Lightning protection compensation capacitor intelligent monitoring system

By designing an intelligent monitoring system for lightning-proof compensation capacitors in a plateau environment, and using sensors and long-term memory networks to predict capacitor status, the problems of capacitor monitoring and fault location are solved, and the stability and equipment safety of the power system are improved.

CN118937839APending Publication Date: 2024-11-12HEBEI UNIV OF SCI & TECH +3
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411013497.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-26
Publication Date
2024-11-12

AI Technical Summary

Technical Problem

In a plateau environment, the performance and life monitoring of track compensation capacitors is difficult to achieve, and the fault positioning is inaccurate, which affects the stability of the power system.

Method used

An intelligent monitoring system for lightning protection compensation capacitors is designed, including parameter acquisition, processing and monitoring subsystem, using sensors to obtain operating parameters, and construct a capacitor state prediction model through long and short-term memory networks, normalized processing and training to generate capacitors Predict the results of the operation status, discover faults in a timely manner and locate them.

Benefits of technology

It improves the safety and stability of the power system, extends the service life of the capacitor, reduces the damage to the capacitor by lightning strikes, and ensures the safe operation of power equipment and systems.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118937839A_ABST
    Figure CN118937839A_ABST
Patent Text Reader

Abstract

The invention discloses an intelligent monitoring system for a lightning protection compensation capacitor, and belongs to the technical field of intelligent monitoring of capacitors, and the system comprises a parameter acquisition subsystem which is used for generating a capacitor operation parameter data set; the parameter processing subsystem is used for carrying out normalization processing on the capacitor operation parameter data set to obtain a normalized data set; the central processing subsystem is used for constructing a long-short-term memory network, training the long-short-term memory network and generating a capacitor state prediction model; and the monitoring subsystem is used for inputting the normalized data set into a prediction model for calculation, obtaining a capacitor operation state prediction result, and giving an alarm based on the capacitor operation state prediction result. For a lightning protection compensation capacitor monitoring system, the method can adapt to the operation state change of the capacitor in different time periods, and improves the adaptability of the system.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention belongs to the technical field of capacitor intelligent monitoring, and in particular relates to an intelligent monitoring system for lightning protection compensation capacitors. Background Art

[0002] In plateau environments, track compensation capacitors play an important role in the power system. The power grid in plateau areas usually faces greater voltage fluctuations and grid instability, which may affect the stability of the power system. Track compensation capacitors can help adjust the power factor of the power system and improve the stability of the power grid. Plateau areas usually have long transmission distances and large power transmission losses. By using track compensation capacitors, the power factor of the power grid can be optimized, the reactive power loss in power transmission can be reduced, and the efficiency of power transmission can be improved.

[0003] Especially during the tourist season or when the climate changes dramatically, the seasonal and geographical conditions in the plateau area may cause large changes in the power load. Track compensation capacitors can help adjust the power factor of the power system and better adapt to changes in the power load. However, at the same time, changeable weather conditions can cause compensation capacitors to be prone to short circuits and have a short service life. How to monitor the performance and life of compensation capacitors in plateau environments, and how to accurately locate capacitor failures have become urgent issues that need to be addressed. Summary of the invention

[0004] In order to solve the above technical problems, the present invention proposes an intelligent monitoring system for lightning protection compensation capacitors to solve the above problems in the prior art.

[0005] To achieve the above object, the present invention provides an intelligent monitoring system for lightning protection compensation capacitors, comprising:

[0006] A parameter acquisition subsystem, used for acquiring operating parameters of the track compensation capacitor based on sensors, converting the operating parameters into digital signals, and generating a capacitor operating parameter data set;

[0007] A parameter processing subsystem, connected to the parameter acquisition subsystem, for normalizing the capacitor operation parameter data set to obtain a normalized data set;

[0008] A central processing subsystem is connected to the parameter processing subsystem and is used to construct a long short-term memory network, and train the long short-term memory network in a progressive order of sample set data acquisition time and using the acquisition time as a constraint condition until the long short-term memory network learns the characteristics of the operating parameters and time and then outputs them to generate a capacitor state prediction model;

[0009] The monitoring subsystem is connected to the central processing subsystem and is used to input the normalized data set into the capacitor state prediction model for calculation, obtain the capacitor operation state prediction result, and issue an alarm based on the capacitor operation state prediction result.

[0010] Preferably, the parameter acquisition subsystem includes:

[0011] A voltage signal measurement module is used to collect the real-time signal of the capacitor through a sensor to obtain a parameter signal;

[0012] A feature extraction module, configured to analyze key features of the parameter signal based on the behavior and performance of the capacitor under different loads to obtain the operating parameters;

[0013] A conversion module, configured to convert the operating parameters into digital signals based on an analog-to-digital converter to generate the capacitor operating parameter data set;

[0014] The operating parameters include temperature characteristics, effective value of harmonics, fundamental wave, unbalance and capacitance.

[0015] Preferably, the parameter processing subsystem includes:

[0016] A data dimension reduction module, used for reducing the dimension of the capacitor operation parameter data set by a principal component analysis method to generate a dimension-reduced data set;

[0017] A feature engineering module, used to increase the features of the dimension reduction data set based on a feature engineering method to generate a feature data set;

[0018] The normalization module is used to normalize the feature data set to obtain the normalized data set.

[0019] Preferably, the central processing subsystem includes:

[0020] A sample set acquisition module, used for acquiring a sample set of a compensation capacitor based on a historical database;

[0021] A sample set processing module, used for dividing the sample set into a training set and a test set;

[0022] A model building module, used to build a long short-term memory network, improve the input gate and forget gate of the long short-term memory network, and generate an improved long short-term memory network;

[0023] A training module, used to train the improved long short-term memory network by using the progressive order of the data acquisition time in the training set and taking the acquisition time as a constraint condition to generate a training model;

[0024] A testing module, used for testing the training model based on the test set and outputting the capacitor state prediction model;

[0025] Preferably, the model building module includes:

[0026] A network construction unit is used to construct a long short-term memory network, and optimize the hyperparameters of the long short-term memory network through a genetic algorithm and the characteristics of the capacitor to obtain a long short-term memory network with optimal parameters;

[0027] Gate calculation module, used to calculate the deep attention gate;

[0028] A replacement module is used to replace the input gate and the forget gate in the long short-term memory network of the optimal parameters with the deep attention gate to generate the improved long short-term memory network.

[0029] Preferably, the training module includes:

[0030] A progressive training unit, used to train the improved long short-term memory network using the progressive order of data acquisition time in the training set, and simultaneously select a cross entropy loss function, and minimize the cross entropy loss function through a stochastic gradient descent algorithm to generate a progressive training model;

[0031] The constrained training unit is used to divide the training set into an input sequence and a target sequence by using time as a constraint condition to train the progressive training model and obtain the training model.

[0032] Preferably, the test module is also used to determine whether the training model has learned the characteristics of the operating parameters and time based on the test results. If so, the capacitor state prediction model is output; if not, the training model is re-trained iteratively.

[0033] Preferably, the monitoring subsystem includes:

[0034] A state calculation module, used for inputting the normalized data set into the capacitor state prediction model for calculation to obtain the future operating state and remaining service life of the capacitor;

[0035] A judgment module, used for calculating the failure probability of the future operation state and remaining service life of the capacitor under high altitude weather conditions;

[0036] The display module is used to transmit the fault probability to the terminal for display, and to display the location of the faulty capacitor by a combined positioning method.

[0037] To achieve the above object, the present invention further provides a lightning protection compensation capacitor, comprising:

[0038] Core connection device, metal casing, capacitor core, short circuit protection device, lead-out cable, grounding device and IoT sensor;

[0039] The core guide device is connected to the metal shell and is used to ensure the structural integrity and electrical performance of the capacitor element during the winding process;

[0040] The capacitor core is arranged inside the metal housing and connected to the anti-short circuit device to improve the transmission performance and stability of the track circuit;

[0041] The anti-short circuit device is arranged inside the metal housing and connected to the lead-out cable for overload protection;

[0042] The grounding device is connected to the metal housing and is used to conduct the stray current accumulated in the track circuit to the ground, while reducing external electromagnetic interference and ensuring signal stability;

[0043] The Internet of Things sensor is connected to the metal shell, and is used to obtain the operating parameters of the track compensation capacitor, convert the operating parameters into digital signals, and generate a capacitor operating parameter data set.

[0044] Compared with the prior art, the present invention has the following advantages and technical effects:

[0045] The LSTM neural network in the present invention can learn and memorize long-term dependencies in time series data, and is therefore suitable for processing complex time-related data. For the lightning protection compensation capacitor monitoring system, it can adapt to changes in the operating state of the capacitor in different time periods and improve the adaptability of the system. By continuously monitoring and predicting the operating state of the capacitor, the lightning protection compensation capacitor intelligent monitoring system can improve the safety and stability of the power system. Timely detection of abnormal operation or potential faults of the capacitor helps to reduce damage to the capacitor caused by lightning strikes, thereby protecting the safe operation of power equipment and systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0047] Figure 1 A schematic diagram of an intelligent monitoring system for lightning protection compensation capacitors according to an embodiment of the present invention;

[0048] Figure 2 Schematic diagram of a lightning protection compensation capacitor according to an embodiment of the present invention. DETAILED DESCRIPTION

[0049] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0050] It should be noted that the steps shown in the flowcharts of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and that, although a logical order is shown in the flowcharts, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0051] Embodiment 1

[0052] like Figure 1 As shown, in this embodiment, an intelligent monitoring system for lightning protection compensation capacitors is provided, including:

[0053] In the power system of the Qinghai-Tibet Plateau, track compensation capacitors are key reactive compensation equipment, and their stable operation is extremely important for the safe and reliable operation of the power grid. Therefore, this embodiment provides an intelligent monitoring system for lightning protection compensation capacitors, including: a parameter acquisition subsystem, a parameter processing subsystem, a central processing subsystem and a monitoring subsystem.

[0054] The parameter acquisition subsystem uses sensor technology to collect real-time operating parameters of track compensation capacitors, such as voltage, current, temperature, etc. After collecting these raw data, the parameter acquisition subsystem will use analog-to-digital conversion technology to convert these analog signals into digital signals to form a capacitor operating parameter data set.

[0055] The parameter processing subsystem is responsible for preprocessing these raw data. Since the collected data may have problems such as noise and outliers, directly using them for model training may affect the accuracy and stability of the model. Therefore, the parameter processing subsystem normalizes the data set to eliminate the dimensional differences between the data so that different parameters have the same weight. After normalization, the quality of the data set has been significantly improved, laying a solid foundation for subsequent model training.

[0056] The central processing subsystem is the core of the entire system. It uses a long short-term memory network (LSTM) to train the normalized data set. LSTM is a neural network that is particularly suitable for processing time series data. It can effectively capture long-term dependencies in the data. During the training process, the central processing subsystem constructs an LSTM network in a progressive order of time based on the acquisition time of the sample set data, and trains the network using the acquisition time as a constraint. As the training progresses, the LSTM network gradually learns the characteristics of operating parameters and time, and finally generates a capacitor state prediction model. This model can predict the future state of the capacitor based on the current operating parameters, providing important decision support for operation and maintenance personnel.

[0057] The monitoring subsystem is responsible for applying the prediction model to actual operation. It inputs the normalized data set into the capacitor state prediction model for calculation and obtains the prediction results of the capacitor operation state. These prediction results can include the remaining life of the capacitor, the possible fault type, etc. Based on these prediction results, the monitoring subsystem can promptly detect potential safety hazards and issue alarm information to remind the operation and maintenance personnel to take appropriate measures to deal with them.

[0058] Further optimizing the scheme, the parameter acquisition subsystem includes:

[0059] A voltage signal measurement module is used to collect the real-time signal of the capacitor through a sensor to obtain a parameter signal;

[0060] A feature extraction module, configured to analyze key features of the parameter signal based on the behavior and performance of the capacitor under different loads to obtain the operating parameters;

[0061] A conversion module, configured to convert the operating parameters into digital signals based on an analog-to-digital converter to generate the capacitor operating parameter data set;

[0062] The operating parameters include temperature characteristics, effective value of harmonics, fundamental wave, unbalance and capacitance.

[0063] Further optimizing the scheme, the parameter processing subsystem includes:

[0064] A data dimension reduction module, used for reducing the dimension of the capacitor operation parameter data set by a principal component analysis method to generate a dimension-reduced data set;

[0065] A feature engineering module, used to increase the features of the dimension reduction data set based on a feature engineering method to generate a feature data set;

[0066] The normalization module is used to normalize the feature data set to obtain the normalized data set.

[0067] In order to further improve the performance of the central processing subsystem, first of all, the sample set acquisition module is the basis of the entire optimization solution. This module is based on the historical database and comprehensively collects relevant sample data of the compensation capacitor. These data not only cover various operating states of the capacitor, but also include key information such as environmental factors and operating conditions.

[0068] Next, the sample set processing module is responsible for reasonably dividing the acquired sample set. This embodiment adopts a random sampling method to divide the sample set into a training set and a test set. The training set is used to build and train the model, while the test set is used to evaluate the performance of the model.

[0069] In the model building module, this embodiment uses a long short-term memory network (LSTM) as the core model. LSTM is a neural network structure suitable for processing sequence data, which can capture long-term dependencies in the data. In order to further improve the performance of the model, this embodiment improves the input gate and forget gate of LSTM. Specifically, this embodiment introduces an attention mechanism so that the model can automatically focus on important information and ignore insignificant details when processing sequence data. Through such improvements, this embodiment can generate a more efficient and accurate capacitor state prediction model.

[0070] The training module is a key link in the entire optimization scheme. In this module, the present embodiment uses the data in the training set to train the model in a progressive order of time. At the same time, the present embodiment uses the acquisition time as a constraint to ensure that the model can fully consider the characteristics of the time series during the training process. Through continuous iteration and optimization, the present embodiment finally generates a well-trained and stable capacitor state prediction model.

[0071] Finally, the test module is used to evaluate the trained model. This embodiment tests the model based on the test set, and evaluates the accuracy and reliability of the model by comparing the predicted results with the actual state. In addition, this embodiment also uses a variety of evaluation indicators, such as accuracy, recall rate, F1 score, etc., to comprehensively evaluate the model performance. Through such testing and analysis, this embodiment can ensure that the capacitor state prediction model finally generated has excellent performance.

[0072] Further optimizing the scheme, the model building module includes:

[0073] A network construction unit is used to construct a long short-term memory network, and optimize the hyperparameters of the long short-term memory network through a genetic algorithm and the characteristics of the capacitor to obtain a long short-term memory network with optimal parameters;

[0074] Gate calculation module, used to calculate the deep attention gate;

[0075] A replacement module is used to replace the input gate and the forget gate in the long short-term memory network of the optimal parameters with the deep attention gate to generate the improved long short-term memory network.

[0076] The improved long short-term memory network is trained through the training set, and each training interval is ten steps, and the last

[0077] The output at one moment is used as the input of the fully connected layer. In order to avoid overfitting, a dropout layer needs to be added at the training time to randomly forget some information. Finally, in order to predict the label, we need to pass the dropout output through a softmax layer to calculate the probability that the sample belongs to each category, and determine the category with the highest probability as the category to which the sample belongs, so as to realize the input of the time series related feature test set into the trained improved long short-term memory network

[0078] The experiments all used a unidirectional two-layer LSTM stacking structure, and used a fully connected layer and a softmax layer as the training model. During the training process, tanh was used as the activation function, and the small batch gradient descent method was used. The batch size was 64. A total of 1500 epochs were trained for 800 training samples, and 10 batches were trained each time. The effect of the model was verified by printing the test results of the test set once. In order to ensure the effectiveness of the comparison, all parameters of the following experiments are exactly the same (except for the first fully connected layer, because the input dimensions are different, but the output dimensions are the same).

[0079] The specific process of network training is as follows:

[0080] Determine the number of input, hidden, and output layers and nodes required based on actual needs.

[0081] Weights and biases are the core parameters of a neural network and need to be randomly initialized to a small value. Weights and biases control the strength and direction of signals transmitted between neurons.

[0082] Starting from the input layer, the prediction is calculated by taking the weighted sum of each neuron and passing it to the next layer.

[0083] The loss function is used to measure the difference between the predicted result and the actual result. Commonly used loss functions include mean squared error (MSE) and cross-entropy.

[0084] Through the back-propagation algorithm, the contribution of each parameter to the loss function is calculated, and then the values ​​of weights and biases are updated according to the gradient descent algorithm to minimize the loss function.

[0085] Through multiple forward propagation and back propagation optimization, the values ​​of weights and biases are continuously updated until the loss function converges.

[0086] Further optimizing the scheme, the training module includes:

[0087] A progressive training unit, used to train the improved long short-term memory network using the progressive order of data acquisition time in the training set, and simultaneously select a cross entropy loss function, and minimize the cross entropy loss function through a stochastic gradient descent algorithm to generate a progressive training model;

[0088] The constrained training unit is used to divide the training set into an input sequence and a target sequence by using time as a constraint condition to train the progressive training model and obtain the training model.

[0089] To further optimize the solution, the test module is also used to determine whether the training model has learned the characteristics of the operating parameters and time based on the test results. If so, the capacitor state prediction model is output; if not, the training model is re-trained iteratively.

[0090] Further optimizing the scheme, the monitoring subsystem includes:

[0091] A state calculation module, used for inputting the normalized data set into the capacitor state prediction model for calculation to obtain the future operating state and remaining service life of the capacitor;

[0092] A judgment module, used for calculating the failure probability of the future operation state and remaining service life of the capacitor under high altitude weather conditions;

[0093] The display module is used to transmit the fault probability to the terminal for display, and to display the location of the faulty capacitor by a combined positioning method.

[0094] The specific judgment process of the judgment module of this application is as follows:

[0095] The current signal I on the secondary side of the track circuit exceeds 1.3 times its rated value, or the operating voltage value Uc of the power capacitor exceeds 1.1 times its rated value, or the reactive power Qc of the power capacitor exceeds 1.35 times its rated value;

[0096] The deviation rate of the three-phase active power from its rated value is greater than 5%;

[0097] The reactive power Qc at rated voltage is less than its rated value;

[0098] When one of the above conditions is met, it means that the power capacitor is in an abnormal working state.

[0099] like Figure 2As shown, this embodiment also discloses a lightning protection compensation capacitor, including:

[0100] Core connection device, metal casing, capacitor core, short circuit protection device, lead-out cable, grounding device and IoT sensor;

[0101] The core guide device is connected to the metal shell and is used to ensure the structural integrity and electrical performance of the capacitor element during the winding process;

[0102] The capacitor core is arranged inside the metal housing and connected to the anti-short circuit device to improve the transmission performance and stability of the track circuit;

[0103] The anti-short circuit device is arranged inside the metal housing and connected to the lead-out cable for overload protection;

[0104] The grounding device is connected to the metal housing and is used to conduct the stray current accumulated in the track circuit to the ground, while reducing external electromagnetic interference and ensuring signal stability;

[0105] The Internet of Things sensor is connected to the metal shell and is used to obtain the operating parameters of the track compensation capacitor, convert the operating parameters into digital signals, and generate a capacitor operating parameter data set.

[0106] The above are only preferred specific implementations of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. An intelligent monitoring system for lightning protection compensation capacitors, characterized in that: include: A parameter acquisition subsystem, used for acquiring operating parameters of the track compensation capacitor based on sensors, converting the operating parameters into digital signals, and generating a capacitor operating parameter data set; A parameter processing subsystem, connected to the parameter acquisition subsystem, for normalizing the capacitor operation parameter data set to obtain a normalized data set; A central processing subsystem is connected to the parameter processing subsystem and is used to construct a long short-term memory network, and train the long short-term memory network in a progressive order of sample set data acquisition time and using the acquisition time as a constraint condition until the long short-term memory network learns the characteristics of the operating parameters and time and then outputs them to generate a capacitor state prediction model; The monitoring subsystem is connected to the central processing subsystem and is used to input the normalized data set into the capacitor state prediction model for calculation, obtain the capacitor operation state prediction result, and issue an alarm based on the capacitor operation state prediction result.

2. The intelligent monitoring system for lightning protection compensation capacitors according to claim 1 is characterized in that: The parameter acquisition subsystem includes: A voltage signal measurement module is used to collect the real-time signal of the capacitor through a sensor to obtain a parameter signal; A feature extraction module, configured to analyze key features of the parameter signal based on the behavior and performance of the capacitor under different loads to obtain the operating parameters; A conversion module, configured to convert the operating parameters into digital signals based on an analog-to-digital converter to generate the capacitor operating parameter data set; The operating parameters include temperature characteristics, effective value of harmonics, fundamental wave, unbalance and capacitance.

3. The intelligent monitoring system for lightning protection compensation capacitors according to claim 1 is characterized in that: The parameter processing subsystem includes: A data dimension reduction module, used for reducing the dimension of the capacitor operation parameter data set by a principal component analysis method to generate a dimension-reduced data set; A feature engineering module, used to increase the features of the dimension reduction data set based on a feature engineering method to generate a feature data set; The normalization module is used to normalize the feature data set to obtain the normalized data set.

4. The intelligent monitoring system for lightning protection compensation capacitors according to claim 1 is characterized in that: The central processing subsystem includes: A sample set acquisition module, used for acquiring a sample set of a compensation capacitor based on a historical database; A sample set processing module, used for dividing the sample set into a training set and a test set; A model building module, used to build a long short-term memory network, improve the input gate and forget gate of the long short-term memory network, and generate an improved long short-term memory network; A training module, used to train the improved long short-term memory network by using the progressive order of the data acquisition time in the training set and taking the acquisition time as a constraint condition to generate a training model; A testing module is used to test the training model based on the test set and output the capacitor state prediction model.

5. The intelligent monitoring system for lightning protection compensation capacitors according to claim 4 is characterized in that: The model building module includes: A network construction unit is used to construct a long short-term memory network, and optimize the hyperparameters of the long short-term memory network through a genetic algorithm and the characteristics of the capacitor to obtain a long short-term memory network with optimal parameters; Gate calculation module, used to calculate the deep attention gate; A replacement module is used to replace the input gate and the forget gate in the long short-term memory network of the optimal parameters with the deep attention gate to generate the improved long short-term memory network.

6. The intelligent monitoring system for lightning protection compensation capacitors according to claim 4 is characterized in that: The training module includes: A progressive training unit, used to train the improved long short-term memory network using the progressive order of data acquisition time in the training set, and simultaneously select a cross entropy loss function, and minimize the cross entropy loss function through a stochastic gradient descent algorithm to generate a progressive training model; The constrained training unit is used to divide the training set into an input sequence and a target sequence by using time as a constraint condition to train the progressive training model and obtain the training model.

7. The intelligent monitoring system for lightning protection compensation capacitors according to claim 4 is characterized in that: The test module is also used to determine whether the training model has learned the characteristics of the operating parameters and time based on the test results. If so, the capacitor state prediction model is output; if not, the training model is re-trained iteratively.

8. The intelligent monitoring system for lightning protection compensation capacitors according to claim 1 is characterized in that: The monitoring subsystem includes: A state calculation module, used for inputting the normalized data set into the capacitor state prediction model for calculation to obtain the future operating state and remaining service life of the capacitor; A judgment module, used for calculating the failure probability of the future operation state and remaining service life of the capacitor under high altitude weather conditions; The display module is used to transmit the fault probability to the terminal for display, and to display the location of the faulty capacitor by a combined positioning method.

9. A lightning protection compensation capacitor, used in the lightning protection compensation capacitor intelligent monitoring system according to any one of claims 1 to 8, characterized in that: include: Core connection device, metal casing, capacitor core, short circuit protection device, lead-out cable, grounding device and IoT sensor; The core guide device is connected to the metal shell and is used to ensure the structural integrity and electrical performance of the capacitor element during the winding process; The capacitor core is arranged inside the metal housing and connected to the anti-short circuit device to improve the transmission performance and stability of the track circuit; The anti-short circuit device is arranged inside the metal housing and connected to the lead-out cable for overload protection; The grounding device is connected to the metal housing and is used to conduct the stray current accumulated in the track circuit to the ground, while reducing external electromagnetic interference and ensuring signal stability; The Internet of Things sensor is connected to the metal shell and is used to obtain the operating parameters of the track compensation capacitor, convert the operating parameters into digital signals, and generate a capacitor operating parameter data set.