Environment-friendly and energy-saving transformer substation

Abstract

The invention relates to an environment-friendly and energy-saving transformer substation. A high-precision sensor is used for monitoring temperature and humidity to predict the humidity degree of electrical equipment of a power transformation cabinet; and controlling the heating temperature and the heating duration of the heater according to the humidity degree so as to dehumidify the electrical equipment. The dilemma that the dehumidification effect cannot be achieved or energy is wasted due to the fact that the relation between the humidity degree and the heating parameters cannot be coordinated and the heat control accuracy of the heater is low in the prior art is solved. In addition, the humidity degree is divided according to whether the electrical equipment breaks down or not, it is clear that the aim of heating and dehumidification is to ensure that the electrical equipment does not break down, and the heating and dehumidification effect and the control precision are improved.

Description

Technical Field

[0001] The invention belongs to the technical field of substations and relates to an environmentally friendly and energy-saving substation. Background Art

[0002] The insulation performance of electrical equipment is crucial to the safe and stable operation of substations. When the air humidity is high, moisture will adhere to the insulating surface of electrical equipment, such as transformer windings and insulators. This moisture will reduce the insulation resistance of the insulating material and deteriorate the insulation performance. For example, in a humid environment, a thin layer of water film may form on the surface of an ordinary insulating porcelain bottle due to condensation of water vapor, and its insulation strength will be greatly reduced, which is easy to cause insulation breakdown and other faults, thereby threatening the normal operation of the substation. By heating and dehumidification, the air humidity can be reduced and the adhesion of moisture on the insulating surface can be reduced, thereby maintaining the good insulation performance of the insulating material and ensuring that the electrical equipment can operate safely and reliably.

[0003] However, the existing technology has low accuracy in controlling the heat of the heater, which is reflected in the heating temperature and heating time: when the heating temperature and heating time are too low, the water vapor may not be dried out and the dehumidification effect cannot be achieved; when the heating temperature and heating time are too high, the power cord is easily burned out, and energy is greatly wasted, which does not meet the requirements of energy conservation and environmental protection. Therefore, in order to coordinate the dehumidification effect with the requirements of energy conservation and environmental protection, it is necessary to develop a technology that can accurately control the heating parameters. Summary of the invention

[0004] In order to solve the above-mentioned problems existing in the prior art, the present invention provides an environmentally friendly and energy-saving substation, which aims to match different heating parameters according to the humidity conditions of the substation cabinet by combining sensor technology, big data analysis technology and intelligent control technology to improve the accuracy of heating control and achieve the purpose of coordinating the dehumidification effect with energy saving and environmental protection.

[0005] The purpose of the present invention can be achieved through the following technical solutions: The present application provides an environmentally friendly and energy-saving substation, including an environmental monitoring module, an environmental analysis module, a parameter matching module and an intelligent control module, wherein the environmental monitoring module, the environmental analysis module, the parameter matching module and the intelligent control module are communicatively connected, wherein: The environmental monitoring module uses a high-precision sensor to monitor the environmental data inside the substation cabinet, and the environmental data includes the ambient temperature and ambient humidity; The environmental analysis module predicts the humidity level of the electrical equipment in the substation cabinet according to the environmental data; the humidity level is divided according to whether it causes the electrical equipment to malfunction; The parameter matching module matches the heating parameters of the heater according to the humidity level, wherein the heating parameters include heating temperature and heating duration; The intelligent control module dehumidifies the electrical equipment of the transformer cabinet according to the matched heating parameters.

[0006] Furthermore, predicting the humidity level of the electrical equipment of the substation cabinet according to the environmental data comprises the following steps: Monitor the environmental data inside the substation cabinet and the condensation water data samples on the surface of electrical equipment during the historical period; Mark and classify data samples according to whether electrical equipment has failed; Determine the critical point of surface condensation water in electrical equipment based on the classification to determine whether there is a fault; Calculate the relative difference between all surface condensation water samples and the critical point; Using a cluster analysis method to divide the data sample into a number of clusters according to the relative difference, and determining the wetness according to the numerical range of the relative difference of each cluster; The artificial intelligence model was trained using environmental data as the independent variable and humidity level as the dependent variable; The newly collected environmental data is input into the trained artificial intelligence model to output the predicted humidity level of the electrical equipment in the substation cabinet.

[0007] Furthermore, the relative difference is calculated as follows: , In the formula, R s Relative difference. S 2 represents a sample of condensation water on a certain surface; S 1 represents the critical point of surface condensation.

[0008] Furthermore, the cluster analysis method specifically comprises the following steps: Selecting a number of data samples as cluster centers, wherein the cluster centers include relative difference samples of critical points of surface condensation water; Calculate the Euclidean distance between all data samples and the cluster center, and assign each data sample to the cluster center with the smallest Euclidean distance to form a cluster; Calculate the average value of each cluster data sample and use the average value as the new cluster center; Repeat the process of calculating the Euclidean distance to assign clusters and calculating the average value until the new cluster center no longer changes, and determine the final cluster; The wetness degree is determined according to the numerical range of the relative difference of the final clusters. The wetness degree is based on the critical point of the surface condensation water, and a value greater than the critical point is a positive value, and a value less than the critical point is a negative value.

[0009] Furthermore, the artificial intelligence model is configured as a BP artificial neural network model, comprising the following steps: Build the network structure and determine the input layer, hidden layer and output layer settings to adapt to the environmental data and humidity prediction results; Preprocessing the historical environmental data used for training and the corresponding electrical equipment humidity degree data, wherein the preprocessing includes normalization processing, outlier processing and missing value processing; Initialize network parameters, including random settings of connection weights and thresholds; The preprocessed data is input into the network through forward propagation calculation to obtain a preliminary predicted moisture level value; Back propagation is performed based on the error between the predicted value and the actual value to adjust the network parameters; Repeat the process of forward propagation and back propagation until the error is lower than the set threshold, completing the post-training evaluation and verification of the model's accuracy.

[0010] Furthermore, matching the heating parameters of the heater according to the humidity level comprises the following steps: Set target wetness level; Subtract the target humidity level from the current humidity level to obtain the target dehumidification level; Set the heating temperature according to the current humidity level; The target heating time is determined by the target dehumidification degree and the set heating temperature.

[0011] Furthermore, the target heating time is calculated as follows: , In the formula, T s Indicates the target heating time; S 3 indicates target dehumidification; W It is the dehumidification efficiency at the set heating temperature, indicating the degree of dehumidification at the heating temperature per unit time.

[0012] Furthermore, setting the heating temperature according to the current humidity level comprises the following steps: Establish a reference table of association between wetness and heating temperature based on historical data samples; According to the current wetness, the associated reference table is retrieved to obtain the initial heating temperature; The initial heating temperature is adjusted in combination with the environmental data to determine the final heating temperature.

[0013] Furthermore, in the intelligent control module, dehumidifying the electrical equipment of the substation cabinet according to the matched heating parameters specifically includes the following steps: Receiving customized heating parameters from a parameter matching module; Establish electrical connection with the heater in the transformer cabinet, perform initialization check on the connected equipment, and check the circuit connectivity and working status of each component; According to the received heating temperature parameter, the target heating temperature of the heater is set, and the target heating temperature is converted into an electrical signal and sent to the heater to start heating; During the heating process, the temperature sensor is used to monitor the actual heating temperature in real time. If deviation is found, it will be adjusted in time, and the actual temperature will be maintained at the set value by fine-tuning the electrical signal; When the heating temperature reaches the standard and is stable, the heating time timing is started, and the heating duration is recorded using the timing unit according to the received parameters; Keep an eye on the timing process and stop heating immediately when the set time is reached; During the dehumidification process, the humidity sensor is used to monitor the changes in air humidity in real time, and the humidity values ​​before and after are compared to determine whether the dehumidification effect has reached the expected goal; If the expectation is not met, the situation will be fed back for re-evaluation and new heating parameters will be matched to perform dehumidification again until the target dehumidification effect is achieved.

[0014] Furthermore, the intelligent control module also uses a high-definition camera to monitor the circuit conditions of the electrical equipment in real time during the heating and dehumidification process. When a fault occurs in the circuit condition, the heating is immediately stopped and the fault is reported.

[0015] Beneficial effects of the present invention: (1) The humidity level of the electrical equipment in the transformer cabinet is predicted by monitoring the temperature and humidity through high-precision sensors; according to the humidity level, the heating temperature and heating time of the heater are controlled to dehumidify the electrical equipment. The present invention solves the dilemma that the prior art cannot coordinate the relationship between the humidity level and the heating parameters, and the heat control accuracy of the heater is low, which leads to the failure to achieve the dehumidification effect or waste of energy.

[0016] (2) Mark and classify data samples according to whether the electrical equipment has a fault; determine the critical point of surface condensation water to determine whether the electrical equipment has a fault according to the classification; calculate the relative difference between all surface condensation water samples and the critical point; use a cluster analysis method to divide the data samples into a number of clusters according to the relative difference, and determine the degree of wetness according to the numerical range of the relative difference of each cluster. The present invention associates the degree of wetness of electrical equipment with the presence or absence of a fault, providing a reference for achieving a dehumidification effect.

[0017] (3) By establishing an artificial intelligence model between environmental data and humidity levels, the accuracy of humidity level judgment is improved, thereby improving the accuracy of heat control of the heater. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings.

[0019] Figure 1 This is a structural diagram of an environmentally friendly and energy-saving substation in the present invention.

[0020] Figure 2 The present invention is a flow chart of predicting the humidity level of electrical equipment in a transformer cabinet according to environmental data in one embodiment of the present invention.

[0021] Figure 3 This is a flow chart of matching heating parameters according to the degree of humidity in one embodiment of the present invention. DETAILED DESCRIPTION

[0022] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the specific implementation mode, structure, characteristics and effects of the present invention are described in detail below in combination with the accompanying drawings and preferred embodiments.

[0023] See also Figure 1-Figure 3 The present application provides an environmentally friendly and energy-saving substation, including an environmental monitoring module, an environmental analysis module, a parameter matching module and an intelligent control module, wherein the environmental monitoring module, the environmental analysis module, the parameter matching module and the intelligent control module are communicatively connected, wherein: The environmental monitoring module uses a high-precision sensor to monitor the environmental data inside the substation cabinet, and the environmental data includes the ambient temperature and ambient humidity; In this embodiment, the high-precision sensor can accurately measure the amount of water vapor in the air inside the substation, and can distinguish the subtle fluctuations of humidity under different working conditions and weather conditions. For example, in humid weather, it can promptly detect a significant increase in air humidity; and in dry weather or when the humidity is reduced due to good ventilation in the substation, it can also accurately reflect the corresponding changes in humidity. The two environmental data, ambient temperature and ambient humidity, are the focus of attention because they have a very critical impact on the normal operation and service life of electrical equipment in the substation. Higher ambient temperatures may cause electrical equipment to overheat, affect its performance and even cause failures; while excessively high ambient humidity may cause the insulation performance of electrical equipment to decrease, metal parts to rust and corrode, and other problems, seriously threatening the safe and stable operation of the substation. Through the environmental monitoring module, these high-precision sensors are continuously used to monitor the ambient temperature and humidity inside the substation in real time, and the accurate data obtained becomes the cornerstone of a series of subsequent analysis, decision-making and control operations of the entire substation system. These data will be accurately transmitted to other related modules such as the environmental analysis module, providing them with detailed and reliable original information, so that the entire substation can achieve intelligent operation management based on these accurate environmental data. For example, dehumidification operations can be performed by reasonably matching heating parameters, thereby achieving the goal of ensuring the normal operation of electrical equipment and achieving energy conservation and environmental protection.

[0024] The environmental analysis module predicts the humidity level of the electrical equipment in the substation cabinet according to the environmental data; the humidity level is divided according to whether it causes the electrical equipment to malfunction; In this embodiment, the environmental analysis module works based on the internal ambient temperature and humidity data of the substation cabinet transmitted by the environmental monitoring module. The ambient temperature affects the water vapor capacity of the air, and the humidity indicates the actual water vapor content, and the two interact with each other. The module will use professional methods to process these data, such as calculating the dew point temperature through temperature and humidity, and comparing the actual temperature with the dew point temperature, to preliminarily judge the condensation trend of water vapor. At the same time, it will also comprehensively consider factors such as the ventilation conditions of the substation cabinet, the heat dissipation characteristics of electrical equipment, and materials. Ventilation can take away water vapor, the heat dissipation characteristics of the equipment affect the surface temperature, and different materials have different water vapor adsorption capabilities. Based on the comprehensive analysis of these various factors, the environmental analysis module uses built-in algorithms and models to quantitatively predict the humidity level of the electrical equipment in the substation cabinet, and presents it in the form of specific values ​​or grades. This prediction result provides a key basis for the subsequent precise matching of heating parameters and the implementation of dehumidification operations, ensuring the normal operation of the equipment and energy saving and environmental protection.

[0025] The parameter matching module matches the heating parameters of the heater according to the humidity level, wherein the heating parameters include heating temperature and heating duration; In this embodiment, when the environmental analysis module accurately predicts the degree of humidity of the electrical equipment in the substation cabinet, the parameter matching module works based on this key information. Its core task is to select the most suitable group for the current humidity level from the many pre-set heating parameter combinations, which involves the accurate matching of the heating temperature and the heating time. For different degrees of humidity, the required dehumidification intensity is different. For example, when the humidity of the electrical equipment is relatively light, it may only take a relatively low heating temperature and a short heating time to effectively remove the water vapor and achieve the purpose of dehumidification; if the humidity is heavy, it is necessary to match a higher heating temperature and a longer heating time to ensure that the water vapor that may exist on the surface and inside of the electrical equipment is completely eliminated to ensure its normal operation. At the same time, in the matching process, the parameter matching module must also fully consider the requirements of energy conservation and environmental protection. It is not possible to increase the heating parameters uncontrollably simply for the pursuit of dehumidification effect, but to select the parameter combination with lower energy consumption as much as possible on the premise of ensuring effective dehumidification, so as to achieve the rational use of energy, so that the substation can not only maintain the electrical equipment in a good dry state, but also meet the current development trend of energy conservation and emission reduction.

[0026] The intelligent control module dehumidifies the electrical equipment of the transformer cabinet according to the matched heating parameters.

[0027] In this embodiment, after the parameter matching module determines the corresponding heating parameters (including specific heating temperature and heating time) according to the humidity of the electrical equipment in the substation cabinet, the intelligent control module starts the precise dehumidification action. It will first establish an effective electrical connection with the heater and other related dehumidification equipment in the substation cabinet to ensure that these devices can be accurately controlled. Then, according to the matching heating temperature parameters, the intelligent control module sends instructions to the heater through a precise control circuit to accurately set the target heating temperature of the heater so that it can be stabilized at the required temperature level for heating operation. During the heating process, the intelligent control module will always pay attention to the progress of the heating time. Once the matching heating time value is reached, it will immediately issue an instruction to stop the heating of the heater to prevent excessive heating from causing energy waste and possible adverse effects on electrical equipment. At the same time, the intelligent control module will also use various sensors installed in the substation cabinet, such as humidity sensors, to monitor the humidity changes inside the substation cabinet in real time. If the humidity has reached the ideal dehumidification effect range before the heating time ends, it also has the right to stop heating in advance, realizing flexible and accurate dehumidification control, and further improving energy efficiency. Through such a series of rigorous and intelligent operations, the intelligent control module performs efficient, accurate, energy-saving and environmentally friendly dehumidification treatment on the electrical equipment in the substation cabinet according to the matching heating parameters, ensuring that the electrical equipment operates stably in a suitable dry environment.

[0028] Furthermore, predicting the humidity level of the electrical equipment of the substation cabinet according to the environmental data comprises the following steps: Monitor the environmental data inside the substation cabinet and the condensation water data samples on the surface of electrical equipment during the historical period; In this embodiment, the amount of condensation on the surface of the electrical equipment can be measured by installing a sensor specifically used to measure the moisture content on the surface of the electrical equipment, such as a capacitive humidity sensor, a resistive humidity sensor, etc. The capacitive humidity sensor measures humidity by using the principle that changes in humidity cause changes in capacitance. When moisture condenses on the surface of the equipment, the capacitance value will change accordingly. Through calibration and conversion, relevant data on the amount of condensation on the surface can be obtained; the resistive humidity sensor is based on the effect of humidity on resistance. Moisture will cause the resistance to change, and then the humidity of the surface is determined based on this change relationship, and the amount of condensation is further calculated. This method can obtain information on the amount of condensation on the surface in real time and relatively accurately.

[0029] Mark and classify data samples according to whether electrical equipment has failed; In this embodiment, by sorting out the internal environment data of the substation cabinet and the condensation water data samples on the surface of the electrical equipment monitored in the historical period, each data sample is marked according to the key situation of whether the electrical equipment fails in the corresponding time period. For example, all data samples corresponding to the failure are marked as "failure class", and those without failure are marked as "normal class". This classification method can clearly distinguish the data under different operating results, help to dig out the characteristics related to the failure and the data rules under normal operating conditions, and prepare for further determining the key factors affecting the operation of the equipment.

[0030] Determine the critical point of surface condensation water in electrical equipment based on the classification to determine whether there is a fault; After completing the above-mentioned marking and classification, the two groups of data, "fault type" and "normal type", are analyzed in depth. The focus is on the data information of the amount of condensation on the surface of electrical equipment. Through statistical analysis, data fitting and other methods, we try to find out under different categories, when the amount of condensation on the surface of electrical equipment reaches what level, it will most likely cause a fault (for "fault type") or still maintain normal operation (for "normal type"). This key condensation value is the so-called critical point of surface condensation. For example, after analysis, it was found that for a certain type of electrical equipment, when the amount of surface condensation exceeds 5 milliliters and is more common in the "fault type" data, then 5 milliliters is determined to be the critical point of surface condensation of the equipment under fault conditions. Determining this critical point is of great significance for subsequent quantitative evaluation of the degree of wetness of the equipment and prediction of failure risks.

[0031] Calculate the relative difference between all surface condensation water samples and the critical point; Once the critical point of surface condensation water volume for electrical equipment failure is determined, the next step is to calculate all collected surface condensation water volume samples one by one. Specifically, the value of each surface condensation water volume sample is subtracted from the value of the corresponding critical point (whether the fault critical point or the normal critical point is used according to the category to which the sample belongs), and the result is the relative difference between the sample and the critical point. For example, if the value of a surface condensation water volume sample is 4 ml, and the corresponding fault critical point of its category is 5 ml, then the relative difference between the sample and the critical point is (4 - 5) / 5 = 0.2. By calculating the relative difference, different surface condensation water volume samples can be quantitatively compared based on the critical point as a reference, which can more clearly show the degree of deviation of each sample from the critical point, providing an important data basis for subsequent cluster analysis.

[0032] A cluster analysis method is used to divide the data sample into a number of clusters according to the relative difference, and the degree of wetness is determined according to the numerical range of the relative difference of each cluster; After obtaining the relative difference between all surface condensation samples and the critical point, cluster analysis is used to process these data. Cluster analysis is a data analysis technique that divides data objects into different groups (i.e., clusters) based on the similarity of data. Here, all data samples are divided into several clusters based on the similarity of relative differences. For example, samples with relative differences in the interval [-0.2, 0] are divided into one cluster, and samples in the interval [0, 0.2] are divided into another cluster, and so on. Then, the degree of wetness corresponding to each cluster is determined according to the numerical range of the relative difference in each cluster. For example, clusters with relative differences in the interval [-0.2, 0] may correspond to the degree of "mild wetness" because the condensation of these samples is relatively close to the critical point of normal operation or slightly lower than the critical point of failure; while clusters with relative differences in the interval [0, 0.2] may correspond to the degree of "moderate wetness", indicating that the condensation of these samples is relatively more deviated from the critical point of normal operation and closer to the critical point of failure. In this way, the originally complex surface condensation water data is converted into intuitively understandable wetness information through cluster analysis, providing a clearer and more operational basis for subsequent operations such as matching heating parameters according to the wetness degree.

[0033] The artificial intelligence model was trained using environmental data as the independent variable and humidity level as the dependent variable; The newly collected environmental data is input into the trained artificial intelligence model to output the predicted humidity level of the electrical equipment in the substation cabinet.

[0034] Furthermore, the relative difference is calculated as follows: , In the formula, R s Relative difference. S 2 represents a sample of condensation water on a certain surface; S 1 represents the critical point of surface condensation.

[0035] Furthermore, the cluster analysis method specifically comprises the following steps: Selecting a number of data samples as cluster centers, wherein the cluster centers include relative difference samples of critical points of surface condensation water; When performing cluster analysis, we must first select several samples from all existing data samples (these samples are processed by the relative difference between the critical point of surface condensation water and the previous calculation) as cluster centers. Here, we deliberately select samples containing the relative difference of the critical point of surface condensation water as cluster centers because these samples are representative of the entire data set, and their relative difference around the critical point can provide a key reference point for subsequent cluster division. For example, if the relative difference of the critical point of surface condensation water has different situations such as -1, 0, and 1, some samples corresponding to some typical values ​​may be selected as cluster centers, so as to better divide different clusters according to the distance relationship between the data and these centers.

[0036] Calculate the Euclidean distance between all data samples and the cluster center, and assign each data sample to the cluster center with the smallest Euclidean distance to form a cluster; For each data sample, the Euclidean distance between it and each selected cluster center is calculated. Euclidean distance is a common method to measure the spatial distance between two points. In this case, it is to measure the distance between the data sample and the cluster center in the "space" with the relative difference as the dimension. Then, each data sample is assigned to the group corresponding to the cluster center with the smallest Euclidean distance, so that different clusters are initially formed. For example, if the Euclidean distance of data sample A with cluster center C1 is calculated to be 3, and the Euclidean distance with cluster center C2 is 2, then data sample A will be assigned to the cluster where cluster center C2 is located.

[0037] Calculate the average value of each cluster data sample and use the average value as the new cluster center; After the clusters are initially formed, the data samples in each cluster need to be further processed. The average value of all data samples in each cluster is calculated on the dimension of relative difference. Then, this average value is used as the new cluster center for the next round of cluster analysis. The purpose of this is to make the cluster center more accurately reflect the overall characteristics of the data in the cluster. With the iterative process, the selection of cluster centers is continuously optimized to make it more consistent with the actual distribution of the data.

[0038] Repeat the process of calculating the Euclidean distance to assign clusters and calculating the average value until the new crystal center no longer changes, and determine the final cluster; Next, repeat the previous process of calculating the Euclidean distance between all data samples and the new cluster center, assigning data samples to the cluster center with the smallest Euclidean distance to form clusters, and calculating the average value of each cluster data sample and using it as the new cluster center. Continue to perform such iterative operations. In each round of iteration, the division of cluster centers and clusters will be adjusted and optimized according to the actual situation of the data samples. After several rounds of iterations, the new cluster center no longer changes, and the final cluster can be determined. This means that after many adjustments, the division of cluster centers and clusters has reached a relatively stable state, which can more accurately reflect the intrinsic relationship between data samples and their relative relationship with the critical point of surface condensation.

[0039] The wetness degree is determined according to the numerical range of the relative difference of the final clusters. The wetness degree is based on the critical point of the surface condensation water, and a value greater than the critical point is a positive value, and a value less than the critical point is a negative value.

[0040] Finally, according to the final clusters determined, the numerical range of the relative difference in each cluster is observed. Since the wetness is defined with reference to the critical point of surface condensation, the wetness corresponding to each cluster is determined according to this numerical range. For example, if the numerical range of the relative difference in a final cluster is [-0.2, -0.1], with reference to the critical point of surface condensation, the relative difference in this range is less than the critical point, so the wetness corresponding to this cluster can be defined as "slightly wet and tends to be dry"; if the numerical range of the relative difference in another final cluster is [0.1, 0.2], and its relative difference is greater than the critical point, then the wetness corresponding to this cluster can be defined as "moderately wet and tends to be humid". In this way, the results of cluster analysis are used to accurately convert the relative difference of data samples into intuitive wetness information, providing a clear basis for subsequent related operations (such as matching heating parameters) based on the wetness.

[0041] Furthermore, the artificial intelligence model is configured as a BP artificial neural network model, comprising the following steps: Build the network structure and determine the input layer, hidden layer and output layer settings to adapt to the environmental data and humidity prediction results; Building a network structure is the primary task of building a BP artificial neural network model. The setting of the input layer needs to be closely centered around the environmental data to be processed, such as the ambient temperature and humidity inside the substation cabinet, and the number of neurons must match the input environmental data dimension in order to accurately receive these data information. The hidden layer plays a key role in complex feature extraction and nonlinear transformation of the input data. The number of layers and the number of neurons in each layer often need to be determined through experiments, experience, or with the help of some optimization algorithms. The purpose is to enable the network to better fit the complex relationship between environmental data and humidity. The output layer is relatively simple, usually set to one neuron, which is used to output the final predicted result of the humidity level of the substation electrical equipment, ensuring that the model can give a predicted value that meets actual needs.

[0042] Preprocessing the historical environmental data used for training and the corresponding electrical equipment humidity degree data, wherein the preprocessing includes normalization processing, outlier processing and missing value processing; Before using historical data for model training, preprocessing operations must be performed to improve data quality and model training effects. Normalization is one of the common methods. It uses specific mathematical transformations to uniformly map environmental data with different value ranges (such as temperature may be in a certain interval, humidity in another interval) and the corresponding humidity data to a specific, easier-to-process interval, such as the [0, 1] interval. This can avoid the problem that some data features are over-emphasized or ignored during network training due to large differences in data magnitude. Outlier processing is for those outliers in the data set that are significantly deviated from other data points. These outliers may be caused by measurement errors, equipment failures, etc. If not processed, they may mislead model training, so appropriate methods (such as deletion, correction, etc.) are needed to minimize their impact on model training. Missing value processing is also important. When there is some missing data in the data set, the missing values ​​can be filled by methods such as mean filling and adjacent value filling to ensure that the data input into the network is complete and valid, thereby providing an accurate and reliable data basis for model training.

[0043] Initialize network parameters, including random settings of connection weights and thresholds; Initializing network parameters is the process of giving an initial state to the BP artificial neural network model. The connection weight determines the strength of information transmission between neurons, while the threshold affects whether the neuron is activated. By randomly setting the connection weight and threshold, an initial starting point is provided for the network. Although this starting point may not be accurate, it lays the foundation for the subsequent training process. In the subsequent training process, these parameters will be gradually optimized and adjusted according to the input data and the errors generated, so that the network can better fit the data and improve the accuracy of the prediction.

[0044] The preprocessed data is input into the network through forward propagation calculation to obtain a preliminary predicted moisture level value; When the network structure is built, the data is preprocessed, and the network parameters are initialized, the forward propagation calculation can begin. The preprocessed historical environmental data samples are input into the input layer of the BP artificial neural network in sequence, and then the data is processed by nonlinear transformation through the hidden layer in sequence according to the connection structure of the network. In the hidden layer, the neurons calculate and output the results based on the input data and the current connection weights and thresholds, and these results are then passed to the next layer until they finally reach the output layer. The output layer calculates a preliminary predicted humidity value based on the received information. This process simulates the forward flow of information from input to output, and realizes a feature mapping of the input environmental data through the calculation of each layer, converting it into a preliminary prediction of the humidity level.

[0045] Back propagation is performed based on the error between the predicted value and the actual value to adjust the network parameters; After completing the forward propagation calculation to obtain the preliminary predicted humidity value, it is necessary to compare this predicted value with the actual humidity value of the electrical equipment (that is, the real humidity value corresponding to the input historical environmental data) to calculate the error between them. Usually, a suitable error measurement indicator such as a confusion matrix is ​​used to calculate this error. Then, based on the calculated error, the back propagation process is started. In the back propagation, the error starts from the output layer and passes through each hidden layer in the opposite direction of the forward propagation. In this process, the connection weights and thresholds of the neurons in each layer are adjusted according to the error in order to reduce the error during the next forward propagation. By continuously performing error calculation and back propagation, the network can gradually learn the correct relationship between the data, making the predicted value closer and closer to the actual value.

[0046] Repeat the process of forward propagation and back propagation until the error is lower than the set threshold, completing the post-training evaluation and verification of the model's accuracy.

[0047] In order to achieve a good training effect for the BP artificial neural network model, it is necessary to repeat the two processes of forward propagation and back propagation. After each forward propagation calculation to obtain the predicted value, the error is calculated by comparing it with the actual value and back propagation is performed to adjust the network parameters. This cycle is repeated until the calculated error is lower than the preset threshold. At this time, it is considered that the model has reached a relatively satisfactory training state. However, completing the training does not mean that the model is necessarily accurate and reliable. The trained model needs to be evaluated and verified. By inputting a part of the reserved data that does not participate in the training (called the validation set) into the trained model, the prediction effect of the model on these data is observed. The prediction accuracy of the model can be evaluated by indicators such as accuracy, precision, and F1 value. By comparing with the error indicators used in the training process, it is determined whether the model is overfitting (that is, the model performs well on the training data, but performs poorly on the data that does not participate in the training). If overfitting is found, it can be adjusted by adjusting the model structure, increasing the amount of data, and using regularization methods to ensure that the model has good versatility and prediction accuracy.

[0048] Furthermore, matching the heating parameters of the heater according to the humidity level comprises the following steps: Set target wetness level; This is the initial setting link of the entire matching heating parameter process. It is necessary to predetermine a desired target humidity level based on multiple factors such as the specific operating requirements, environmental conditions, and characteristics of the equipment itself of the electrical equipment in the substation. This target humidity level is an idealized humidity state value that can ensure the normal operation of the electrical equipment and meet the requirements of energy conservation and environmental protection. For example, based on past experience and the performance analysis of this type of electrical equipment in different humidity environments, it is determined that the humidity of the surface of the electrical equipment and the surrounding environment is maintained within a certain relative humidity value range, such as 30% - 40%, as the target humidity level. It provides a clear direction and final state indicator to be achieved for subsequent dehumidification operations.

[0049] Subtract the target humidity level from the current humidity level to obtain the target dehumidification level; After the target humidity level is determined, the dehumidification level to be achieved must be further determined in combination with the humidity level of the electrical equipment in the substation cabinet obtained through actual monitoring and analysis. Through simple mathematical operations, the difference between the current humidity level and the target humidity level is the target dehumidification level to be achieved. This target dehumidification level intuitively reflects how much humidity needs to be reduced from the current humidity level in order for the electrical equipment to reach the predetermined ideal humidity state. For example, if the relative humidity corresponding to the current humidity level is 60%, and the relative humidity of the set target humidity level is 40%, then the target dehumidification level is 60% - 40% = 20%, which means that the humidity needs to be reduced by 20% through subsequent heating and dehumidification operations.

[0050] Set the heating temperature according to the current humidity level; This step is the key part of flexibly adjusting the heating parameters according to the actual situation. Different current humidity levels require different heating temperatures to achieve effective dehumidification effects. Generally speaking, when the current humidity level is high, that is, the humidity is high, it may be necessary to set a relatively high heating temperature in order to speed up the evaporation of water vapor and cause the humidity to drop rapidly; and when the current humidity level is low, a lower heating temperature can be set accordingly, which can achieve the purpose of dehumidification and avoid energy waste caused by excessive heating. This setting process will comprehensively consider many factors such as the current humidity level, the ventilation conditions in the substation cabinet, the material and heat dissipation characteristics of the electrical equipment, etc., and accurately determine the heating temperature value suitable for the current humidity level through specific algorithms or reference to the pre-established humidity level and heating temperature correspondence table.

[0051] Determine the target heating time by the target dehumidification and the set heating temperature; After determining the target dehumidification degree and heating temperature, the last step is to calculate the target heating time required to achieve the expected dehumidification effect. The determination of the target heating time is closely related to the target dehumidification degree and the set heating temperature. A higher target dehumidification degree may require a longer heating time to ensure sufficient water vapor evaporation to achieve effective reduction of humidity; and the set heating temperature will also affect the setting of the heating time. A higher heating temperature can usually speed up the evaporation of water vapor and may shorten the heating time to a certain extent, but this also requires comprehensive consideration of factors such as energy consumption and the impact on electrical equipment. By establishing relevant mathematical models or based on past actual operating experience, combined with the current target dehumidification degree and the set heating temperature, after a series of calculations or analyses, a target heating time value that can achieve the target dehumidification degree and take into account energy conservation and environmental protection requirements is finally determined, so that the subsequent intelligent control module can perform precise dehumidification operations on the electrical equipment of the substation according to this time.

[0052] Furthermore, the target heating time is calculated as follows: , In the formula, T s Indicates the target heating time; S 3 indicates target dehumidification; W It is the dehumidification efficiency at the set heating temperature, indicating the degree of dehumidification at the heating temperature per unit time.

[0053] Furthermore, setting the heating temperature according to the current humidity level comprises the following steps: Establish a reference table of association between wetness and heating temperature based on historical data samples; This step is the basis of the entire process of setting the heating temperature. First, a large number of historical data samples need to be collected. These samples cover the environmental data inside the substation cabinet at different times and the corresponding electrical equipment humidity data. At the same time, they also record the heating temperature settings used to achieve effective dehumidification under the circumstances at that time. Through in-depth analysis and statistical processing of these rich historical data, the inherent laws and correlations between humidity and heating temperature are found. For example, when the humidity is in a certain range (such as relative humidity between 50%-60%), after many practical verifications, it is found that a certain heating temperature range (such as 40℃-50℃) can achieve a better dehumidification effect and take into account energy saving and environmental protection requirements. Based on such analysis results, different levels of humidity are matched with the corresponding appropriate heating temperature range or specific values, so as to establish a detailed reference table of the correlation between humidity and heating temperature. This reference table will serve as an important basis for the subsequent setting of heating temperature to ensure that the set temperature can effectively dehumidify at different humidity levels.

[0054] According to the current wetness, the associated reference table is retrieved to obtain the initial heating temperature; After obtaining the humidity level information of the current substation electrical equipment, the next step is to search the established reference table of humidity level and heating temperature according to the humidity level. By accurately comparing the current humidity level with the humidity level levels divided in the reference table, find the corresponding items that match it. Extract the corresponding heating temperature range or specific value from the matching items. The extracted temperature value is the initial heating temperature that is preliminarily set. It is summarized based on a large amount of historical data and practical experience. Under normal circumstances, it can play a certain dehumidification role for electrical equipment at the current humidity level, but it may need to be further optimized and adjusted according to actual conditions.

[0055] The initial heating temperature is adjusted in combination with the environmental data to determine the final heating temperature.

[0056] Although the initial heating temperature is obtained by searching the associated reference table, in reality, relying solely on this temperature may not achieve the best dehumidification effect or meet the requirements of energy saving and environmental protection. Therefore, it is necessary to further adjust the initial heating temperature in combination with the current environmental data.

[0057] The current environmental data includes but is not limited to factors such as ambient temperature, ambient humidity, and ventilation conditions in the substation cabinet. For example, if the current ambient temperature is low, it may be necessary to appropriately increase the initial heating temperature to speed up the evaporation of water vapor and ensure the dehumidification effect; if the ambient humidity is already high, it may be necessary to slightly increase the temperature to cope with greater dehumidification pressure; and good ventilation conditions may allow the initial heating temperature to be appropriately lowered, because ventilation can speed up the dissipation of water vapor and may also achieve better dehumidification effects at relatively low temperatures. By comprehensively considering the impact of these environmental data on the dehumidification effect, using the corresponding algorithm or based on past experience, the initial heating temperature is reasonably adjusted, and finally a final heating temperature value that can effectively dehumidify and meet energy-saving and environmental protection requirements is determined, so that it can be used for subsequent precise dehumidification operations on the electrical equipment in the substation cabinet.

[0058] Furthermore, in the intelligent control module, dehumidifying the electrical equipment of the substation cabinet according to the matched heating parameters specifically includes the following steps: Receiving customized heating parameters from a parameter matching module; Establish electrical connection with the heater in the transformer cabinet, perform initialization check on the connected equipment, and check the circuit connectivity and working status of each component; According to the received heating temperature parameter, the target heating temperature of the heater is set, and the target heating temperature is converted into an electrical signal and sent to the heater to start heating; During the heating process, the temperature sensor is used to monitor the actual heating temperature in real time. If deviation is found, it will be adjusted in time, and the actual temperature will be maintained at the set value by fine-tuning the electrical signal; When the heating temperature reaches the standard and is stable, the heating time timing is started, and the heating duration is recorded using the timing unit according to the received parameters; Keep an eye on the timing process and stop heating immediately when the set time is reached; During the dehumidification process, the humidity sensor is used to monitor the changes in air humidity in real time, and the humidity values ​​before and after are compared to determine whether the dehumidification effect has reached the expected goal; If the expectation is not met, the situation will be fed back for re-evaluation and new heating parameters will be matched to perform dehumidification again until the target dehumidification effect is achieved.

[0059] Furthermore, the intelligent control module also uses a high-definition camera to monitor the circuit conditions of the electrical equipment in real time during the heating and dehumidification process. When a fault occurs in the circuit condition, the heating is immediately stopped and the fault is reported.

[0060] In this embodiment, the intelligent control module has an important additional function when heating and dehumidifying the electrical equipment of the substation cabinet. It uses a high-definition camera to monitor the line condition of the electrical equipment in real time. The high-definition camera can accurately capture the details of the line with its high resolution, such as whether the connection is tight, whether there is damage or deformation, and whether there are signs of abnormal heating. Once a line condition fault is detected, the intelligent control module will take immediate action. On the one hand, it will immediately stop the ongoing heating operation to prevent the fault from being aggravated by continued heating and avoid causing more serious safety accidents; on the other hand, it will quickly report the fault to the monitoring center, operation and maintenance personnel and other relevant responsible entities of the substation, so that they can know the approximate location of the fault, possible types and other key information in time, and arrange professional personnel to rush to the site for inspection and treatment as soon as possible, so as to improve the protection mechanism and ensure the safe and stable operation of the substation.

[0061] The above description is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technical personnel in this field can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any brief modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the content of the technical solution of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. An environmentally friendly and energy-saving substation, characterized by: It includes an environment monitoring module, an environment analysis module, a parameter matching module and an intelligent control module, wherein the environment monitoring module, the environment analysis module, the parameter matching module and the intelligent control module are communicatively connected, wherein: The environmental monitoring module uses a high-precision sensor to monitor the environmental data inside the substation cabinet, and the environmental data includes the ambient temperature and ambient humidity; The environmental analysis module predicts the humidity level of the electrical equipment in the substation cabinet according to the environmental data, and the humidity level is divided according to whether it causes the electrical equipment to fail, including the following steps: Monitor the environmental data inside the substation cabinet and the condensation water data samples on the surface of electrical equipment during the historical period; Mark and classify data samples according to whether the electrical equipment has a fault or not; Determine the critical point of surface condensation water in electrical equipment based on the classification to determine whether there is a fault; Calculate the relative difference between all surface condensation water samples and the critical point; Using a cluster analysis method to divide the data sample into a number of clusters according to the relative difference, and determining the wetness according to the numerical range of the relative difference of each cluster; The artificial intelligence model was trained using environmental data as the independent variable and humidity level as the dependent variable; The newly collected environmental data is input into the trained artificial intelligence model, and the predicted humidity level of the electrical equipment in the substation cabinet is output; The parameter matching module matches the heating parameters of the heater according to the humidity level, wherein the heating parameters include heating temperature and heating duration; The intelligent control module dehumidifies the electrical equipment of the transformer cabinet according to the matched heating parameters.

2. The environmentally friendly and energy-saving substation according to claim 1, characterized in that: The relative difference is calculated as follows: ， In the formula, R s Relative difference. S 2 represents a sample of condensation water on a certain surface; S 1 represents the critical point of surface condensation.

3. The environmentally friendly and energy-saving substation according to claim 1 is characterized in that: The cluster analysis method specifically comprises the following steps: Selecting a number of data samples as cluster centers, wherein the cluster centers include relative difference samples of critical points of surface condensation water; Calculate the Euclidean distance between all data samples and the cluster center, and assign each data sample to the cluster center with the smallest Euclidean distance to form a cluster; Calculate the average value of each cluster data sample and use the average value as the new cluster center; Repeat the process of calculating the Euclidean distance to assign clusters and calculating the average value until the new cluster center no longer changes, and determine the final cluster; The wetness degree is determined according to the numerical range of the relative difference of the final clusters. The wetness degree is based on the critical point of the surface condensation water, and a value greater than the critical point is a positive value, and a value less than the critical point is a negative value.

4. The environmentally friendly and energy-saving substation according to claim 1 is characterized in that: The artificial intelligence model is configured as a BP artificial neural network model, comprising the following steps: Build the network structure and determine the input layer, hidden layer and output layer settings to adapt to the environmental data and humidity prediction results; Preprocessing the historical environmental data used for training and the corresponding electrical equipment humidity degree data, wherein the preprocessing includes normalization processing, outlier processing and missing value processing; Initialize network parameters, including random settings of connection weights and thresholds; The preprocessed data is input into the network through forward propagation calculation to obtain a preliminary predicted moisture level value; Back propagation is performed based on the error between the predicted value and the actual value to adjust the network parameters; Repeat the process of forward propagation and back propagation until the error is lower than the set threshold, completing the post-training evaluation and verification of the model's accuracy.

5. The environmentally friendly and energy-saving substation according to claim 1, characterized in that: The step of matching the heating parameters of the heater according to the humidity level comprises the following steps: Set target wetness level; Subtract the target humidity level from the current humidity level to obtain the target dehumidification level; Set the heating temperature according to the current humidity level; The target heating time is determined by the target dehumidification degree and the set heating temperature.

6. The environmentally friendly and energy-saving substation according to claim 5, characterized in that: The target heating time is calculated as follows: ， In the formula, T s Indicates the target heating time; S 3 indicates target dehumidification; W It is the dehumidification efficiency at the set heating temperature, indicating the degree of dehumidification at the heating temperature per unit time.

7. The environmentally friendly and energy-saving substation according to claim 5, characterized in that: The step of setting the heating temperature according to the current humidity level comprises the following steps: Establish a reference table of association between wetness and heating temperature based on historical data samples; According to the current wetness, the associated reference table is retrieved to obtain the initial heating temperature; The initial heating temperature is adjusted in combination with the environmental data to determine the final heating temperature.

8. The environmentally friendly and energy-saving substation according to claim 1, characterized in that: In the intelligent control module, dehumidifying the electrical equipment of the substation cabinet according to the matched heating parameters specifically includes the following steps: Receiving customized heating parameters from a parameter matching module; Establish electrical connection with the heater in the transformer cabinet, perform initialization check on the connected equipment, and check the circuit connectivity and working status of each component; According to the received heating temperature parameter, the target heating temperature of the heater is set, and the target heating temperature is converted into an electrical signal and sent to the heater to start heating; During the heating process, the temperature sensor is used to monitor the actual heating temperature in real time. If deviation is found, it will be adjusted in time, and the actual temperature will be maintained at the set value by fine-tuning the electrical signal; When the heating temperature reaches the standard and is stable, the heating time timing is started, and the heating duration is recorded using the timing unit according to the received parameters; Keep an eye on the timing process and stop heating immediately when the set time is reached; During the dehumidification process, the humidity sensor is used to monitor the changes in air humidity in real time, and the humidity values ​​before and after are compared to determine whether the dehumidification effect has reached the expected goal; If the expectation is not met, the situation will be fed back for re-evaluation and new heating parameters will be matched to perform dehumidification again until the target dehumidification effect is achieved.

9. The environmentally friendly and energy-saving substation according to claim 1, characterized in that: The intelligent control module also uses a high-definition camera to monitor the circuit conditions of the electrical equipment in real time during the heating and dehumidification process. When a fault occurs in the circuit condition, the heating is immediately stopped and the fault is reported.

Images