A humidity management system and method for a drainage type switchgear cabinet

Abstract

The present invention relates to the technical field of humidity control for drainage-type switchgear, and in particular to a humidity management system and method for drainage-type switchgear. By disposing a humidity sensor network within the switchgear, humidity data at different points within the switchgear is acquired, generating a humidity data matrix. This data matrix is then combined with external temperature and humidity information acquired by an environmental acquisition module to generate an environmental parameter matrix, thereby improving monitoring accuracy and overcoming the limitations of local monitoring. A real-time determination unit based on a residual neural network and a deep learning-based prediction and determination unit analyze the humidity data matrix and the environmental parameter matrix, accurately distinguishing between non-humidity, natural humidity, and accidental humidity, thereby selecting an appropriate dehumidification strategy, effectively reducing the risk of equipment failure caused by misjudgment and delayed response, and saving the energy consumption cost of the dehumidification device.

Description

Technical Field

[0001] The present invention relates to the technical field of humidity control of a drainage type switch cabinet, and in particular to a humidity management system and method for a drainage type switch cabinet. Background Art

[0002] In modern power systems, switchgear, as crucial power distribution and control equipment, is widely used in various industrial and residential applications. The stable operation of switchgear is crucial to the safety and reliability of power systems. However, in humid or rainy environments, excessive humidity can easily occur inside switchgear, leading to degraded insulation performance, increased electrical failures, and even serious accidents such as short circuits and fires.

[0003] Traditional humidity control methods primarily rely on simple dehumidification devices, such as fixed dehumidifiers or fans, which are manually or timed to control dehumidification. However, this approach suffers from slow response and limited dehumidification effectiveness in the face of complex environmental changes. Furthermore, traditional methods are unable to determine the specific humidity type within the cabinet and intelligently adjust based on real-time changes in environmental parameters, resulting in high energy consumption costs for dehumidification devices. Furthermore, humidity monitoring is limited to localized monitoring, making it difficult to effectively prevent equipment damage and failures caused by unexpected moisture. Summary of the Invention

[0004] To solve the above problems, the present invention provides a drainage-type switch cabinet humidity management system and method. By setting up a humidity sensor network in the cabinet and combining residual network training and learning with environmental acquisition parameters, the humidity type is identified to match the drainage and dehumidification strategy. The determination of humidity type and prediction of humidity conditions are realized and corresponding measures are taken, which saves the energy consumption cost of the dehumidification device and effectively avoids equipment damage and failure due to humidity.

[0005] To achieve the above object, the technical solution adopted by the present invention is:

[0006] A moisture management system for a drainage-type switch cabinet comprises a humidity sensor network, a humidity analysis module, and a drainage control module connected in sequence, and further comprises an environment acquisition module connected to the humidity analysis module;

[0007] The humidity sensor network is used to obtain humidity data at different points in the switch cabinet, construct a humidity data matrix according to the spatial position of the points, and collect the humidity data matrix of the continuous time line to construct a humidity matrix set;

[0008] The environment acquisition module is used to collect temperature data and humidity data outside the switch cabinet, and build an environmental parameter matrix based on the temperature data and humidity data outside the switch cabinet;

[0009] The humidity analysis module includes a real-time determination unit and a prediction determination unit;

[0010] The real-time determination unit is used to analyze the humidity matrix set and the environmental parameter matrix based on the residual neural network to determine the humidity type, wherein the humidity type includes non-humidity, natural humidity and accidental humidity;

[0011] The prediction and determination unit is configured to perform the following steps:

[0012] Based on deep learning, correlation analysis is performed on the humidity matrix set to predict the humidity change trend of each point in the future time period and extract potential humidity anomaly areas;

[0013] Determine whether the humidity abnormal area coincides with the switch in the cabinet and send the coincidence determination result to the drainage control module;

[0014] The drainage control module is used to match the preset drainage strategy according to the moisture type and the coincidence judgment result, and control the drainage facilities in the cabinet to drain and dehumidify according to the drainage strategy.

[0015] Furthermore, the humidity sensor network includes a processor and a plurality of humidity sensors disposed in different areas within the switch cabinet, wherein the processor is connected to the humidity sensors;

[0016] The humidity sensor is used to obtain humidity data;

[0017] The processor is used to record the spatial position information of all humidity sensors, integrate all humidity data with the same timestamp and arrange them according to the spatial position information of the humidity sensors to construct a humidity data matrix.

[0018] Furthermore, constructing the environmental parameter matrix based on the temperature data and humidity data outside the switch cabinet includes the following steps:

[0019] Obtain temperature and humidity data at several different points on the outer surface of the switchgear;

[0020] The temperature data and humidity data are arranged in the order of preset points to construct the environmental parameter matrix.

[0021] Furthermore, the analysis of the humidity matrix set and the environmental parameter matrix based on the residual neural network includes the following steps:

[0022] Performing data preprocessing on the humidity matrix set and the environmental parameter matrix to generate a preprocessed data matrix, wherein the data preprocessing includes data cleaning and normalization;

[0023] A residual neural network model is constructed based on the preprocessed data matrix, and high-dimensional features are extracted layer by layer through several residual blocks to generate feature maps;

[0024] The feature map is weightedly analyzed through the attention mechanism to generate the moisture type judgment result.

[0025] Furthermore, the residual neural network model is trained by the following steps:

[0026] Construct a training dataset, take the historical humidity matrix set and the environmental parameter matrix as input data, and take the humidity type and humidity change trend as target output to generate a labeled data sample set;

[0027] Initialize the parameters of the residual neural network model, use the stochastic gradient descent method for initial training, and calculate the error of each iteration through the cross entropy loss function;

[0028] Perform error back propagation training on the residual neural network model by adjusting the weight parameters of the residual block layer by layer;

[0029] The trained residual neural network model is verified by the K-fold cross-validation method.

[0030] Furthermore, the formula of the residual neural network model is as follows:

[0031] ;

[0032] in, is the input data, including the preprocessed data matrix; Output results to the network, including the determination of humidity types or the prediction of humidity change trends; and is the weight matrix of the residual block; and is the bias term; is the rectified linear unit activation function; is the Sigmoid activation function.

[0033] Furthermore, the attention mechanism is trained by the following steps:

[0034] Taking the historical humidity matrix set, environmental parameter matrix and corresponding humidity type as input data, a time series feature set is constructed through time series to generate a training sample set;

[0035] Initialize the query, key, and value matrices of the attention mechanism model, calculate feature weights for the input data, and assign weights based on the relative importance of different humidity feature points;

[0036] Based on the training samples and weight calculation results, continue to adjust the query, key, and value matrices.

[0037] Furthermore, the drainage facilities in the cabinet include: a drying device, a fan, a plurality of vents and a drain port at the bottom of the cabinet.

[0038] Furthermore, the preset drainage strategy includes:

[0039] If the humidity type is non-humidity and the coincidence determination result is no coincidence, turn off the drying and dehumidifying device, fan and cabinet bottom drain, and open the vents;

[0040] If the humidity type is non-humidity and the coincidence determination result is coincidence, turn on the drying and dehumidifying device and close the fan, vents, and drain outlet at the bottom of the cabinet;

[0041] If the humidity type is natural, turn on the drying and dehumidifying device and close the fan, vents and drain outlet at the bottom of the cabinet;

[0042] If the humidity is accidental, turn on the fan, vents, and drains at the bottom of the cabinet, and turn off the drying and dehumidifying devices.

[0043] A method for humidity management of a drainage-type switch cabinet, applied to the aforementioned humidity management system for a drainage-type switch cabinet, comprises the following steps:

[0044] Obtain humidity data at different points in the switch cabinet, construct a humidity data matrix based on the spatial position of the points, and collect the humidity data matrix of the continuous time line to construct a humidity matrix set;

[0045] Collect temperature and humidity data outside the switch cabinet, and build an environmental parameter matrix based on the temperature and humidity data outside the switch cabinet;

[0046] Analyzing the humidity matrix set and the environmental parameter matrix based on the residual neural network to determine the humidity type, which includes non-humidity, natural humidity and accidental humidity;

[0047] Based on deep learning, correlation analysis is performed on the humidity matrix set to predict the humidity change trend of each point in the future time period and extract potential humidity anomaly areas;

[0048] Determine whether the abnormal humidity area coincides with the switch inside the cabinet;

[0049] The preset drainage strategy is matched according to the moisture type and the coincidence judgment result, and the drainage facilities in the control cabinet are controlled according to the drainage strategy to drain and dehumidify.

[0050] The beneficial effects of the present invention are as follows: the present invention solves the problems of slow response speed and limited dehumidification effect of traditional humidity control methods in complex environments by setting up a humidity sensor network, an environmental acquisition module and an intelligent humidity analysis module. The humidity sensor network can accurately obtain humidity data at different points inside the switch cabinet, construct a humidity data matrix, and generate an environmental parameter matrix in combination with the external temperature and humidity information provided by the environmental acquisition module. This data acquisition method not only improves the accuracy of monitoring, but also overcomes the local monitoring limitations of traditional monitoring. The real-time judgment unit based on the residual neural network and the prediction and judgment unit based on deep learning perform in-depth analysis of the collected humidity and environmental data. Through the real-time judgment unit, this solution can accurately distinguish different types of humidity conditions, including non-humidity, natural humidity and accidental humidity, so as to make corresponding dehumidification strategy selections for different types of humidity. This intelligent analysis method effectively reduces the risk of equipment failure due to misjudgment or delayed response. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 It is a structural diagram of a drainage-type switch cabinet humidity management system in the present invention.

[0052] Figure 2 It is a flow chart of the steps of analyzing the humidity matrix set and the environmental parameter matrix based on the residual neural network in the present invention. DETAILED DESCRIPTION

[0053] See also Figure 1-2 As shown, the present invention relates to a moisture management system for a drainage type switch cabinet, comprising a humidity sensor network, a humidity analysis module and a drainage control module connected in sequence, and also comprising an environment acquisition module, wherein the environment acquisition module is connected to the humidity analysis module;

[0054] The humidity sensor network is used to obtain humidity data at different points in the switch cabinet, construct a humidity data matrix according to the spatial position of the points, and collect the humidity data matrix of the continuous time line to construct a humidity matrix set;

[0055] The environment acquisition module is used to collect temperature data and humidity data outside the switch cabinet, and build an environmental parameter matrix based on the temperature data and humidity data outside the switch cabinet;

[0056] The humidity analysis module includes a real-time determination unit and a prediction determination unit;

[0057] The real-time determination unit is used to analyze the humidity matrix set and the environmental parameter matrix based on the residual neural network to determine the humidity type, wherein the humidity type includes non-humidity, natural humidity and accidental humidity;

[0058] The prediction and determination unit is configured to perform the following steps:

[0059] Based on deep learning, correlation analysis is performed on the humidity matrix set to predict the humidity change trend of each point in the future time period and extract potential humidity anomaly areas;

[0060] Determine whether the humidity abnormal area coincides with the switch in the cabinet and send the coincidence determination result to the drainage control module;

[0061] The drainage control module is used to match the preset drainage strategy according to the moisture type and the coincidence judgment result, and control the drainage facilities in the cabinet to drain and dehumidify according to the drainage strategy.

[0062] It should be noted that the humidity sensor network consists of several humidity sensors evenly distributed across key locations within the switchgear, including the top and bottom of the switchgear, and around key electrical components. Each humidity sensor is equipped with an independent embedded processing unit that acquires humidity data in real time and transmits the data to the central processing unit. Based on the sensor location data, the central processing unit constructs a humidity data matrix according to a preset three-dimensional coordinate system. The data is sampled at regular intervals to generate a continuous timeline humidity matrix set. The environmental acquisition module uses highly sensitive temperature and humidity sensors to collect real-time temperature and humidity data from the switchgear's external environment. This module connects to the humidity analysis module, transmits the collected external environmental data to the system's central database, and constructs an environmental parameter matrix based on the external temperature and humidity data. The environmental parameter matrix supplements the humidity data matrix, providing environmental context for humidity changes, thereby improving data analysis accuracy. The humidity analysis module includes a built-in computing unit, including a real-time determination unit and a predictive determination unit. The real-time determination unit uses a hardware accelerator based on a residual neural network to simultaneously analyze the humidity matrix set and the environmental parameter matrix to determine the humidity type within the switchgear. The real-time judgment unit can quickly process large amounts of environmental and humidity data, complete calculation tasks within milliseconds, and classify the data according to the humidity type to ensure real-time performance.

[0063] Furthermore, the humidity sensor network includes a processor and a plurality of humidity sensors disposed in different areas within the switch cabinet, wherein the processor is connected to the humidity sensors;

[0064] The humidity sensor is used to obtain humidity data;

[0065] The processor is used to record the spatial position information of all humidity sensors, integrate all humidity data with the same timestamp and arrange them according to the spatial position information of the humidity sensors to construct a humidity data matrix.

[0066] In some embodiments, each humidity sensor has independent spatial location information that identifies its specific location within the switchgear. These sensors are installed in key areas, such as the main switch, branch wiring terminals, and around the control module, to comprehensively cover the humidity conditions within the switchgear. All humidity sensors synchronously collect humidity data at preset time intervals. The sensor's embedded microprocessor performs preliminary data processing, including data filtering and signal denoising, to reduce errors. A central processing unit (CPU) integrates the humidity data from each humidity sensor. Specifically, the processor receives and records the data from each sensor and its corresponding spatial location information. To ensure data timeliness and consistency, the CPU timestamps the data from different sensors and integrates humidity data with the same timestamp to form a time series dataset. During the data integration process, the CPU arranges the data from each sensor according to three-dimensional spatial coordinates based on the humidity sensor's spatial location information to construct a humidity data matrix. This matrix presents the humidity distribution within the switchgear as a three-dimensional grid, with the data at each grid node representing the humidity value at a specific location at a specific point in time. By constructing a humidity data matrix with a continuous timeline, the system can monitor humidity trends in real time, providing a foundation for subsequent data analysis and prediction.

[0067] Furthermore, constructing the environmental parameter matrix based on the temperature data and humidity data outside the switch cabinet includes the following steps:

[0068] Obtain temperature and humidity data at several different points on the outer surface of the switchgear;

[0069] The temperature data and humidity data are arranged in the order of preset points to construct the environmental parameter matrix.

[0070] In some embodiments, a number of high-precision temperature and humidity sensors installed on the exterior of the switchgear acquire environmental data at various locations in real time. These sensors are evenly distributed across various exterior areas of the switchgear, such as the top, sides, and bottom, according to a pre-set layout rule to comprehensively capture the temperature and humidity conditions of the switchgear's environment. Each sensor utilizes highly sensitive components that accurately respond to even the smallest temperature and humidity changes, ensuring data reliability and accuracy. The system then arranges the collected temperature and humidity data according to a pre-set point order. During data collection, each sensor's data is accompanied by its specific spatial coordinate information. Based on this coordinate information, the central processor sorts the temperature and humidity data from all sensors into a pre-set order. This sorting is based on the switchgear's geometric structure to ensure that the spatial distribution of the data in the environmental parameter matrix remains consistent with the actual locations of the collection points. Next, an environmental parameter matrix is constructed based on the sorted data. The environmental parameter matrix is a two-dimensional array, where each row represents the temperature and humidity values at a specific collection point, and each column corresponds to a different collection time point. This matrix structure enables the system to analyze environmental changes at different points in a continuous time period in real time. Through the dynamic changes of each data in the matrix, the potential impact of the external environment on the humidity inside the switchgear can be effectively monitored and predicted.

[0071] Furthermore, the analysis of the humidity matrix set and the environmental parameter matrix based on the residual neural network includes the following steps:

[0072] Performing data preprocessing on the humidity matrix set and the environmental parameter matrix to generate a preprocessed data matrix, wherein the data preprocessing includes data cleaning and normalization;

[0073] A residual neural network model is constructed based on the preprocessed data matrix, and high-dimensional features are extracted layer by layer through several residual blocks to generate feature maps;

[0074] The feature map is weightedly analyzed through the attention mechanism to generate the moisture type judgment result.

[0075] In some embodiments, the humidity matrix and the environmental parameter matrix are first preprocessed to ensure data accuracy and consistency. The first step in preprocessing is data cleaning, which involves detecting and removing possible outliers or sensor failure data. For example, sudden changes in values and missing data are imputed or smoothed. Next, the processed data is normalized to reduce dimensionality differences between different data sources. Normalization uses the Min-Max method to map all data to the range [0, 1], generating a preprocessed data matrix that provides a unified input format for subsequent neural network analysis. Based on this preprocessed data matrix, a residual neural network model is constructed. The core of this model is to extract high-dimensional features layer by layer using multiple residual blocks. Each residual block consists of two layers of convolution operations with a 3x3 kernel size, designed to extract local spatial features. Batch normalization is performed after each convolution layer to accelerate convergence and prevent vanishing gradients. Residual blocks use shortcut connections to directly add inputs to outputs, forming skip connections to maintain feature consistency and prevent degradation in deep networks. The residual neural network model, based on the complexity of the input data, employs a strategy of increasing depth layer by layer to generate a feature map that describes the underlying feature relationships between different humidity types. This feature map undergoes further weighted analysis using an attention mechanism. The attention mechanism aims to enhance key features and suppress irrelevant or noisy data. Specifically, the attention mechanism calculates the importance weight of each feature point using the feature channels in the feature map, which are the query, key, and value. Weights are generated based on the similarity between feature points and normalized using a softmax function. Finally, the attention mechanism rescales the feature map based on the weighted results, highlighting data points that are critical to determining humidity type. Based on this analysis process, the system generates the humidity type determination results, which include three categories: non-humidity, naturally occurring humidity, and accidental humidity. These determinations serve as the basis for subsequent drainage strategy selection, ensuring a scientific and targeted drainage solution and improving the efficiency and reliability of humidity control in switchgear under variable environmental conditions.

[0076] Furthermore, the residual neural network model is trained by the following steps:

[0077] Construct a training dataset, take the historical humidity matrix set and the environmental parameter matrix as input data, and take the humidity type and humidity change trend as target output to generate a labeled data sample set;

[0078] Initialize the parameters of the residual neural network model, use the stochastic gradient descent method for initial training, and calculate the error of each iteration through the cross entropy loss function;

[0079] Perform error back propagation training on the residual neural network model by adjusting the weight parameters of the residual block layer by layer;

[0080] The trained residual neural network model is verified by the K-fold cross-validation method.

[0081] In some embodiments, a training dataset is first constructed. This dataset consists of a historical humidity matrix and an environmental parameter matrix as input data sources. Specifically, the humidity matrix records the humidity distribution inside the switchgear at different time points, while the environmental parameter matrix captures the changes in the external temperature and humidity environment during the corresponding time period. The target output consists of two components: the humidity type inside the switchgear, which includes three categories: non-humidity, natural humidity, and accidental humidity; and a predicted humidity trend. To ensure the quality of the training data, all data samples are manually reviewed and labeled to ensure label accuracy, ultimately forming a labeled data sample set. After the dataset is constructed, the parameters of the residual neural network model are initialized. Parameter initialization uses the Xavier or He initialization method to optimize the initial state of the neural network and ensure that the gradient remains stable during forward and backward propagation. Stochastic gradient descent (SGD) is used for model optimization, which effectively reduces computational overhead on large datasets. During the initial training phase, the system calculates the output through forward propagation and uses the cross-entropy loss function to evaluate the difference between the output and the target label. The cross-entropy loss function is suitable for multi-classification problems. It quantifies the model's classification accuracy and provides a basis for subsequent error adjustment. Next, the model enters the error backpropagation training phase. The backpropagation algorithm calculates gradient information and adjusts the weight parameters of the residual block layer by layer. Specifically, in each iteration, the system calculates the partial derivative of the loss function with respect to each weight parameter and updates the parameters based on the learning rate to gradually reduce prediction error. The residual block preserves essential information about the input features through shortcut connections, enabling efficient feature learning in deep networks and avoiding the vanishing gradient problem. Finally, the trained model is validated using the K-fold cross-validation method. This validation process divides the dataset into K subsets, selecting one subset at a time as the validation set and the remaining subsets as the training set. Multiple rounds of iterative validation are performed to comprehensively evaluate the model's generalization ability. The final model performance is determined by calculating metrics such as accuracy, precision, and recall from each validation round. If the validation results do not meet expected standards, the system will adjust the network structure or hyperparameters based on the validation feedback to further optimize the model.

[0082] Furthermore, the formula of the residual neural network model is as follows:

[0083] ;

[0084] in, is the input data, including the preprocessed data matrix; Output results to the network, including the determination of humidity types or the prediction of humidity change trends; and is the weight matrix of the residual block; and is the bias term; is the rectified linear unit activation function; is the Sigmoid activation function.

[0085] Specifically, input data Represents the humidity matrix set and environmental parameter matrix after data preprocessing, which are used as input features of the neural network. The input data has been normalized and cleaned to ensure data consistency and stability. The weight matrix in the residual block is and , the parameters used for convolution operations. The learning of the weight matrix is optimized by gradient descent during training, aiming to maximize the representation ability of the features. and The function of is to ensure that the neural network can produce appropriate output responses even when there is no input signal, and to prevent the output value of the activation function from being fixed in a certain state. Nonlinear activation is performed after each convolutional layer, truncating negative values to zero, effectively enhancing the model's nonlinear representation capabilities. The Sigmoid activation function is used to normalize the final output value to the range [0, 1] and is suitable for probabilistic prediction tasks, such as humidity classification. In the above formula, the design of the residual block is a key feature of the model. By introducing shortcut connections between network layers, input data is passed directly to the output of the deep network, thus avoiding the gradient vanishing problem. This structure allows the network to extract deep features while retaining the original information of the input data, improving the model's learning efficiency on complex datasets. By stacking multiple layers of residual blocks, the model is able to gradually capture high-dimensional correlation features in humidity and environmental data, providing accurate decision-making basis for subsequent dehumidification and drainage strategies.

[0086] Furthermore, the attention mechanism is trained by the following steps:

[0087] Taking the historical humidity matrix set, environmental parameter matrix and corresponding humidity type as input data, a time series feature set is constructed through time series to generate a training sample set;

[0088] Initialize the query, key, and value matrices of the attention mechanism model, calculate feature weights for the input data, and assign weights based on the relative importance of different humidity feature points;

[0089] Based on the training samples and weight calculation results, continue to adjust the query, key, and value matrices.

[0090] In some embodiments, the first step in the training process is to construct a time-series feature set. This data is obtained by obtaining a historical humidity matrix set, an environmental parameter matrix, and its corresponding humidity type, and using this data as the input data source. When constructing the training sample set, the continuous humidity and environmental data are organized into a time-series data format using time series analysis methods. Specifically, each time-series sample contains a humidity matrix for multiple time steps and the corresponding external environmental parameters at that time, ensuring that the sample set reflects the dynamic characteristics of humidity changes. These time-series samples not only record past humidity information but also clearly annotate the humidity type at that time, which is used for supervised learning of the subsequent model. Secondly, during the training phase, the core parameters of the attention mechanism model are initialized, including the query matrix (Query), key matrix (Key), and value matrix (Value). These matrices are initialized using a random distribution to ensure parameter diversity. The query, key, and value matrices are used to calculate feature weights for the input data. During training, the model encodes the input humidity matrix and environmental parameter matrix, generating corresponding query, key, and value vectors for each feature point. The query and key vectors are used to calculate similarities between feature points to form attention weights. The model calculates the relative importance of each humidity feature point through an inner product operation and then normalizes the weights using the Softmax function to determine the relative importance of each feature point within the overall dataset. Finally, based on the generated training samples and the previously calculated feature weights, the parameters of the query, key, and value matrices are further adjusted. In each iteration, the model calculates the prediction error of the attention weights through backpropagation and adjusts the matrix parameters based on the gradient information to reduce the error. This process ensures that the model accurately identifies the importance of different humidity feature points and assigns appropriate weights to highlight features that are critical to determining humidity type. At the end of training, the attention mechanism model is able to dynamically focus on the correlation between different humidity and environmental feature points, improving the accuracy and robustness of humidity type prediction.

[0091] Furthermore, the drainage facilities in the cabinet include: a drying device, a fan, a plurality of vents and a drain port at the bottom of the cabinet.

[0092] In some embodiments, a drying device utilizes an integrated heating unit and intelligent control module, installed at strategic locations within the switchgear. By precisely controlling the power output of the heating element, the drying device automatically activates when humidity is too high, raising the temperature of the air inside the switchgear and accelerating the evaporation and removal of moisture. Built-in temperature and humidity sensors monitor the ambient conditions within the switchgear in real time, ensuring that the heating process does not cause localized overheating or damage electronic equipment. In emergency situations involving high humidity, the drying device can operate in high-power mode to quickly reduce humidity to a safe level. Secondly, a fan system accelerates air circulation within the switchgear, supplementing the drying device's dehumidification effectiveness. Fans are installed in the upper and lower areas of the switchgear, utilizing a multi-directional design to effectively circulate air and direct humid air toward the drainage channel. Furthermore, the fan speed and operating mode can be intelligently adjusted by the drainage control module to adapt to varying humidity conditions. For example, when the humidity is naturally low, the fan operates at a medium speed to enhance air circulation; in the case of unexpected humidity, the fan operates at a high speed to rapidly remove moisture. Several ventilation openings are located on the top and sides of the cabinet, and feature sealed regulating valves to prevent the free entry of outside air under normal conditions. The opening and closing status of the ventilation openings is controlled in real time by a humidity analysis module, dynamically adjusting according to humidity changes to ensure that the humidity inside the cabinet is always maintained within a stable range. The ventilation openings are also equipped with dust screens and filters to prevent external contaminants from entering the cabinet. Finally, the drain outlet at the bottom of the cabinet is a core component of the drainage facility, used to collect and discharge excess condensed water from the cabinet. The drain outlet uses an automatically opening and closing electronically controlled valve that automatically opens to drain accumulated water when the humidity sensor detects condensation, and automatically closes when the humidity returns to normal levels. The drain outlet is also equipped with an anti-clogging device to ensure long-term operational reliability.

[0093] Furthermore, the preset drainage strategy includes:

[0094] If the humidity type is non-humidity and the coincidence determination result is no coincidence, turn off the drying and dehumidifying device, fan and cabinet bottom drain, and open the vents;

[0095] If the humidity type is non-humidity and the coincidence determination result is coincidence, turn on the drying and dehumidifying device and close the fan, vents, and drain outlet at the bottom of the cabinet;

[0096] If the humidity type is natural, turn on the drying and dehumidifying device and close the fan, vents and drain outlet at the bottom of the cabinet;

[0097] If the humidity is accidental, turn on the fan, vents, and drains at the bottom of the cabinet, and turn off the drying and dehumidifying devices.

[0098] In some embodiments, when the humidity type is determined to be non-humidity and the overlap determination result indicates no overlap, the overall humidity within the switchgear is normal and the predicted abnormal humidity area does not involve critical equipment. In this case, to save energy, the system shuts down all dehumidification equipment, including drying units, fans, and cabinet drains, while simultaneously opening the vents to maintain stable internal humidity through natural air circulation. This strategy ensures that under normal humidity conditions, unnecessary energy consumption is avoided while maintaining good air exchange. If the humidity type is non-humidity, but the overlap determination result indicates an abnormal humidity area that overlaps with critical equipment, the system adopts a more cautious approach. In this case, the system activates the drying units to accelerate dehumidification to prevent potential hazards posed by humidity to critical equipment. Simultaneously, the fans, vents, and cabinet drains are shut down to prevent unintended moisture from entering. In the case of natural humidity, the increase in humidity is due to natural environmental fluctuations, such as seasonal humidity increases. In this case, the system activates only the drying units to reduce humidity in the air through heating. All other equipment, including fans, vents, and cabinet drains, remains shut down to prevent external humid air from entering the cabinet and ensure effective dehumidification. This strategy focuses on gently regulating overall humidity to maintain a stable environment within the switchgear. When the humidity type is determined to be unexpected, it indicates a potential sudden increase in humidity within the switchgear, possibly caused by equipment failure or a sudden change in the external environment. The system immediately initiates all aggressive dehumidification measures, including fans, vents, and under-cabinet drains, to rapidly remove moisture from the cabinet. During this time, the drying unit remains disabled to prevent excessive energy consumption due to heating. Fans operate at high speeds, vents are fully opened to ensure rapid exhaust of moisture, and under-cabinet drains continue to remove accumulated water until humidity returns to normal. These pre-set drainage strategies, based on intelligent judgment of humidity type and overlap determination results, enable refined management in different humidity scenarios, ensuring the safe operation of the switchgear. Each strategy is scientifically designed to minimize potential threats to equipment posed by humidity while minimizing energy consumption.

[0099] The present invention also includes a method for managing humidity in a drainage-type switch cabinet, comprising the following steps:

[0100] Obtain humidity data at different points in the switch cabinet, construct a humidity data matrix based on the spatial position of the points, and collect the humidity data matrix of the continuous time line to construct a humidity matrix set;

[0101] Collect temperature and humidity data outside the switch cabinet, and build an environmental parameter matrix based on the temperature and humidity data outside the switch cabinet;

[0102] Analyzing the humidity matrix set and the environmental parameter matrix based on the residual neural network to determine the humidity type, which includes non-humidity, natural humidity and accidental humidity;

[0103] Based on deep learning, correlation analysis is performed on the humidity matrix set to predict the humidity change trend of each point in the future time period and extract potential humidity anomaly areas;

[0104] Determine whether the abnormal humidity area coincides with the switch inside the cabinet;

[0105] The preset drainage strategy is matched according to the moisture type and the coincidence judgment result, and the drainage facilities in the control cabinet are controlled according to the drainage strategy to drain and dehumidify.

[0106] The above embodiments are merely descriptions of preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary engineering technicians in this field should fall within the scope of protection determined by the claims of the present invention.

Claims

1. A drainage type switch cabinet humidity management system, characterized in that: It includes a humidity sensor network, a humidity analysis module and a drainage control module connected in sequence, and also includes an environment acquisition module, which is connected to the humidity analysis module; The humidity sensor network is used to obtain humidity data at different points in the switch cabinet, construct a humidity data matrix according to the spatial position of the points, and collect the humidity data matrix of the continuous time line to construct a humidity matrix set; The environment acquisition module is used to collect temperature data and humidity data outside the switch cabinet, and build an environmental parameter matrix based on the temperature data and humidity data outside the switch cabinet; The humidity analysis module includes a real-time determination unit and a prediction determination unit; The real-time determination unit is used to analyze the humidity matrix set and the environmental parameter matrix based on the residual neural network to determine the humidity type, wherein the humidity type includes non-humidity, natural humidity and accidental humidity; The prediction and determination unit is configured to perform the following steps: Based on deep learning, correlation analysis is performed on the humidity matrix set to predict the humidity change trend of each point in the future time period and extract potential humidity anomaly areas; Determine whether the humidity abnormal area coincides with the switch in the cabinet and send the coincidence determination result to the drainage control module; The drainage control module is used to match the preset drainage strategy according to the moisture type and the coincidence judgment result, and control the drainage facilities in the cabinet to drain and dehumidify according to the drainage strategy; The analysis of the humidity matrix set and the environmental parameter matrix based on the residual neural network includes the following steps: Performing data preprocessing on the humidity matrix set and the environmental parameter matrix to generate a preprocessed data matrix, wherein the data preprocessing includes data cleaning and normalization; A residual neural network model is constructed based on the preprocessed data matrix, and high-dimensional features are extracted layer by layer through several residual blocks to generate feature maps; The feature map is weightedly analyzed through the attention mechanism to generate the moisture type determination result; The formula of the residual neural network model is as follows: ； in, is the input data, including the preprocessed data matrix; Output results to the network, including the determination of humidity types or the prediction of humidity change trends; and is the weight matrix of the residual block; and is the bias term; is the rectified linear unit activation function; is the Sigmoid activation function; The residual neural network model is trained by the following steps: Construct a training dataset, take the historical humidity matrix set and the environmental parameter matrix as input data, and take the humidity type and humidity change trend as target output to generate a labeled data sample set; Initialize the parameters of the residual neural network model, use the stochastic gradient descent method for initial training, and calculate the error of each iteration through the cross entropy loss function; Perform error back propagation training on the residual neural network model by adjusting the weight parameters of the residual block layer by layer; The trained residual neural network model is verified by the K-fold cross-validation method; The attention mechanism is trained by the following steps: Taking the historical humidity matrix set, environmental parameter matrix and corresponding humidity type as input data, a time series feature set is constructed through time series to generate a training sample set; Initialize the query, key, and value matrices of the attention mechanism model, calculate feature weights for the input data, and assign weights based on the relative importance of different humidity feature points; Based on the training samples and weight calculation results, continue to adjust the query, key and value matrices; The drainage facilities in the cabinet include: a drying device, a fan, several vents and a drain at the bottom of the cabinet; The preset drainage strategy includes: If the humidity type is non-humidity and the coincidence determination result is no coincidence, turn off the drying and dehumidifying device, fan and cabinet bottom drain, and open the vents; If the humidity type is non-humidity and the coincidence determination result is coincidence, turn on the drying and dehumidifying device and close the fan, vents, and drain outlet at the bottom of the cabinet; If the humidity type is natural, turn on the drying and dehumidifying device and close the fan, vents and drain outlet at the bottom of the cabinet; If the humidity is accidental, turn on the fan, vents, and drains at the bottom of the cabinet, and turn off the drying and dehumidifying devices.

2. A drainage switch cabinet humidity management system according to claim 1, characterized in that: The humidity sensor network includes a processor and a plurality of humidity sensors arranged in different areas of the switch cabinet, wherein the processor is connected to the humidity sensors; The humidity sensor is used to obtain humidity data; The processor is used to record the spatial position information of all humidity sensors, integrate all humidity data with the same timestamp and arrange them according to the spatial position information of the humidity sensors to construct a humidity data matrix.

3. The moisture management system for a drainage switch cabinet according to claim 1, characterized in that: The construction of the environmental parameter matrix based on the temperature data and humidity data outside the switch cabinet includes the following steps: Obtain temperature and humidity data at several different points on the outer surface of the switchgear; The temperature data and humidity data are arranged in the order of preset points to construct the environmental parameter matrix.

4. A method for humidity management of a drainage type switch cabinet, applied to a humidity management system of a drainage type switch cabinet according to any one of claims 1 to 3, characterized in that: The following steps are involved: Obtain humidity data at different points in the switch cabinet, construct a humidity data matrix based on the spatial position of the points, and collect the humidity data matrix of the continuous time line to construct a humidity matrix set; Collect temperature and humidity data outside the switch cabinet, and build an environmental parameter matrix based on the temperature and humidity data outside the switch cabinet; The humidity matrix set and the environmental parameter matrix are analyzed based on the residual neural network to determine the humidity type, which includes non-humidity, natural humidity and accidental humidity; Based on deep learning, correlation analysis is performed on the humidity matrix set to predict the humidity change trend of each point in the future time period and extract potential humidity anomaly areas; Determine whether the abnormal humidity area coincides with the switch inside the cabinet; The preset drainage strategy is matched according to the moisture type and the coincidence judgment result, and the drainage facilities in the control cabinet are controlled according to the drainage strategy to drain and dehumidify.

Images