Self-adaptive heat dissipation power distribution cabinet temperature regulation and control method and system

Abstract

The invention relates to the technical field of power distribution cabinet temperature control, and discloses a self-adaptive heat dissipation power distribution cabinet temperature regulation and control method and system, and the method comprises the steps: obtaining a temperature distribution record which comprises a timestamp, a position identifier and a temperature value; performing data preprocessing on the temperature distribution record to obtain standard temperature data; performing dynamic change analysis on the standard temperature data to obtain temperature change characteristics; inputting the temperature change characteristics into a pre-constructed temperature change prediction model, and outputting future temperature data; performing risk assessment on the future temperature data, and judging and triggering a heat dissipation demand signal according to an assessment result; calling a pre-established heat dissipation resource allocation database according to the heat dissipation demand signal, obtaining an optimal equipment scheduling combination, and generating an equipment starting instruction set; and activating a corresponding heat dissipation equipment operation module through the equipment starting instruction set. According to the method, accurate and efficient temperature regulation and control are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of temperature control of power distribution cabinets, and in particular to a temperature control method and system for power distribution cabinets with adaptive heat dissipation. Background Art

[0002] In modern industry and intelligent device management, particularly in the critical power distribution link of smart grids, temperature control is crucial for ensuring stable operation, extending the lifespan, and preventing accidents for power equipment (such as distribution cabinets). Efficient temperature management in complex operating environments presents significant challenges. First, real-time and comprehensive temperature monitoring at multiple points within the distribution cabinet is required to capture subtle variations and spatial distribution differences, placing high demands on the density and accuracy of the sensor network. Second, temperature fluctuations exhibit both periodic patterns and sudden fluctuations, necessitating the construction of accurate prediction models using multi-dimensional data (time series and spatial gradients) to anticipate potential risks. Furthermore, dynamic and optimal allocation of cooling resources must be implemented under complex operating conditions, ensuring timely response to local overheating demands while also balancing overall load and energy efficiency to avoid resource waste or inadequate response. The core of these challenges lies in establishing a closed-loop control mechanism, from precise sensing and intelligent prediction to dynamic execution, to adapt to environmental variability and ensure continuous stability.

[0003] In one existing technology, a small number of temperature sensors (usually only 2-3) are first installed in fixed locations within the cabinet to collect temperature data at fixed intervals (e.g., every 5 minutes). The system presets a single static threshold (e.g., 70°C). When any sensor reading exceeds this threshold, the preset cooling response mode is immediately triggered: regardless of the temperature distribution characteristics or the degree of overshoot, all cooling devices (e.g., fans) are activated and run at maximum power for a fixed period of time (e.g., 10 minutes). During operation, the device status or temperature change trends are not monitored; only a simple timer is used to control the cooling duration. After the preset operating time is reached, all devices are automatically shut down and the next round of data collection is awaited. If sensor data is missing or abnormal, the system only records the error log without performing data repair or policy adjustments, and continues to execute the collection and response process according to the original cycle. However, in the existing technology, due to the use of a small number of sensors in fixed positions and fixed time intervals to collect data, there are blind spots in temperature monitoring, which cannot fully reflect the actual temperature distribution inside the distribution cabinet; at the same time, the trigger mechanism based on a single static threshold lacks the ability to predict the temperature change trend. Regardless of the degree of local overheating, a unified heat dissipation response mode is adopted, which may cause insufficient heat dissipation and lead to the risk of equipment overheating, and may also waste energy due to excessive heat dissipation; in addition, the system lacks real-time monitoring and dynamic adjustment functions during the heat dissipation process, and cannot optimize the strategy according to the actual heat dissipation effect. The processing of data anomalies only stays at the recording level, and an effective closed-loop control mechanism has not been formed, which ultimately leads to insufficient accuracy and adaptability of temperature management.

[0004] In summary, the existing technology has the problem of being unable to achieve accurate and efficient temperature control. Summary of the Invention

[0005] The present invention provides a temperature control method and system for a power distribution cabinet with adaptive heat dissipation, so as to achieve precise and efficient temperature control.

[0006] In a first aspect, in order to solve the above technical problems, the present invention provides a temperature control method for a power distribution cabinet with adaptive heat dissipation, comprising: Obtaining a temperature distribution record, wherein the temperature distribution record includes a timestamp, a location identifier, and a temperature value; Performing data preprocessing on the temperature distribution record to obtain standard temperature data; Performing dynamic change analysis on the standard temperature data to obtain temperature change characteristics; Inputting the temperature change characteristics into a pre-built temperature change prediction model to output future temperature data; Performing a risk assessment on the future temperature data, and determining whether to trigger a heat dissipation demand signal based on the assessment result; Calling a pre-established heat dissipation resource allocation database according to the heat dissipation demand signal to obtain an optimal device scheduling combination and generate a device startup instruction set; The corresponding heat dissipation device operation module is activated through the device startup instruction set.

[0007] In an optional embodiment, the performing data preprocessing on the temperature distribution record to obtain standard temperature data includes: De-noising the temperature distribution record using a median filter algorithm to obtain de-noised temperature data; Using a linear interpolation algorithm to fill missing values ​​in the denoised temperature data to obtain filled temperature data; Using a moving average algorithm to smooth and correct the filled temperature data to obtain smoothed temperature data; The smoothed temperature data is standardized using a Z-score algorithm to obtain standard temperature data.

[0008] In an optional embodiment, the performing dynamic change analysis on the standard temperature data to obtain temperature change characteristics includes: Using a fixed time window algorithm to divide the standard temperature data into temperature data of independent time periods to generate segmented temperature data; According to the segmented temperature data, the temperature variation amplitude is calculated by using the range statistics method and the temperature spatial gradient is calculated by using the adjacent point gradient method; The temperature variation amplitude and the temperature spatial gradient are combined into a comprehensive feature vector to generate a temperature variation feature.

[0009] In an optional embodiment, the process of constructing the temperature change prediction model includes: Obtaining historical temperature change characteristics and inputting them into the input layer of the preliminarily constructed long short-term memory network model for training, thereby obtaining a predicted temperature output by the output layer of the long short-term memory network model; Input the predicted temperature and the pre-stored actual temperature into the mean square error loss function to calculate the loss value; According to the loss value, the gradient of the output layer of the long short-term memory network model is calculated by the back propagation algorithm, and the gradient is passed forward layer by layer by the chain rule to calculate the gradient of the parameters of each layer to obtain gradient data; Based on the gradient data and the preset learning rate, the parameters of each layer of the long short-term memory network model are updated through the Adam optimizer; The parameters of each layer are updated repeatedly until the training times of the long short-term memory network model are greater than the preset times, or when the loss value data of the long short-term memory network model is less than the preset loss threshold, the training is determined to be completed and a temperature change prediction model is obtained.

[0010] In an optional embodiment, performing risk assessment on the future temperature data and determining triggering a heat dissipation demand signal based on the assessment result includes: Comparing the future temperature data with a preset temperature threshold, and when the future temperature data exceeds the temperature threshold, marking the location identifier corresponding to the future temperature data as a high-risk point, and generating risk distribution data, wherein the risk distribution data includes a risk location, a risk temperature, and a risk time; Multiplying the risk temperature by the risk time to obtain a risk value, and generating a priority list in descending order according to the risk values; Based on the risk location, the K-means algorithm is used to identify the dense high-temperature area and obtain the heat dissipation area; According to the heat dissipation area, a load balancing algorithm is used to assign a heat dissipation intensity level; The priority list, the heat dissipation areas, and the heat dissipation intensity levels are used as heat dissipation requirement signals.

[0011] In an optional embodiment, calling a pre-established heat dissipation resource allocation database according to the heat dissipation demand signal to obtain an optimal device scheduling combination and generate a device startup instruction set includes: Based on the heat dissipation demand signal, extracting the device status data and heat dissipation parameter range of the heat dissipation area from a pre-established heat dissipation resource allocation database, and filtering out a list of available devices through conditional query based on the device status data and the heat dissipation parameter range; According to the available device list, combined with the priority list and the heat dissipation intensity level, a polling algorithm is used to perform load balancing and output an optimal device scheduling combination; According to the optimal device scheduling combination, a device startup instruction set is generated through an industrial communication protocol.

[0012] In an optional embodiment, activating the corresponding heat dissipation device operation module by using the device startup instruction set includes: Parsing the device startup instruction set into an activation signal corresponding to the device ID to generate an activation signal set; A PID algorithm is used to adjust the heat dissipation device according to the activation signal set and the temperature distribution record.

[0013] In a second aspect, the present invention provides a temperature control system for a power distribution cabinet with adaptive heat dissipation, comprising: A data acquisition module is used to acquire temperature distribution records, wherein the temperature distribution records include a timestamp, a location identifier, and a temperature value; A data processing module, configured to perform data preprocessing on the temperature distribution records to obtain standard temperature data; A temperature characteristic analysis module is used to perform dynamic change analysis on the standard temperature data to obtain temperature change characteristics; A future temperature prediction module, configured to input the temperature change characteristics into a pre-built temperature change prediction model and output future temperature data; a heat dissipation demand analysis module, configured to perform a risk assessment on the future temperature data and determine whether to trigger a heat dissipation demand signal based on the assessment result; An optimal device combination module is used to call a pre-established heat dissipation resource allocation database according to the heat dissipation demand signal, obtain an optimal device scheduling combination, and generate a device startup instruction set; An execution module is configured to activate the corresponding heat dissipation device operation module through the device startup instruction set. In a third aspect, the present invention further provides an electronic device comprising a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, it implements any of the above-mentioned methods for temperature control of a power distribution cabinet with adaptive heat dissipation.

[0014] In a fourth aspect, the present invention also provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute any one of the above-mentioned adaptive heat dissipation distribution cabinet temperature control methods.

[0015] Compared with the prior art, the present invention has the following beneficial effects: (1) Through the data preprocessing process combining the median filter algorithm, linear interpolation algorithm, moving average algorithm and Z-score normalization algorithm, the problems of large sensor data noise, many missing values ​​and insufficient standardization in the existing technology are solved, the accuracy and consistency of temperature data are improved, and reliable input is provided for subsequent dynamic analysis, ensuring the comprehensiveness and real-time performance of temperature monitoring.

[0016] (2) Based on the fixed time window algorithm, the temperature data is divided, the temperature change amplitude is calculated by combining the range statistics method and the neighboring point gradient method to calculate the spatial gradient, and a comprehensive feature vector is generated. The vector is then input into the LSTM model for training and prediction. This overcomes the defect that the static threshold mechanism cannot predict the temperature trend, and achieves accurate prediction of future temperature data, thus identifying potential overheating risks in advance.

[0017] (3) Risk value calculation, priority list generation and K-means algorithm are used to identify dense high-temperature areas, and the load balancing algorithm is used to allocate heat dissipation intensity levels to solve the problem of insufficient local heat dissipation or energy waste caused by unified heat dissipation response, realize intelligent optimization allocation of heat dissipation resources, improve heat dissipation efficiency and reduce energy consumption.

[0018] (4) The list of available devices is screened through conditional queries of the heat dissipation resource allocation database, and the load balancing output is optimized by combining the polling algorithm. The PID algorithm is used to dynamically adjust the operation of the heat dissipation equipment to form a closed-loop control mechanism. This overcomes the shortcomings of the existing technology in lacking real-time monitoring and dynamic adjustment, ensures the adaptability of the heat dissipation process and the stability of the system, and extends the life of the equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a schematic flow chart of a method for temperature control of a power distribution cabinet with adaptive heat dissipation provided by a first embodiment of the present invention; Figure 2 It is a schematic structural diagram of a temperature control system for a power distribution cabinet with adaptive heat dissipation provided by the second embodiment of the present invention. DETAILED DESCRIPTION

[0020] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0021] Reference Figure 1 The first embodiment of the present invention provides a temperature control method for a power distribution cabinet with adaptive heat dissipation, comprising the following steps: S11, obtaining a temperature distribution record, wherein the temperature distribution record includes a timestamp, a location identifier, and a temperature value; S13, performing data preprocessing on the temperature distribution record to obtain standard temperature data; S13, performing dynamic change analysis on the standard temperature data to obtain temperature change characteristics; S14, inputting the temperature change characteristics into a pre-built temperature change prediction model to output future temperature data; S15, performing a risk assessment on the future temperature data, and determining whether to trigger a heat dissipation demand signal based on the assessment result; S16, calling a pre-established heat dissipation resource allocation database according to the heat dissipation demand signal, obtaining an optimal device scheduling combination, and generating a device startup instruction set; S17, activating the corresponding heat dissipation device operation module through the device startup instruction set.

[0022] In step S11 , a temperature distribution record is obtained, where the temperature distribution record includes a timestamp, a location identifier, and a temperature value.

[0023] Specifically, the system collects temperature distribution records in real time through a distributed temperature sensor network deployed in key areas inside the distribution cabinet. The temperature distribution record consists of three core fields: a timestamp records the precise moment of data collection (formatted as year, month, day, hour, minute, second, and millisecond), a location identifier corresponds to the unique installation coordinates of the sensor within the distribution cabinet (using a three-dimensional grid coding system for the cabinet, such as an XYZ partition identifier), and a temperature value is the Celsius temperature value detected by the sensor (with accuracy meeting industrial standards). The data collection process uses a time synchronization protocol to ensure that all sensors operate under a unified time reference. The collected raw data is transmitted to the central processing unit via the industrial bus, forming a structured data table stored in a real-time database. This step eliminates traditional monitoring blind spots through a high-density sensor layout, laying the data foundation for precise control.

[0024] In step S12, data preprocessing is performed on the temperature distribution record to obtain standard temperature data.

[0025] In a specific embodiment, the data preprocessing of the temperature distribution record to obtain standard temperature data comprises: The temperature distribution record is denoised by a median filtering algorithm to obtain denoised temperature data; The denoised temperature data is filled with missing values by a linear interpolation algorithm to obtain filled temperature data; The filled temperature data is smoothed by a moving average algorithm to obtain smoothed temperature data; The smoothed temperature data is standardized by a Z-score algorithm to obtain standard temperature data.

[0026] Specifically, the temperature value field in the original temperature distribution record is first processed by a median filtering algorithm. Taking each position identifier as the center, a sliding window (the window size is preset to 3x3 grid according to the physical size of the power distribution cabinet and the sensor spacing) is selected from the adjacent sensors, the median of all temperature values in the window is taken after sorting, and the original value is replaced to generate denoised temperature data. This operation eliminates transient interference noise of the sensor and retains the true temperature change trend.

[0027] Subsequently, the denoised temperature data is filled with missing values. When it is detected that a position identifier has data missing in consecutive time stamps, the filling value at the missing time is calculated linearly according to the temperature values of the two valid time stamps before and after the position, and the filled temperature data is generated. The filling trigger condition is that the missing duration is less than a preset safety interval (the interval is determined by the maximum allowed temperature monitoring interruption time of the device, and is set to a certain multiple of the sampling period through thermodynamic simulation verification).

[0028] Then, the filled temperature data is processed by a moving average algorithm. Taking the current time stamp as the center, the arithmetic mean of the temperature values at the same position in the continuous time window (the window length is set to an integer multiple of the thermal response time of the device according to the temperature fluctuation characteristics) is taken, the current value is replaced to generate smoothed temperature data, and non-periodic random fluctuations are suppressed.

[0029] Finally, the smoothed temperature data is standardized by a Z-score algorithm. The temperature mean and standard deviation of each position identifier are calculated based on historical operation data, the current temperature value is subtracted from the corresponding position mean and then divided by the standard deviation to generate standard temperature data conforming to the normal distribution. The standardization parameters are calculated and stored in advance through offline analysis of three months of historical data.

[0030] This preprocessing procedure eliminates noise interference, ensures data continuity, smooths short-term fluctuations, and unifies the dimension, so that the subsequent analysis module obtains reliable input with consistent time and space dimensions, laying a foundation for accurate temperature prediction.

[0031] In step S13, a dynamic change analysis is performed on the standard temperature data to obtain temperature change characteristics.

[0032] In a specific embodiment, the dynamic change analysis of the standard temperature data to obtain the temperature change characteristics includes: Using a fixed time window algorithm to divide the standard temperature data into temperature data of independent time periods to generate segmented temperature data; According to the segmented temperature data, the temperature variation amplitude is calculated by using the range statistics method and the temperature spatial gradient is calculated by using the adjacent point gradient method; The temperature variation amplitude and the temperature spatial gradient are combined into a comprehensive feature vector to generate a temperature variation feature.

[0033] Specifically, the system uses a fixed time window algorithm to process standard temperature data. A fixed time window is set according to the thermal time constant of the power distribution cabinet. The standard temperature data with continuous timestamps is cut according to the window boundary to generate segmented temperature data containing independent time periods. Each segmented temperature data contains a temperature value sequence of all position identifiers in the window. Two-dimensional feature extraction is performed based on the segmented temperature data: in the time dimension, the temperature variation amplitude is calculated using the extreme difference statistics method for the temperature value sequence of each position identifier in a single time window. Specifically, the temperature variation amplitude value is generated by taking the difference between the maximum and minimum temperature values ​​of the position in the window; in the spatial dimension, the temperature spatial gradient is calculated using the neighboring point gradient method for the temperature distribution at the end of each time window. After selecting the target position identifier, the temperature difference between it and the nearest neighbor position identifier in the X / Y / Z direction in the three-dimensional grid code is calculated and divided by the physical distance. The modulus of the gradient vector in each direction is taken as the spatial gradient value. The above features are aligned according to the location identifier and time window. Each location identifier corresponds to a temperature change amplitude value and a temperature spatial gradient value in each time window. They are merged into a two-dimensional comprehensive feature vector, and finally a structured temperature change feature containing four fields: location identifier, time window number, temperature change amplitude, and temperature spatial gradient is generated. This step breaks through the limitations of the single-point temperature threshold of the existing technology through dual-dimensional spatiotemporal feature extraction. The temperature change amplitude quantifies the intensity of thermal fluctuations at local points to identify abnormal temperature rise. The temperature spatial gradient reveals the direction and rate of heat conduction in the cabinet and locates the source of heat accumulation. The comprehensive feature vector contains both temporal dynamics and spatial correlation to provide a physically meaningful input for the prediction model. The final output temperature change feature accurately characterizes the evolution law of the distribution cabinet temperature field, enabling the subsequent prediction module to predict the heat diffusion path and achieve precise spatiotemporal positioning of heat dissipation intervention.

[0034] In step S14, the temperature change characteristics are input into a pre-built temperature change prediction model, and future temperature data is output.

[0035] In a specific embodiment, the process of constructing the temperature change prediction model includes: Obtaining historical temperature change characteristics and inputting them into the input layer of the preliminarily constructed long short-term memory network model for training, thereby obtaining a predicted temperature output by the output layer of the long short-term memory network model; Input the predicted temperature and the pre-stored actual temperature into the mean square error loss function to calculate the loss value; According to the loss value, the gradient of the output layer of the long short-term memory network model is calculated by the back propagation algorithm, and the gradient is passed forward layer by layer by the chain rule to calculate the gradient of the parameters of each layer to obtain gradient data; Based on the gradient data and the preset learning rate, the parameters of each layer of the long short-term memory network model are updated through the Adam optimizer; The parameters of each layer are updated repeatedly until the training times of the long short-term memory network model are greater than the preset times, or when the loss value data of the long short-term memory network model is less than the preset loss threshold, the training is determined to be completed and a temperature change prediction model is obtained.

[0036] Specifically, the system uses a dataset of temperature variation characteristics accumulated during historical operation as a training data source. This dataset includes fields such as location identifier, time window number, temperature variation amplitude, and temperature spatial gradient. These features are fed into the input layer of a pre-built long-short-term memory network model. The model's internal time series processing mechanism generates predicted temperature values ​​at the output layer.

[0037] When the long short-term memory network (LSTM) processes temperature feature data to generate predicted temperature, the model uses a sophisticated gating mechanism to collaboratively filter, update, and transmit spatiotemporal information. The specific workflow is as follows: When temperature change features (location identifier, time window number, temperature change amplitude, spatial gradient) enter the LSTM input layer, the model first splits the feature vector into the input data for the current moment. This data is then simultaneously fed into three key gating units: the forget gate, the input gate, and the output gate. Each gating unit dynamically regulates the information flow using self-learned weight parameters.

[0038] The forget gate first analyzes the relevance of historical memories. It receives the current input features (e.g., "temperature change at location A in time window 5: +0.8°C, spatial gradient: 1.2°C / m") and the previous hidden state (including the temperature evolution trend in the previous window) and generates a forgetting coefficient between 0 and 1 through an activation function. For example, if it detects a sudden increase in the current spatial gradient, the forget gate will reduce the weight retained for the earlier low-temperature steady-state memory (e.g., setting the memory coefficient for the 0.2°C / m gradient at the same location three days ago to 0.1). This effectively weakens the influence of outdated patterns in the neural network.

[0039] The input gate performs two tasks in parallel: first, it evaluates the value of the current feature through the "input screening layer." For example, a sudden temperature change (e.g., 1.5°C) is identified as a critical signal and assigned a high weight of 0.9. Second, it fuses the filtered features with historical states through the "memory enhancement layer" to generate new candidate memory content. For example, combining the current high temperature gradient with the previously accumulated trend of declining heat dissipation efficiency can form a new memory fragment of "accelerated heat accumulation."

[0040] At this point, the cell state (memory backbone) begins updating: the forget gate coefficient is multiplied by the old state (e.g., weakening the regular fluctuation memory from three hours ago), plus the input gate coefficient is multiplied by the new candidate memory (e.g., strengthening the newly generated "heat accumulation acceleration" feature), forming an updated memory that integrates spatiotemporal dynamics. This is equivalent to using physical laws to simulate the heat conduction and accumulation process in the distribution cabinet within the neural network's memory channel.

[0041] Finally, the output gate controls the release of prediction information: based on the updated cell state and the current input, it calculates output coefficients to determine which memory features influence the final prediction. For example, if the "accelerated heat accumulation" feature is present in the memory backbone and the current input indicates that the cooling device is idle, the output gate amplifies the weight of this risk signal (coefficient 0.95), causing the output layer to generate a higher-than-usual predicted temperature value (such as 72.3°C). Simultaneously, the current hidden state is passed to the next time step, carrying a persistent warning signal, such as "heat accumulation at location A has not eased."

[0042] The entire processing mechanism forms a closed loop: under the collaborative filtering of the gating system, key patterns are continuously strengthened (such as sudden temperature rise) and redundant information is gradually attenuated (such as steady-state noise), ultimately generating a predicted value that conforms to the laws of thermodynamic evolution at the output layer.

[0043] The predicted temperature value and the actual temperature value of the same time window stored in the real-time database are input into the mean square error loss function, and the square mean of the prediction error is calculated as the loss value.

[0044] Based on this loss value, backpropagation is performed: first, the gradient of the output layer is calculated. This gradient is then propagated forward layer by layer to the hidden and input layers using the chain rule. Gradient data for the weight and bias parameters of each layer is then calculated. The parameters of each layer are updated using the Adam optimizer combined with a preset learning rate (optimized using a grid search method on a validation set). The parameter update process is iteratively executed, terminating when the number of training iterations exceeds a preset maximum number (set to the number of iterations at which the loss stabilizes according to the early stopping strategy) or when the loss falls below a preset loss threshold (determined by a specific percentile of the historical prediction error distribution). This results in the final temperature change prediction model.

[0045] During the prediction phase, the current temperature change features generated in step S13 are input into the trained model, which then outputs future temperature data, including the future time window number, location identifier, and predicted temperature value. This prediction model uses a long-short-term memory network to model the temporal dependencies of spatiotemporal features. This model overcomes the limitations of existing static threshold mechanisms, accurately capturing the inertial laws and spatial conduction effects of temperature changes. This allows for a quantitative prediction of future temperature field evolution trends, providing a forward-looking basis for cooling decisions and avoiding the risk of equipment overheating due to delayed responses.

[0046] In step S15, a risk assessment is performed on the future temperature data, and a heat dissipation demand signal is triggered based on the assessment result.

[0047] In a specific embodiment, the performing risk assessment on the future temperature data and determining triggering a heat dissipation demand signal based on the assessment result includes: Comparing the future temperature data with a preset temperature threshold, and when the future temperature data exceeds the temperature threshold, marking the location identifier corresponding to the future temperature data as a high-risk point, and generating risk distribution data, wherein the risk distribution data includes a risk location, a risk temperature, and a risk time; Multiplying the risk temperature by the risk time to obtain a risk value, and generating a priority list in descending order according to the risk values; Based on the risk location, the K-means algorithm is used to identify the dense high-temperature area and obtain the heat dissipation area; According to the heat dissipation area, a load balancing algorithm is used to assign a heat dissipation intensity level; The priority list, the heat dissipation areas, and the heat dissipation intensity levels are used as heat dissipation requirement signals.

[0048] Specifically, the system compares the predicted temperature values ​​in future temperature data with preset temperature thresholds (set based on the temperature resistance rating of the insulation material in the power distribution cabinet and the equipment's safe operating specifications) for each location identifier. If the predicted temperature at a location identifier exceeds the threshold, the location is marked as a high-risk point, and risk distribution data is generated, including the risk location (location identifier field), risk temperature (predicted temperature value), and risk duration (the number of windows in which the temperature exceeds the limit).

[0049] Calculate the risk value of each high-risk point based on the risk distribution data: take the product of risk temperature and risk time as the quantitative indicator. Sort all high-risk points from high to low risk value to generate a priority list (the data structure includes location identifier, risk value, and sorting sequence number).

[0050] The K-means algorithm is used to process the three-dimensional coordinate data of risk locations: the number of physical space partitions of the distribution cabinet is used as the number of cluster centers, and the high-risk points with the closest Euclidean distance are grouped into the same cluster. The coordinate range of the dense high-temperature area is output as the heat dissipation area (the data structure includes the area number and boundary coordinates).

[0051] Based on the sum of the risk values ​​of all high-risk points in the heat dissipation area, the heat dissipation intensity level is assigned through a load balancing algorithm: the sum of the risk values ​​is divided into three levels according to a preset proportional interval (the level division threshold is determined by fitting the maximum cooling capacity of the heat dissipation equipment with historical heat load data), and a mapping relationship between the heat dissipation area and the intensity level is generated.

[0052] The final output is a cooling demand signal consisting of three core fields: priority list, cooling area, and cooling intensity level. This step addresses the shortcomings of the existing unified cooling response: risk values ​​quantify the degree of overheating risk, and the priority list ensures that critical points are prioritized; K-means clustering identifies heat accumulation areas to avoid fragmentation of cooling resources; and intensity levels match actual heat load requirements, enabling precise hierarchical control of cooling resources, improving cooling efficiency and reducing energy consumption at the decision-making level.

[0053] In step S16, a pre-established heat dissipation resource allocation database is called according to the heat dissipation demand signal to obtain an optimal device scheduling combination and generate a device startup instruction set.

[0054] In a specific embodiment, calling a pre-established heat dissipation resource allocation database according to the heat dissipation demand signal to obtain an optimal device scheduling combination and generate a device startup instruction set includes: Based on the heat dissipation demand signal, extracting the device status data and heat dissipation parameter range of the heat dissipation area from a pre-established heat dissipation resource allocation database, and filtering out a list of available devices through conditional query based on the device status data and the heat dissipation parameter range; According to the available device list, combined with the priority list and the heat dissipation intensity level, a polling algorithm is used to perform load balancing and output an optimal device scheduling combination; According to the optimal device scheduling combination, a device startup instruction set is generated through an industrial communication protocol.

[0055] Specifically, based on the cooling zone field in the cooling demand signal, the system extracts the device status data and cooling parameter ranges for all cooling devices within that zone from a pre-established cooling resource allocation database. The device status data includes fields such as device ID, operating status, and accumulated operating hours, while the cooling parameter range includes fields such as minimum cooling temperature and maximum air volume range. A conditional query filters out devices whose operating status is idle and whose cooling parameter ranges meet the required cooling intensity level for the zone. A list of available devices is generated, containing device IDs, location coordinates, and cooling capacity values. Combined with the priority list and cooling intensity level fields in the cooling demand signal, a round-robin algorithm is used to perform load balancing. High-risk location identifiers are traversed in order of priority. For each location identifier, the next unassigned device is selected from the available device list. The device's cooling capacity range is then matched to the cooling intensity level of the cooling zone to which the location belongs (the mapping between intensity levels and capacity ranges is predefined by the device performance curve). When a device's remaining cooling capacity meets the required cooling intensity, a binding is established between the device ID and location identifier, and the device's remaining capacity value is updated. The optimal device scheduling combination, consisting of the device ID, bound location identifier, and startup parameters, is output. Finally, the startup parameters are converted into a device startup instruction set consisting of the device ID, instruction code, parameter values, and execution timestamp via the industrial communication protocol. When converting the optimal device scheduling combination into an executable instruction set, data conversion and transmission must be completed via the industrial communication protocol. The specific process is as follows: the system first matches the preset device control protocol template based on the device ID, location identifier, and startup parameters in the optimal device scheduling combination. Taking the widely used Modbus protocol in the industrial sector as an example, the device ID is mapped to the device station address in the protocol. The startup parameters (such as fan speed percentage) are converted to corresponding register values ​​according to the protocol specification. For example, 70% speed is quantized into an integer value of 7000 (corresponding to the range of 0-10000), and the target register address (such as the speed register 0x1000) is determined. The system automatically generates binary instructions that conform to the Modbus frame structure: the frame header contains the device station ID and function code (such as 06 for writing a single register), followed by the register address and parameter values ​​in the subsequent bytes, and finally a CRC checksum is appended to ensure transmission integrity. Simultaneously, the execution timestamp is converted into a delayed trigger mechanism, allowing the controller to send the instruction at the specified time. The resulting instruction set is output as a structured list. Each instruction clearly includes the device ID, binary instruction stream, actual parameter values, and precise execution time. This instruction is transmitted in real time to the device control unit via an industrial bus (such as RS-485 or Ethernet), driving the cooling device to operate on demand. The entire process uses a protocol conversion layer to accurately map decision instructions to physical signals, and an error retransmission mechanism ensures reliability.This step solves the problem of rigid device scheduling in existing technologies through a dynamic load balancing mechanism. Conditional queries ensure that device availability matches capabilities, polling allocation avoids overloading of single devices, and intensity level binding achieves precise matching of heat dissipation capabilities. The final generated instruction set optimizes the correspondence between heat dissipation resources and thermal risk distribution, reducing device energy consumption while ensuring heat dissipation effects.

[0056] In step S17, the corresponding heat dissipation device operation module is activated through the device startup instruction set.

[0057] In a specific embodiment, activating the corresponding heat dissipation device operation module through the device startup instruction set includes: Parsing the device startup instruction set into an activation signal corresponding to the device ID to generate an activation signal set; A PID algorithm is used to adjust the heat dissipation device according to the activation signal set and the temperature distribution record.

[0058] Specifically, in step S17, the system activates the cooling device operation module using the device startup instruction set. The specific processing process is as follows: the system parses each instruction in the device startup instruction set into an activation signal recognizable by the device control unit. Each instruction contains a device ID, instruction code, parameter value, and execution timestamp fields. The system then generates an activation signal set indexed by the device ID. This set data structure is a mapping of device IDs to startup parameter key-value pairs. Based on the activation signal set and temperature distribution records in a real-time database, a PID algorithm is used to dynamically adjust the operation of the cooling device. First, the input parameters are bound. The target temperature value obtains the preset temperature target value from the cooling intensity level field of the cooling demand signal based on the location identifier bound to the device ID (this value is predetermined by the mapping relationship between the cooling intensity level and the device performance curve). The real-time temperature value extracts the actual temperature value from the latest temperature distribution record using the location identifier. The historical deviation reads the temperature deviation record at the location corresponding to the device ID at consecutive timestamps. PID parameter calculation is then performed. The proportional term calculates the difference between the actual temperature and the target temperature at the current moment. The integral term accumulates the temperature deviation value of the device-bound location within a preset time period (the time period is determined by the device's thermal response speed). The differential term calculates the rate of change of temperature between the current moment and the previous timestamp. Finally, the output control is performed. The weighted sum of the three calculation results generates a correction value for the device operating parameters (the weight coefficient is calibrated through the device cooling characteristics experiment). This correction value is sent to the device control unit in real time via the industrial bus to dynamically adjust the fan speed or cooling power. This step solves the uncontrollable defects of the heat dissipation process in the existing technology through a closed-loop feedback mechanism. Instruction analysis ensures that the control signal corresponds accurately to the physical device. The PID algorithm dynamically adjusts the device output according to real-time temperature changes. The target temperature and intensity level are bound to achieve adaptive heat dissipation intensity, and finally form a complete control closed loop of prediction-decision-execution-feedback. While eliminating the risk of overheating, energy waste is avoided, and the core goal of precise temperature control of the distribution cabinet is achieved.

[0059] Reference Figure 2 The second embodiment of the present invention provides a temperature control system for a power distribution cabinet with adaptive heat dissipation, comprising: A data acquisition module is used to acquire temperature distribution records, wherein the temperature distribution records include a timestamp, a location identifier, and a temperature value; A data processing module, configured to perform data preprocessing on the temperature distribution records to obtain standard temperature data; A temperature characteristic analysis module is used to perform dynamic change analysis on the standard temperature data to obtain temperature change characteristics; A future temperature prediction module, configured to input the temperature change characteristics into a pre-built temperature change prediction model and output future temperature data; a heat dissipation demand analysis module, configured to perform a risk assessment on the future temperature data and determine whether to trigger a heat dissipation demand signal based on the assessment result; An optimal device combination module is used to call a pre-established heat dissipation resource allocation database according to the heat dissipation demand signal, obtain an optimal device scheduling combination, and generate a device startup instruction set; The execution module is used to activate the corresponding heat dissipation device operation module through the device startup instruction set.

[0060] It should be noted that the adaptive heat dissipation distribution cabinet temperature control device provided in an embodiment of the present invention is used to execute all the process steps of the adaptive heat dissipation distribution cabinet temperature control method of the above embodiment. The working principles and beneficial effects of the two correspond one to one, so they will not be repeated here.

[0061] An embodiment of the present invention further provides an electronic device. The electronic device includes: a processor, a memory, and a computer program stored in the memory and executable on the processor, such as a temperature control program for a power distribution cabinet with adaptive heat dissipation. When the processor executes the computer program, the steps of the above-mentioned embodiments of the temperature control method for a power distribution cabinet with adaptive heat dissipation are implemented, such as Figure 1 Alternatively, when the processor executes the computer program, the functions of the modules / units in the above-mentioned device embodiments are realized, such as the temperature control module of the power distribution cabinet with adaptive heat dissipation.

[0062] Exemplarily, the computer program may be divided into one or more modules / units, which are stored in the memory and executed by the processor to implement the present invention. The one or more modules / units may be a series of computer program instruction segments capable of implementing specific functions, and the instruction segments are used to describe the execution process of the computer program in the electronic device.

[0063] The electronic device may be a computing device such as a desktop computer, notebook, PDA, or smart tablet. The electronic device may include, but is not limited to, a processor and memory. Those skilled in the art will appreciate that the aforementioned components are merely examples of electronic devices and do not constitute a limitation of the electronic device. The electronic device may include more or fewer components than those described above, or a combination of certain components, or different components. For example, the electronic device may also include input / output devices, network access devices, buses, and the like.

[0064] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor. The processor is the control center of the electronic device and connects various parts of the entire electronic device using various interfaces and lines.

[0065] The memory can be used to store the computer programs and / or modules. The processor implements the various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory and accessing the data stored in the memory. The memory may primarily include a program storage area and a data storage area. The program storage area may store an operating system and at least one application required for a function (such as a sound playback function or an image playback function); the data storage area may store data generated based on the use of the mobile phone (such as audio data, a phone book, etc.). Furthermore, the memory may include high-speed random access memory and non-volatile memory, such as a hard disk, internal memory, a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, at least one disk storage device, a flash memory device, or other volatile solid-state storage device.

[0066] If the module / unit integrated into the electronic device is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention can implement all or part of the process steps in the above-mentioned method embodiments by using a computer program to instruct the relevant hardware. The computer program can be stored in a computer-readable storage medium. When executed by a processor, the computer program can implement the steps of each of the above-mentioned method embodiments. The computer program includes computer program code, which can be in source code form, object code form, executable file, or some intermediate form. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard drive, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal, and software distribution medium. It should be noted that the content of the computer-readable medium can be appropriately increased or decreased based on the requirements of legislation and patent practice in a jurisdiction. For example, in some jurisdictions, based on legislation and patent practice, computer-readable media does not include electric carrier signals and telecommunication signals.

[0067] It should be noted that the device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed across multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment. In addition, in the drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art can understand and implement the present invention without inventive effort.

[0068] The specific embodiments described above further illustrate the objectives, technical solutions, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A temperature control method for a power distribution cabinet with adaptive heat dissipation, characterized in that: include: Obtaining a temperature distribution record, wherein the temperature distribution record includes a timestamp, a location identifier, and a temperature value; Performing data preprocessing on the temperature distribution record to obtain standard temperature data; Performing dynamic change analysis on the standard temperature data to obtain temperature change characteristics; Inputting the temperature change characteristics into a pre-built temperature change prediction model to output future temperature data; Performing a risk assessment on the future temperature data, and determining whether to trigger a heat dissipation demand signal based on the assessment result; Calling a pre-established heat dissipation resource allocation database according to the heat dissipation demand signal to obtain an optimal device scheduling combination and generate a device startup instruction set; The corresponding heat dissipation device operation module is activated through the device startup instruction set.

2. The temperature control method of the power distribution cabinet with adaptive heat dissipation according to claim 1, characterized in that: The step of performing data preprocessing on the temperature distribution record to obtain standard temperature data includes: De-noising the temperature distribution record using a median filter algorithm to obtain de-noised temperature data; Using a linear interpolation algorithm to fill missing values ​​in the denoised temperature data to obtain filled temperature data; Using a moving average algorithm to smooth and correct the filled temperature data to obtain smoothed temperature data; The smoothed temperature data is standardized using a Z-score algorithm to obtain standard temperature data.

3. The temperature control method of the power distribution cabinet with adaptive heat dissipation according to claim 1, characterized in that: The dynamic change analysis of the standard temperature data to obtain temperature change characteristics includes: Using a fixed time window algorithm to divide the standard temperature data into temperature data of independent time periods to generate segmented temperature data; According to the segmented temperature data, the temperature variation amplitude is calculated by using the range statistics method and the temperature spatial gradient is calculated by using the adjacent point gradient method; The temperature variation amplitude and the temperature spatial gradient are combined into a comprehensive feature vector to generate a temperature variation feature.

4. The temperature control method of a power distribution cabinet with adaptive heat dissipation according to claim 1, characterized in that: The process of constructing the temperature change prediction model includes: Obtaining historical temperature change characteristics and inputting them into the input layer of the preliminarily constructed long short-term memory network model for training, thereby obtaining a predicted temperature output by the output layer of the long short-term memory network model; Input the predicted temperature and the pre-stored actual temperature into the mean square error loss function to calculate the loss value; According to the loss value, the gradient of the output layer of the long short-term memory network model is calculated by the back propagation algorithm, and the gradient is passed forward layer by layer by the chain rule to calculate the gradient of the parameters of each layer to obtain gradient data; Based on the gradient data and the preset learning rate, the parameters of each layer of the long short-term memory network model are updated through the Adam optimizer; The parameters of each layer are updated repeatedly until the training times of the long short-term memory network model are greater than the preset times, or when the loss value data of the long short-term memory network model is less than the preset loss threshold, the training is determined to be completed and a temperature change prediction model is obtained.

5. The temperature control method of a power distribution cabinet with adaptive heat dissipation according to claim 1, characterized in that: The risk assessment of the future temperature data and triggering a heat dissipation demand signal according to the assessment result include: Comparing the future temperature data with a preset temperature threshold, and when the future temperature data exceeds the temperature threshold, marking the location identifier corresponding to the future temperature data as a high-risk point, and generating risk distribution data, wherein the risk distribution data includes a risk location, a risk temperature, and a risk time; Multiplying the risk temperature by the risk time to obtain a risk value, and generating a priority list in descending order according to the risk values; Based on the risk location, the K-means algorithm is used to identify the dense high-temperature area and obtain the heat dissipation area; According to the heat dissipation area, a load balancing algorithm is used to assign a heat dissipation intensity level; The priority list, the heat dissipation areas, and the heat dissipation intensity levels are used as heat dissipation requirement signals.

6. The temperature control method for a power distribution cabinet with adaptive heat dissipation according to claim 5, characterized in that: The calling of a pre-established heat dissipation resource allocation database according to the heat dissipation demand signal to obtain an optimal device scheduling combination and generate a device startup instruction set includes: Based on the heat dissipation demand signal, extracting the device status data and heat dissipation parameter range of the heat dissipation area from a pre-established heat dissipation resource allocation database, and filtering out a list of available devices through conditional query based on the device status data and the heat dissipation parameter range; According to the available device list, combined with the priority list and the heat dissipation intensity level, a polling algorithm is used to perform load balancing and output an optimal device scheduling combination; According to the optimal device scheduling combination, a device startup instruction set is generated through an industrial communication protocol.

7. The temperature control method of a power distribution cabinet with adaptive heat dissipation according to claim 1, characterized in that: The activating the corresponding heat dissipation device operation module by the device startup instruction set includes: Parsing the device startup instruction set into an activation signal corresponding to the device ID to generate an activation signal set; A PID algorithm is used to adjust the heat dissipation device according to the activation signal set and the temperature distribution record.

8. A temperature control system for a power distribution cabinet with adaptive heat dissipation, characterized in that: include: A data acquisition module is used to acquire temperature distribution records, wherein the temperature distribution records include a timestamp, a location identifier, and a temperature value; A data processing module, configured to perform data preprocessing on the temperature distribution records to obtain standard temperature data; A temperature characteristic analysis module is used to perform dynamic change analysis on the standard temperature data to obtain temperature change characteristics; A future temperature prediction module, configured to input the temperature change characteristics into a pre-built temperature change prediction model and output future temperature data; a heat dissipation demand analysis module, configured to perform a risk assessment on the future temperature data and determine whether to trigger a heat dissipation demand signal based on the assessment result; An optimal device combination module is used to call a pre-established heat dissipation resource allocation database according to the heat dissipation demand signal, obtain an optimal device scheduling combination, and generate a device startup instruction set; The execution module is used to activate the corresponding heat dissipation device operation module through the device startup instruction set.

9. An electronic device, characterized in that: It includes a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, and when the processor executes the computer program, it implements the temperature control method of the power distribution cabinet with adaptive heat dissipation as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium includes a stored computer program, wherein when the computer program is running, the device where the computer-readable storage medium is located is controlled to execute the temperature control method for a power distribution cabinet with adaptive heat dissipation according to any one of claims 1 to 7.

Images