401: Obtain the first weighting coefficient based on the temperature at each location on the two-dimensional infrared temperature distribution image, and obtain the second weighting coefficient based on the temperature at each location on the initial three-dimensional temperature field distribution cloud map; Step 402: When the two-dimensional infrared temperature distribution image and the initial three-dimensional temperature field distribution cloud map have the same position, the temperatures at each position on the two-dimensional infrared temperature distribution image and the temperatures at each position on the initial three-dimensional temperature field distribution cloud map are weighted and summed to obtain the fusion temperature of N1 preset positions, where N1 is a positive integer greater than 1, the sum of the first weighting coefficient and the second weighting coefficient is 1, and the first weighting coefficient is greater than the second weighting coefficient; when the two-dimensional infrared temperature distribution image and the initial three-dimensional temperature field distribution cloud map have different positions, the temperatures at each position on the initial three-dimensional temperature field distribution cloud map are used as the fusion temperature of N2 preset positions, where N2 is a positive integer greater than 1, and N1 + N2 = N; Step 403: Based on the fusion temperature of N preset locations, determine the fusion three-dimensional temperature field distribution cloud map of N preset locations within the target control cabinet; Step 500: Input the real-time load current, fan status and ambient temperature of the target control cabinet into the trained control cabinet location temperature prediction model to obtain the predicted temperature of N preset locations inside the target control cabinet. Step 600: Summing the predicted temperature and preset temperature increment at N preset locations within the target control cabinet to obtain the dynamic early warning temperature threshold at the N preset locations within the target control cabinet. Step 700: Based on the fused three-dimensional temperature field distribution cloud map of N preset locations within the target control cabinet, and combined with the dynamic early warning temperature threshold of the N preset locations within the target control cabinet, a temperature early warning result is obtained.

2. The intelligent temperature early warning method for the control cabinet according to claim 1, characterized in that, Step 200 includes: Step 201: Obtain the average distance between the i-th target position of the target control cabinet and N preset positions, where i = 1, ..., n, and n is a positive integer greater than 1; and obtain the first mapping relationship between the average distance between each target position of the target control cabinet and M preset positions in the corresponding target area and the mean temperature variance between the M preset positions, where M is a positive integer greater than 1. Step 202: Based on the average distance between the i-th target position of the target control cabinet and N preset positions and the first mapping relationship, calculate the estimated temperature value of the i-th target position of the target control cabinet; Step 203: Based on the estimated temperature values ​​of the n target locations and the measured temperature values ​​of the N preset locations inside the target control cabinet, determine the initial three-dimensional temperature field distribution cloud map covering the entire internal space of the target control cabinet.

3. The intelligent temperature early warning method for the control cabinet according to claim 2, characterized in that, Step 201, obtaining the first mapping relationship between the average distance between each target location of the target control cabinet and M preset locations within the corresponding target area and the average temperature variance among the M preset locations, includes: Step 2011: Obtain the three-dimensional coordinates and corresponding measured temperature values ​​of N preset positions inside the target control cabinet, and normalize the three-dimensional coordinates of the N preset positions inside the target control cabinet to obtain normalized three-dimensional coordinates. Step 2012: Based on the normalized three-dimensional coordinates of N preset positions in the target control cabinet and the corresponding measured temperature values, calculate the distance between every two preset positions and the mean temperature variance of the corresponding N preset positions in the target control cabinet. Step 2013: Fit the distance between any two preset positions and the corresponding mean temperature variance of the N preset positions in the target control cabinet to obtain the first mapping relationship.

4. The intelligent temperature early warning method for the control cabinet according to claim 2, characterized in that, Before step 201, the following are also included: Determine the digital model inside the target control cabinet, as well as the preset high-temperature zone and preset low-temperature zone of the digital model; The digital model inside the target control cabinet is divided into n1 first voxel grids and n2 second voxel grids, wherein the size of the first voxel grid is smaller than the size of the second voxel grid, the first voxel grid corresponds to the preset high temperature region, and the second voxel grid corresponds to the preset low temperature region, and n1 and n2 are both positive integers greater than 1. Based on n1 first voxel grids, determine n1 target locations; based on n2 second voxel grids, determine n2 target locations, where n1 + n2 = n.

5. The intelligent temperature early warning method for the control cabinet according to claim 1, characterized in that, Before step 500, the following are also included: Step 5001: Obtain the historical operating data of the target control cabinet. The historical operating data includes the load current, fan status, ambient temperature at each time, and the measured temperature values ​​at N preset locations inside the target control cabinet at each time. Step 5002: Divide the dataset consisting of the historical running data into a training set and a test set according to a preset ratio; Step 5003: Construct an initial prediction model for outputting the predicted temperature at N preset locations inside the target control cabinet. The input of the initial prediction model is the load current, fan status and ambient temperature data in the training set, and the output is the predicted temperature at N preset locations inside the target control cabinet. Step 5004: Using the data from the training set as input, the measured temperature values ​​at N preset locations as supervision targets, setting the parameters of the initial prediction model, and iteratively training the initial prediction model; Step 5005: Input the data of the test set into the trained prediction model, and evaluate the prediction accuracy of the trained prediction model using a preset accuracy threshold. If the evaluation result meets the preset accuracy threshold, the trained prediction model is determined as the trained control cabinet position temperature prediction model. If the evaluation result does not meet the preset accuracy threshold, adjust the model structure or model parameters, return to step 5004 and re-iterate the training until the preset accuracy threshold is met.

6. The intelligent temperature early warning method for the control cabinet according to claim 1, characterized in that, Before step 600, the following are also included: When the predicted temperature is greater than the preset temperature, a first preset temperature increment is determined; when the predicted temperature is not greater than the preset temperature, a second preset temperature increment is determined, wherein the second preset temperature increment is less than the first preset temperature increment, and the value range of the preset temperature increment is 5℃-10℃.

7. A temperature intelligent early warning device for a control cabinet, characterized in that, The intelligent temperature early warning device for the control cabinet includes: The acquisition module is used to acquire the real-time load current, fan status and ambient temperature of the target control cabinet, the three-dimensional coordinates of N preset positions inside the target control cabinet and the corresponding measured temperature values, where N is a positive integer greater than 1. The three-dimensional temperature field determination module is used to determine an initial three-dimensional temperature field distribution cloud map covering the entire internal space of the target control cabinet based on the three-dimensional coordinates of N preset positions inside the target control cabinet and the corresponding measured temperature values. A two-dimensional temperature distribution determination module is used to determine the two-dimensional infrared temperature distribution images of each surface of the target control cabinet; The temperature fusion module is used to fuse the initial three-dimensional temperature field distribution cloud map and the two-dimensional infrared temperature distribution image to obtain a fused three-dimensional temperature field distribution cloud map of N preset locations inside the target control cabinet. The temperature fusion module can also be used to obtain a first weighting coefficient based on the temperature at each location on the two-dimensional infrared temperature distribution image, and a second weighting coefficient based on the temperature at each location on the initial three-dimensional temperature field distribution cloud map. When the two-dimensional infrared temperature distribution image and the initial three-dimensional temperature field distribution cloud map have the same location, the temperatures at each location on the two-dimensional infrared temperature distribution image and the temperatures at each location on the initial three-dimensional temperature field distribution cloud map are weighted and summed to obtain the fused temperatures of N1 preset locations, where N1 is a positive integer greater than 1, the sum of the first weighting coefficient and the second weighting coefficient is 1, and the first weighting coefficient is greater than the second weighting coefficient. When the two-dimensional infrared temperature distribution image and the initial three-dimensional temperature field distribution cloud map have different locations, the temperatures at each location on the initial three-dimensional temperature field distribution cloud map are used as the fused temperatures of N2 preset locations, where N2 is a positive integer greater than 1, and N1 + N2 = N. Based on the fused temperatures of the N preset locations, the fused three-dimensional temperature field distribution cloud map of the N preset locations within the target control cabinet is determined. The predicted temperature calculation module is used to input the real-time load current, fan status and ambient temperature of the target control cabinet into the trained control cabinet location temperature prediction model to obtain the predicted temperature of N preset locations inside the target control cabinet. The warning temperature threshold calculation module is used to sum the predicted temperature and preset temperature increment of N preset locations in the target control cabinet to obtain the dynamic warning temperature threshold of N preset locations in the target control cabinet. The temperature warning module is used to obtain a temperature warning result based on the fused three-dimensional temperature field distribution cloud map of N preset locations within the target control cabinet, combined with the dynamic warning temperature threshold of the N preset locations within the target control cabinet.

8. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the intelligent temperature early warning method for the control cabinet as described in any one of claims 1 to 6.

9. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the intelligent temperature early warning method for the control cabinet according to any one of claims 1 to 6.

Images