Data identification method and system for cable thermal degradation characteristic gas

Abstract

The invention discloses a data identification method and system for cable thermal degradation characteristic gas, and relates to the technical field of cable fault detection, and the method comprises the steps: carrying out the high-precision synchronous collection of a plurality of gases released in the cable thermal degradation process; the collected data is preprocessed; and packaging and classifying the preprocessed data, and detecting whether an abnormal condition exists or not. According to the method, the SVM model is adopted for classification and recognition, the characteristics of different gases can be efficiently classified, and whether the data deviates from the normal range or not can be recognized through anomaly detection. The optimal solution is searched by dynamically adjusting the wolf pack position, so that the SVM model can adapt to the complex cable degradation state, and the recognition accuracy is improved. The thermal degradation state of the cable can be analyzed according to real-time monitoring data, the degradation process and possible fault types can be accurately predicted, early warning is provided, the accidental fault risk of power equipment is reduced, and safe and stable operation of a power system is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of cable fault detection, and particularly to a data recognition method and system for characteristic gases of cable thermal degradation. Background Art

[0002] With the advancement of the global electrification process, the demand for electricity in the industrial, commercial, and residential sectors continues to increase. The reliability and safety of power equipment have become the focus of people's attention. As an important part of power transmission, cables need to ensure their stable operation.

[0003] Since cables are exposed to complex and variable environments all year round, especially under some conditions such as high voltage, humidity, and chemical corrosion, the degradation process of cables will be accelerated. Long-term overload operation, the influence of environmental temperature changes, and cable aging itself and other factors may lead to the decomposition and damage of cable insulation materials. In the early stage of cable thermal degradation, cables usually release a series of characteristic gases, such as carbon monoxide, methane, carbon dioxide, etc. These gases are important indicators of cable thermal degradation. By identifying the types and concentration changes of these gases, the process of cable degradation and the type of fault can be more effectively reflected. By analyzing the combination of different gases and their concentration changes, it can be distinguished whether the cable is in the early stage of degradation, the mid-stage of deterioration, or approaching the fault critical point, so as to achieve more accurate fault prediction and protection, and thus avoid serious accidents.

[0004] The traditional cable condition monitoring mainly relies on temperature sensors and manual monitoring, and these methods have certain limitations. Therefore, it is of great significance to develop a data recognition method and system for characteristic gases of cable thermal degradation. Summary of the Invention

[0005] In view of the above existing problems, the present invention is proposed.

[0006] Therefore, the technical problem solved by the present invention is: how to obtain the process of cable degradation and the type of fault through the types and concentration changes of the released gases.

[0007] To solve the above technical problem, the present invention provides the following technical solution: A data recognition method for characteristic gases of cable thermal degradation, including: synchronously and highly accurately collecting a variety of gases released during the cable thermal degradation process; preprocessing the collected data; packing and classifying the preprocessed data, and simultaneously detecting whether there are abnormal situations.

[0008] As a preferred scheme of the data recognition method for characteristic gases of cable thermal degradation described in the present invention, wherein: the preprocessing includes filtering and denoising processing, feature extraction, and normalization, and the feature extraction includes the average value, maximum value, minimum value, change rate, and standard deviation of the gas concentration.

[0009] As a preferred solution of the data recognition method for the characteristic gases of cable thermal degradation described in the present invention, wherein: the filtering and denoising process includes moving average filtering, low-pass filtering and median filtering, the median filtering is to take the middle value after sorting the values within the window, and the moving average filtering is expressed as:

[0010]

[0011] The low-pass filtering is expressed as:

[0012] y[n] = a0·x[n] + a1·x[n - 1] + b1·y[n - 1]

[0013] where y[n] is the filtering output at the current time n, x[n - i] is the input at the past i-th time, N is the window size, x[n] is the current input value, y[n] is the current output value, and a0, a1, b1 are filtering coefficients.

[0014] As a preferred solution of the data recognition method for the characteristic gases of cable thermal degradation described in the present invention, wherein: the average gas concentration is expressed as:

[0015]

[0016] The maximum value and the minimum value are respectively expressed as:

[0017] x max = max(x[1], x[2], …, x[N])

[0018] x min = min(x[1], x[2], …, x[N])

[0019] The change rate is expressed as:

[0020]

[0021] The standard deviation is expressed as:

[0022]

[0023] where x[i] is the i-th sampling value, N is the total number of samplings, x[N] is the latest gas concentration value, x[1] is the earliest gas concentration value, and μ is the mean value.

[0024] As a preferred solution of the data recognition method for the characteristic gases of cable thermal degradation described in the present invention, wherein: the normalization is expressed as:

[0025]

[0026] Perform differential calculation after normalization:

[0027] Δx = x[i] - x[i - 1]

[0028] where x norm is the data after normalization, x[i] is the current sampling value, and x[i - 1] is the previous sampling value.

[0029] As a preferred solution of the data recognition method for the characteristic gases of cable thermal deterioration described in the present invention, wherein: the classification includes classifying and abnormally detecting four gas data x1, x2, x3, x4 through an SVM model, and the decision function of the SVM model is:

[0030] f(X) = w T X + b

[0031] Find an optimal hyperplane through SVM to maximize the distance from the sample points to the hyperplane, and the optimization objective is expressed as:

[0032]

[0033] where w is the weight vector, X = [x1, x2, x3, x4] is the input feature vector of the four gas data, b is the bias term, C is the penalty parameter, and ξ i is the slack variable.

[0034] As a preferred solution of the data recognition method for the characteristic gases of cable thermal deterioration described in the present invention, wherein: during the training process of the SVM model, the Grey Wolf Optimization (GWO) algorithm is used to optimize the hyperparameters of the SVM model, including: initializing the wolf pack, randomly initializing the positions of the wolf pack, each position representing a parameter combination of the SVM, dividing the wolf pack into α, β, δ levels, and each level of wolf updates its own position according to the positions of α, β, δ. The position update is expressed as:

[0035]

[0036] Calculate through the following formula:

[0037]

[0038] is a coefficient vector used to adjust the search behavior of the wolf pack, and it, together with helps the wolf pack dynamically adjust its position during the search process. Its calculation formula is:

[0039]

[0040] where is a linearly decreasing vector, is a random vector greater than 1, whose values perform global exploration in the solution space, while those less than 1 cause the wolf pack to perform local exploration around the current optimal solution; is a random vector, is the position of the current wolf, is the position of the current best solution, is a coefficient vector used to control the convergence and exploration of the wolf pack;

[0041] During classification, the SVM model detects whether the data deviates from the normal range. When the classification margin of a data point is too small, or the output value of the SVM model is far from the normal class boundary, it can be marked as abnormal data.

[0042] In a second aspect, another object of the present invention is to provide a data identification system for characteristic gases of cable thermal deterioration, including: a multi-channel gas collection module, a data processing module, a data transmission module, and a monitoring and alarm module; the multi-channel gas collection module is used to synchronously and highly accurately collect various gases released during the cable thermal deterioration process; the data processing module is used to preprocess the collected data; the data transmission module is used to package and transmit the data; the monitoring and alarm module is used to detect whether there is an abnormal situation and alarm for the abnormal situation.

[0043] In a third aspect, a computer device includes a memory and a processor. When the processor executes the computer program, the steps of the data identification method for characteristic gases of cable thermal deterioration as described above are implemented.

[0044] In a fourth aspect, a computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the data identification method for characteristic gases of cable thermal deterioration as described above are implemented.

[0045] Advantages of the present invention: The data recognition method and system for cable thermal degradation characteristic gases provided by the present invention effectively remove interference during the acquisition process through filtering and denoising processing, ensuring the purity of the data. By extracting multiple characteristics of the gas concentration and analyzing the gas concentration changes from multiple dimensions, it provides richer data support for the accurate recognition of the cable degradation process. The normalization processing of the data enables the comparison of data at different time points, eliminates the differences in units or dimensions, and at the same time, the differential calculation can effectively detect the fluctuations in the data to assist in identifying abnormal changes. By using the SVM model for classification and recognition, it can not only efficiently classify the characteristics of different gases but also identify whether the data deviates from the normal range through anomaly detection. The decision function of the SVM model can help clearly define the boundary between normal and abnormal data, improving the accuracy of fault prediction. By dynamically adjusting the positions of the wolf pack to search for the optimal solution, the SVM model can adapt to complex cable degradation states and improve the accuracy of recognition. It can analyze the thermal degradation state of the cable based on real-time monitoring data, accurately predict the degradation process and possible fault types, provide early warnings, reduce the risk of unexpected failures of power equipment, and ensure the safe and stable operation of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0047] Figure 1 It is the overall flowchart of a data recognition method for cable thermal degradation characteristic gases provided by an embodiment of the present invention;

[0048] Figure 2 It is the design schematic diagram of a data recognition system for cable thermal degradation characteristic gases provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0049] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following will provide a detailed description of the specific embodiments of the present invention with reference to the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, not all of them. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the scope of protection of the present invention.

[0050] In the following description, many specific details are set forth in order to provide a thorough understanding of the present invention. However, the present invention may be practiced in other ways than those specifically described herein. Those skilled in the art can make similar generalizations without departing from the spirit of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below.

[0051] Embodiment 1

[0052] Referring to Figure 1 , for an embodiment of the present invention, a method for data recognition of characteristic gases of cable thermal deterioration is provided, including:

[0053] S1: Synchronously and highly accurately collect a variety of gases released during the cable thermal deterioration process;

[0054] Furthermore, in the multi-channel gas collection module, a variety of gases released during the cable thermal deterioration process are synchronously and highly accurately collected. High-precision collection means that the sampling accuracy reaches 12 bits or more, and synchronous collection means that the data of multiple gas sensors are simultaneously input into the STM32F407 main control chip for real-time processing.

[0055] S2: Preprocess the collected data;

[0056] Furthermore, digital filter algorithms such as moving average filtering, low-pass filtering, and median filtering can be used;

[0057] Further, the moving average filter can smooth data fluctuations in the short term. For the STM32F407 microprocessor, it can be implemented in the following way:

[0058]

[0059] where y[n] is the filtered output at the current time n, x[n - i] is the input at the past i-th time, and N is the window size;

[0060] Furthermore, the low-pass filter can achieve a denoising function. The implementation method of the IIR (Infinite Impulse Response) low-pass filter is as follows:

[0061] y[n] = a0·x[n] + a1·x[n - 1] + b1·y[n - 1]

[0062] where x[n] is the current input value, y[n] is the current output value, a0, a1, and b1 are filter coefficients. Each time new data arrives, the filtered output is calculated, and the previous input and output values are retained for recursive calculation;

[0063] It should be noted that the median filter is suitable for eliminating spike noise, mainly by sorting the values within the window and taking the middle value, which can be implemented through a simple sorting algorithm;

[0064] Furthermore, feature extraction is performed on the gas data to extract key statistics from the data, such as the average value, maximum value, change rate, etc.;

[0065] It should be noted that the mean is used to calculate the average gas concentration over a certain period of time, and the implementation formula is:

[0066]

[0067] where x[i] is the i-th sampling value and N is the total number of samplings;

[0068] Furthermore, the maximum and minimum values are used to detect extreme values in the gas concentration to help identify peaks and valleys:

[0069] x max = max(x[1], x[2], …, x[N])

[0070] x min = min(x[1], x[2], …, x[N])

[0071] Furthermore, the change rate is used to measure the rate of change of the gas concentration over time to identify the rising or falling trend of the concentration:

[0072]

[0073] where x[N] is the latest gas concentration value, x[1] is the earliest gas concentration value, and N is the number of samplings within the time window. The change rate can be used to detect situations of rapid concentration changes;

[0074]

[0075] where μ is the mean, x[i] is the i-th sampling value, and N is the number of samples. The standard deviation can be used to detect the stability of the gas concentration. A larger standard deviation indicates more fluctuations;

[0076] It should be noted that the preprocessing step also includes normalization and data packaging to ensure the consistency of the data during transmission and analysis. Normalization can scale the data to a specific range to eliminate the influence of different dimensions. Normalization can be achieved using the following formula:

[0077]

[0078] x norm is the normalized data, x min and x max are the minimum and maximum values of the data respectively. Through normalization, it is ensured that all sensor data is compared and analyzed on the same scale;

[0079] Furthermore, differential calculation is used to obtain the changing trend of data and is suitable for processing multi-channel data. The differential formula is as follows:

[0080] Δx = x[i] - x[i - 1]

[0081] where x[i] is the current sampling value and x[i - 1] is the previous sampling value. Differential calculation can reflect the dynamic changes of data and help identify features with rapid changes.

[0082] S3: After preprocessing the data, pack and classify the data, and at the same time detect whether there are abnormal situations.

[0083] Furthermore, after the data processing is completed, the data is packed into a data frame for transmission; the data packet includes a timestamp, a sensor ID, the processed gas data, and the extracted feature values; the specific format of data packing can adopt a binary format or use a standard JSON format for transmission; the terminal device adopts a low-rate data transmission mode of LORA to ensure that the device reduces power consumption while performing long-distance data transmission.

[0084] Furthermore, use the support vector machine (SVM) machine learning algorithm to separate data of different categories by finding an optimal hyperplane.

[0085] Assume that four types of data, namely x1, x2, x3, and x4, are input, then their feature vector is:

[0086] X = [x1, x2, x3, x4]

[0087] Use SVM to perform classification recognition and anomaly detection on the four types of gas data. The decision function of the SVM model is:

[0088] f(X) = w T X + b

[0089] where w is the weight vector, X = [x1, x2, x3, x4] is the feature vector of the four types of input gas data, and b is the bias term;

[0090] Through the training of SVM, the model can learn the boundaries between different types of gas data to help with classification recognition; by finding an optimal hyperplane through SVM, the distance from the sample points to the hyperplane is maximized, and the optimization objective is:

[0091] where C is the penalty parameter used to control the tolerance of misclassification, and ξ i is the slack variable used to handle the case where the data is linearly inseparable;

[0092] Furthermore, during the SVM training process, the Grey Wolf Optimization (GWO) algorithm is used to optimize hyperparameters such as the penalty parameter of the SVM model to achieve the optimal performance of the classifier; GWO simulates the hunting process of a grey wolf pack. First, the wolf pack is initialized, and the positions of the wolf pack are randomly initialized, with each position representing a parameter combination of the SVM; the wolf pack is divided into α, β, and δ levels, and each level of wolf updates its position according to the positions of α, β, and δ. The position update can be based on the following formula:

[0093]

[0094] where is the position of the current wolf (i.e., the SVM parameter), is the position of the current best solution (i.e., the current optimal SVM parameter), is a coefficient vector used to control the convergence and exploration of the wolf pack. It changes gradually with the increase in the number of iterations, thereby adjusting the search behavior of the grey wolf;

[0095] It should be noted that can be calculated through the following formula:

[0096]

[0097] where is a linearly decreasing vector with an initial value of 2 and a final value of 0. It is used to control the exploration and exploitation of the search process. As the number of iterations increases, gradually decreases from 2 to 0, is a random vector, and the value of each element ranges from [0, 1]. controls the convergence and search behavior of the wolf pack around the prey position. A value greater than 1 allows the wolf pack to conduct global exploration in the solution space, while a value less than 1 prompts the wolf pack to conduct local exploration around the current optimal solution;

[0098] Furthermore, is a coefficient vector used to adjust the search behavior of the wolf pack. It, together with helps the wolf pack dynamically adjust its position during the search process. Its calculation formula is:

[0099]

[0100] where, is a random vector, and the value of each element ranges from [0, 1], ensuring the randomness and diversity of the wolf pack during the search process;

[0101] It should be noted that after multiple iterations, the GWO algorithm will obtain the optimal SVM parameters; the SVM model optimized by GWO is used to classify the four types of gas data and simultaneously detect whether there are abnormal conditions;

[0102] Furthermore, during the classification, SVM can detect whether the data deviates from the normal range. Abnormalities can be detected by observing the classification margin of SVM. If the margin of a data point is too small, or the output value of SVM is far from the normal class boundary, it can be marked as abnormal data;

[0103] Furthermore, the conditions for anomaly detection can be set as:

[0104] |f(X)| < threshold

[0105] If the value of the SVM decision function f(X) is lower than a certain threshold (i.e., the sample point is close to the classification boundary), then the data point is considered an abnormal point; when the monitoring and alarm module detects an abnormality, it immediately triggers an alarm signal to facilitate timely handling of cable faults.

[0106] Embodiment 2

[0107] Refer to Figure 2 , which is an embodiment of the present invention, provides a data identification system for characteristic gases of cable thermal degradation, including: a channel gas collection module, a data processing module, a data transmission module, and a monitoring and alarm module; the channel gas collection module is connected to the main control CPU, and the data transmission module includes a LORA module, N terminal devices, a LoRaWAN gateway module, an STM32H7 microcontroller, an Ethernet PHY chip, a router, and a LoRaWAN network server.

[0108] The multi-channel gas collection module is used to synchronously collect various gases released during the cable thermal degradation process with high precision. The main control CPU uses an STM32F407 chip and is connected to multiple multi-channel gas collection modules through its built-in ADC interface. Each gas collection module communicates with the STM32F407 through an I2C or SPI interface and is used to detect the characteristic gases released during the early cable thermal degradation process. The sensor converts the collected analog gas signal into a digital signal through these interfaces and transmits it to the STM32F407 microcontroller for data processing. The 12-bit ADC of the STM32F407 can achieve high-precision data collection. The multi-channel gas collection module synchronously collects various gases released during the cable thermal degradation process with high precision. Synchronous collection means that the data of multiple gas sensors are simultaneously input into the STM32F407 main control chip for real-time processing. The data processing module is used to preprocess the collected data, and the STM32F407 filters and denoises the collected data.

[0109] The data transmission module is used to package and transmit data. After the data processing is completed, STM32F407 needs to package the data into a data frame for transmission through the LORA module. The LORA module communicates with STM32F407 through the SPI interface and is responsible for sending and receiving data. The LORA module transmits the data to the LoRaWAN gateway for processing.

[0110] The N terminal devices include Terminal Device 1, Terminal Device 2, …, Terminal Device N, etc., where N is greater than or equal to 4. The N terminal devices are placed at different monitoring points of the cable to cover and monitor the working status of the cable. The LoRaWAN gateway consists of an STM32H7 microcontroller, a LoRaWAN gateway module, and an Ethernet chip. The LoRaWAN gateway is responsible for receiving the gas data sent by multiple STM32F407 terminal nodes through the LORA module. The LoRaWAN gateway is connected to the router through the Ethernet interface, and the Ethernet PHY chip in the gateway is used to upload the data. The router forwards the gas data uploaded by the LoRaWAN gateway to the network server. The network server uploads the received gas data to the application server, and the application server is responsible for further processing and analyzing the data for real-time monitoring of the thermal degradation status of the cable and providing fault warnings.

[0111] The application server of the monitoring and alarm system first receives the data from the LoRaWAN gateway and the network server. When the monitoring and alarm module detects an abnormality, it immediately triggers an alarm signal to promptly handle the cable fault.

[0112] Embodiment 3

[0113] An embodiment of the present invention, which is different from the previous two embodiments, is as follows:

[0114] If the said function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to enable a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs, etc., which can store program codes.

[0115] The logic and / or steps represented in the flowchart or otherwise described herein can, for example, be considered as a definitional sequence of executable instructions for implementing logical functions, which can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, a system including a processor, or other systems that can fetch and execute instructions from the instruction execution system, apparatus, or device. As used in this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.

[0116] More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection portion (electronic device) having one or more wirings, a portable computer diskette (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which the program can be printed, since the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or otherwise processing as appropriate, and then stored in a computer memory.

[0117] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, any one or a combination of the following techniques well known in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0118] Example 4

[0119] An embodiment of the present invention provides a method for data identification of characteristic gases of cable thermal deterioration. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0120] Select cable samples of different types and specifications (such as high-voltage cables, low-voltage cables, etc.). Keep the temperature at 25°C ± 2°C in the laboratory environment to simulate common cable working environments, and keep the humidity within the range of 50% ± 10%. Simulate the degradation process of cables under different loads, voltages, and environmental conditions, and conduct early, mid-term, and late-stage degradation tests respectively. Collect data from each cable sample once per minute for 48 hours of continuous monitoring. The data includes the concentrations of different gases (such as CO, CH4, CO2).

[0121] Table 1: SVM Model Classification and Anomaly Detection Results

[0122]

[0123]

[0124] Table 2: Cable Thermal Degradation Gas Concentration Data and Classification Results

[0125]

[0126] As the cable degradation process progresses, the gas concentrations (such as CO, CH4, CO2) gradually increase, especially in the late-stage degradation phase, where the increase in gas concentration is particularly significant. The changes in the maximum concentration, minimum concentration, and average concentration also clearly reflect the different stages of cable degradation.

Claims

1. A data identification method for characteristic gases of cable thermal degradation, characterized in that: include: High-precision synchronous acquisition of multiple gases released during cable thermal degradation; Preprocess the collected data; The preprocessed data is packaged and classified, and any anomalies are detected.

2. The data identification method for characteristic gas of cable thermal degradation according to claim 1, characterized in that: The preprocessing includes filtering and denoising, feature extraction, and normalization. The feature extraction includes the average value, maximum value, minimum value, change rate, and standard deviation of gas concentration.

3. The data identification method for characteristic gas of cable thermal degradation according to claim 2, characterized in that: The filtering and denoising process includes sliding average filtering, low-pass filtering and median filtering. The median filtering is to sort the values ​​in the window and take the middle value. The sliding average filtering is expressed as: The low-pass filtering is expressed as: y[n]=a0·x[n]+a1·x[n-1]+b1·y[n-1] Where y[n] is the filter output at the current time n, x[ni] is the input at the i-th time in the past, N is the window size, x[n] is the current input value, y[n] is the current output value, and a0, a1, and b1 are filter coefficients.

4. The data identification method for characteristic gas of cable thermal degradation according to claim 3, characterized in that: The average gas concentration is expressed as: The maximum and minimum values ​​are expressed as: x max =max(x[1],x[2],…,x[N]) x min =min(x[1],x[2],…,x[N]) The rate of change is expressed as: The standard deviation is expressed as: Where x[i] is the i-th sampling value, N is the total number of samples, x[N] is the latest gas concentration value, x[1] is the earliest gas concentration value, and μ is the mean.

5. The method for identifying the data of characteristic gases of cable thermal degradation according to claim 4, characterized in that: The normalization is expressed as: After normalization, perform difference calculation: Δx=x[i]-x[i-1] Among them, x norm is the normalized data, x[i] is the current sampling value, and x[i-1] is the previous sampling value.

6. The method for identifying the data of characteristic gases of cable thermal degradation according to claim 5, characterized in that: The classification includes classifying and identifying four gas data x1, x2, x3, and x4 and detecting anomalies through an SVM model. The decision function of the SVM model is: f(X)=w T X+b An optimal hyperplane is found through SVM to maximize the distance from the sample point to the hyperplane. The optimization objective is expressed as: Among them, w is the weight vector, X = [x1, x2, x3, x4] is the input feature vector of the four gas data, b is the bias term, C is the penalty parameter, ξ i is a slack variable.

7. The method for identifying the data of characteristic gases of cable thermal degradation according to claim 6, characterized in that: In the SVM model training process, the gray wolf optimization GWO algorithm is used to optimize the hyperparameters of the SVM model, including: initializing the wolf pack, randomly initializing the position of the wolf pack, each position represents a parameter combination of the SVM, dividing the wolf pack into α, β, and δ levels, and each level of wolf updates its position according to the position of α, β, and δ. The position update is expressed as: Calculate using the following formula: is a coefficient vector used to adjust the wolf pack's search behavior, which is related to Together, they help the wolves dynamically adjust their positions during the search process. The calculation formula is: in, is a linearly decreasing vector, is a random vector greater than 1 The value is explored globally in the solution space, and the value less than 1 The value makes the wolf pack conduct local exploration around the current optimal solution; is a random vector, is the current position of the wolf, is the current best solution position, is a coefficient vector used to control the convergence and exploration of the wolf pack; While classifying, the SVM model detects whether the data deviates from the normal range. When the classification margin of a data point is too small, or the output value of the SVM model is far away from the normal category boundary, it can be marked as abnormal data.

8. A system using the data identification method for cable thermal degradation characteristic gas according to any one of claims 1 to 7, characterized in that: include: Multi-channel gas collection module, data processing module, data transmission module and monitoring alarm module; The multi-channel gas collection module is used to perform high-precision synchronous collection of multiple gases released during the thermal degradation of the cable; The data processing module is used to pre-process the collected data; The data transmission module is used to package and transmit data; The monitoring alarm module is used to detect whether there is an abnormal situation and to alarm for the abnormal situation.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for data identification of characteristic gases of thermal degradation of cables according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for data identification of characteristic gases of thermal degradation of cables according to any one of claims 1 to 7 are implemented.

Images