Cable joint fault detection method and device, electronic equipment and storage medium

Abstract

The invention provides a cable joint fault detection method and device, electronic equipment and a storage medium, and belongs to the technical field of cable fault detection, and the method comprises the steps: correcting a cable joint temperature based on a temperature compensation coefficient and cable environment temperature data, and obtaining a target joint temperature; calculating a partial discharge effective value based on the partial discharge pulse amplitude data of the cable joint, and calculating a contact resistance change rate based on the contact resistance data of the cable joint; constructing fusion features based on the target joint temperature, the partial discharge effective value and the contact resistance change rate, and determining a fault judgment result based on the fusion features; and if the fault determination result is a fault, determining a fault type determination result based on the fusion feature, the historical cable joint temperature data, the historical cable environment temperature data, the historical partial discharge pulse amplitude data and the historical contact resistance data. According to the invention, accurate determination of the fault type of the cable joint can be realized.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of cable fault detection, and more particularly to a cable joint fault detection method and device, an electronic device, and a storage medium. BACKGROUND

[0002] The cable joint is a key component for connecting cable segments in a power transmission system, and its performance directly affects the safe and stable operation of the power network. During long-term operation, the cable joint is susceptible to factors such as environmental temperature and humidity, mechanical vibration, and electrical load impact, which can cause local overheating, insulation aging, and poor contact, etc. If not detected and processed in a timely manner, it may cause line tripping, equipment burning, and even large-scale power outage accidents. Therefore, accurate detection of cable joint faults has important engineering significance.

[0003] Currently, cable joint fault detection mainly relies on single parameter monitoring or simple threshold judgment methods. For example, the joint surface temperature is monitored by a temperature sensor, and when the temperature exceeds a preset threshold, it is determined to be a fault; or a partial discharge detector is used to collect discharge signals, and the insulation state is determined based on the discharge amount. However, the existing technology has obvious defects: on the one hand, environmental interference factors can easily cause parameter monitoring values to be distorted, resulting in false positives or false negatives; on the other hand, it is difficult to distinguish fault types by using only a single parameter or static threshold, which cannot meet the needs of the power system for cable joint fault detection. SUMMARY

[0004] The purpose of the present application is to provide a cable joint fault detection method and device, an electronic device, and a storage medium to accurately determine the type of cable joint fault.

[0005] The first aspect of the embodiments of the present application provides a cable joint fault detection method, comprising: obtaining cable joint temperature and cable environment temperature data; determining the scene type based on the voltage level and laying method of the cable joint, and determining the temperature compensation coefficient based on the scene type; correcting the cable joint temperature based on the temperature compensation coefficient and the cable environment temperature data to obtain the target joint temperature; calculating the partial discharge effective value based on the partial discharge pulse amplitude data of the cable joint, and calculating the contact resistance change rate based on the contact resistance data of the cable joint; constructing a fusion feature based on the target joint temperature, the partial discharge effective value, and the contact resistance change rate, and determining a fault determination result based on the fusion feature; If the fault determination result is a fault, then the fault type determination result is determined based on the fusion feature, the historical cable joint temperature data, the historical cable environment temperature data, the historical partial discharge pulse amplitude data of the cable joint, and the historical contact resistance data of the cable joint.

[0006] A second aspect of this application provides a cable joint fault detection device, comprising: The temperature correction module is used to acquire cable joint temperature and cable ambient temperature data; determine the scenario type based on the voltage level and laying method of the cable joint; determine the temperature compensation coefficient based on the scenario type; and correct the cable joint temperature based on the temperature compensation coefficient and cable ambient temperature data to obtain the target joint temperature. The electrical data processing module is used to calculate the effective value of partial discharge based on the partial discharge pulse amplitude data of the cable joint, and to calculate the contact resistance change rate based on the contact resistance data of the cable joint. The fault determination module is used to construct fusion features based on the target joint temperature, the effective value of partial discharge and the rate of change of contact resistance, and to determine the fault determination result based on the fusion features. The fault type determination module is used to determine the fault type determination result based on fusion characteristics, historical cable joint temperature data, historical cable ambient temperature data, historical partial discharge pulse amplitude data of cable joints, and historical contact resistance data of cable joints if the fault determination result is a fault.

[0007] A third aspect of this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor executes the computer program to implement the steps of the above-described cable joint fault detection method.

[0008] A fourth aspect of this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the cable joint fault detection method described above.

[0009] The beneficial effects of the cable joint fault detection method and device, electronic equipment, and storage medium provided in this application are as follows: This application first determines the actual scenario based on the voltage level and installation method of the cable joint, and then matches the corresponding temperature correction value. This correction value eliminates the interference of ambient temperature on the joint temperature, making the final joint temperature data more consistent with the actual operating conditions and avoiding misjudgment of faults due to ambient temperature. This application no longer relies on a single data point, but combines three key pieces of information—the corrected joint temperature, partial discharge intensity, and contact resistance change—to determine whether a fault exists. If a fault is determined, historical operating data is also referenced. By observing the trends reflected in the historical data, the issue is accurately identified as localized overheating, insulation aging, or poor contact, allowing for direct and accurate problem identification during maintenance, avoiding blind troubleshooting.

[0010] In conclusion, the embodiment of the present application can reduce fault misjudgment and omission, accurately locate the fault type, reduce the risk of power line tripping, speed up the repair, reduce the power outage time, and ensure the stable supply of power. BRIEF DESCRIPTION OF DRAWINGS

[0011] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings needed to be used in the embodiments or prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0012] Figure 1 The flowchart of the cable joint fault detection method provided by an embodiment of the present application is shown. Figure 2 The structural block diagram of the cable joint fault detection device provided by an embodiment of the present application is shown. Figure 3 The schematic block diagram of the electronic device provided by an embodiment of the present application is shown. DETAILED DESCRIPTION

[0013] In the following description, specific details are set forth in order to provide a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application can be practiced in other embodiments without these specific details. In other instances, well-known systems, devices, circuits, and methods have not been described in detail in order to avoid obscuring the description of the present application.

[0014] In order to make the purpose, technical solutions and advantages of the present application clearer, the following will be described by specific embodiments in conjunction with the drawings.

[0015] Reference Figure 1 , Figure 1 The flowchart of the cable joint fault detection method provided by an embodiment of the present application is shown. The method can be executed by an electronic device, specifically, the method can include S101 to S104.

[0016] S101: Obtain cable joint temperature and cable environment temperature data; determine the scene type based on the voltage level and laying mode of the cable joint, determine the temperature compensation coefficient based on the scene type; correct the cable joint temperature based on the temperature compensation coefficient and the cable environment temperature data to obtain the target joint temperature.

[0017] In the embodiment, the scene type includes a high-voltage scene or a non-high-voltage scene, and a closed scene or an open scene; the scene type is determined based on the voltage level and the laying mode of the cable joint, and specifically includes: determining the scene type as a high-voltage scene or a non-high-voltage scene based on the voltage level of the cable joint; and determining the scene type as a closed scene or an open scene based on the laying mode of the cable joint.

[0018] In the embodiment, the temperature compensation coefficient is determined based on the scene type, and specifically includes: If the scene type is a high-voltage scene and a closed scene, the temperature compensation coefficient is determined as a first temperature compensation coefficient. If the scene type is a high-voltage scene and an open scene, the temperature compensation coefficient is determined as a second temperature compensation coefficient. If the scene type is a medium-low-voltage scene and a closed scene, the temperature compensation coefficient is determined as a third temperature compensation coefficient. If the scene type is a medium-low-voltage scene and an open scene, the temperature compensation coefficient is determined as a fourth temperature compensation coefficient. The fourth temperature compensation coefficient is greater than the first temperature compensation coefficient and less than the third temperature compensation coefficient, and the first temperature compensation coefficient is greater than the second temperature compensation coefficient.

[0019] In the embodiment, the cable joint temperature refers to the temperature generated by the cable joint itself during operation, which is a basic parameter for reflecting whether the joint has an overheating fault. The cable environment temperature data refers to the temperature information of the surrounding environment of the cable joint, which is used to analyze the interference of the environmental temperature on the joint temperature. The voltage level of the cable joint refers to the voltage range that the cable joint is designed to withstand, which is used to distinguish between high-voltage scenes and non-high-voltage scenes and adapt to different temperature interference characteristics. The laying mode of the cable joint refers to the specific form of the cable joint installation with the cable, which is used to distinguish between closed scenes and open scenes and adapt to different heat dissipation conditions. The scene type refers to the classification of the cable joint operation scene formed according to the combination of the voltage level and the laying mode, and the scene type can provide a classification standard for matching the corresponding temperature compensation coefficient and ensure the pertinence of the compensation coefficient. The high-voltage scene refers to a scene in which the rated voltage of the cable joint is in a high-voltage range (usually ≥110kV), and the high-voltage scene can represent the characteristics of the joint in this scene, such as high rated temperature rise and low environmental temperature interference ratio, which provides a basis for selecting the compensation coefficient. The non-high-voltage scene, i.e. the medium-low-voltage scene, refers to a scene in which the rated voltage of the cable joint is in a medium-low-voltage range (usually ≤35kV), and the non-high-voltage scene can represent the characteristics of the joint in this scene, such as low rated temperature rise and high environmental temperature interference ratio, which provides a basis for selecting the compensation coefficient.

[0020] The closed scenario refers to a laying scenario of the cable joint in an environment with limited heat dissipation conditions (such as direct burial and pipeline laying), and the closed scenario can represent the characteristics that the change of the ambient temperature in this scenario is easily directly conducted to the joint and the interference is significant, thereby providing a basis for the selection of the compensation coefficient. The open scenario refers to a laying scenario of the cable joint in an environment with good heat dissipation conditions (such as cable trench and bridge laying), and the open scenario can represent the characteristics that the ambient temperature interference in this scenario is weakened and the heat dissipation is efficient, thereby providing a basis for the selection of the compensation coefficient. The temperature compensation coefficient refers to a correction parameter for offsetting the interference of the ambient temperature on the joint temperature, and the ambient temperature data can be corrected by combining the ambient temperature data to ensure that the corrected data is consistent with the actual operation state of the joint.

[0021] The first temperature compensation coefficient refers to a temperature compensation coefficient used when the scenario type is a high-voltage scenario and a closed scenario, which can adapt to the interference characteristics of the high rated temperature rise of the high-voltage joint and the closed environment, and realize accurate temperature correction. The second temperature compensation coefficient refers to a temperature compensation coefficient used when the scenario type is a high-voltage scenario and an open scenario, which can adapt to the weak interference characteristics of the high rated temperature rise of the high-voltage joint and the open environment, and realize accurate temperature correction. The third temperature compensation coefficient refers to a temperature compensation coefficient used when the scenario type is a medium / low-voltage scenario and a closed scenario, which can adapt to the strong interference characteristics of the low rated temperature rise of the medium / low-voltage joint and the closed environment, and realize accurate temperature correction. The fourth temperature compensation coefficient refers to a temperature compensation coefficient used when the scenario type is a medium / low-voltage scenario and an open scenario, which can adapt to the medium interference characteristics of the low rated temperature rise of the medium / low-voltage joint and the open environment, and realize accurate temperature correction. The target joint temperature refers to the cable joint temperature corrected by the temperature compensation coefficient and the ambient temperature data, which is the real temperature data after eliminating the environmental interference, and provides reliable temperature basis for subsequent fault detection.

[0022] The consideration behind the present embodiment is that the cable joint temperature monitoring is easily interfered by the ambient temperature, and the interference degree is jointly determined by the joint voltage grade and the laying method. In the high-voltage scenario, the rated temperature rise of the joint is high (usually 30-50°C), the ambient temperature fluctuation accounts for a low proportion of the total temperature rise, and the interference effect is weak; in the medium / low-voltage scenario, the rated temperature rise of the joint is low (usually 10-20°C), the ambient temperature interference accounts for a high proportion, and the effect is stronger. The closed scenario has poor heat dissipation, the change of the ambient temperature is easily directly conducted to the joint, and the interference is significant; the open scenario has good heat dissipation, and the environmental interference is weakened. Based on this logic, the present embodiment divides the scenario type according to the voltage grade + laying method, matches different temperature compensation coefficients for different scenarios, and can counteract the environmental interference. The stronger the interference of the scenario, the greater the compensation coefficient, so as to ensure that the corrected target joint temperature is accurate and avoid the correction deviation caused by a single coefficient, thereby providing a reliable data basis for subsequent fault detection.

[0023] For example, the embodiment can be applied to the temperature correction of the 10kV cable joint (medium and low voltage) and the 220kV power cable joint (high voltage) of the urban power distribution network, as follows: (1) Obtain the cable joint temperature and cable environment temperature data. For the 10kV power distribution network cable joint, the embodiment can install a PT100 platinum resistance temperature sensor at the junction of the metal shielding layer and the insulating layer of the joint, and the sensor is connected to the local monitoring terminal through a wire. The joint temperature data is collected every 10 seconds. The embodiment can install a temperature and humidity sensor in the cable well where the joint is located, 1 meter away from the joint and away from the air outlet. The environment temperature data is also collected every 10 seconds. Both types of data are attached with time stamps and transmitted to the real-time database storage of the monitoring terminal. For the 220kV power cable joint, the embodiment can use the same type of sensor, and the installation position is adjusted to the vicinity of the stress cone of the joint (the core heating area of the high-voltage joint). The environment temperature sensor is installed at a fixed monitoring point in the cable tunnel where the joint is located, and the temperature data is collected in real time.

[0024] (2) Determine the scene type based on the voltage level and laying method of the cable joint. First, the embodiment can determine the voltage level by checking the product nameplate or operation and maintenance archives of the cable joint. If the rated voltage is marked as 10kV, it is determined as a medium and low voltage scene (not a high voltage scene). If the rated voltage is marked as 220kV, it is determined as a high voltage scene. Then the embodiment can determine the laying method by checking the cable laying engineering drawings or on-site inspection. If the 10kV cable joint adopts the direct burial method (the joint is buried in the underground soil without a special heat dissipation channel), it is determined as a closed scene. If the 220kV cable joint adopts the cable trench laying (the joint is located in the underground cable trench with a cover plate, which has natural ventilation conditions), it is determined as an open scene. Finally, the scene type is obtained by combining: the 10kV direct buried joint is a medium and low voltage scene and a closed scene, and the 220kV cable trench joint is a high voltage scene and an open scene.

[0025] (3) The embodiment can determine the matching temperature compensation coefficient for different scene types according to historical data analysis. For example, the medium and low voltage scene and the closed scene correspond to the third temperature compensation coefficient, with a value of 0.4; the medium and low voltage scene and the open scene correspond to the fourth temperature compensation coefficient, with a value of 0.3; the high voltage scene and the closed scene correspond to the first temperature compensation coefficient, with a value of 0.2; and the high voltage scene and the open scene correspond to the second temperature compensation coefficient, with a value of 0.1. According to the scene type determined in the second step, the 10kV direct buried joint matches the third temperature compensation coefficient 0.4, and the 220kV cable trench joint matches the second temperature compensation coefficient 0.1.

[0026] (4) Based on the temperature compensation coefficient and the cable ambient temperature data, the cable joint temperature is corrected to obtain the target joint temperature. For a 10kV direct-buried joint, the embodiment can obtain the joint temperature data (such as 78℃) and the ambient temperature data (such as 32℃) at a certain timestamp, subtract the standard reference temperature (25℃) from the ambient temperature to obtain the ambient temperature deviation (7℃), multiply the ambient temperature deviation by the third temperature compensation coefficient 0.4 to obtain the environmental interference value (2.8℃), and finally subtract the environmental interference value from the joint temperature to obtain the target joint temperature 75.2℃. For a 220kV cable trench joint, the embodiment can retrieve the joint temperature data (such as 85℃) and the ambient temperature data (such as 28℃) at the same timestamp, calculate the ambient temperature deviation 3℃, multiply the ambient temperature deviation by the second temperature compensation coefficient 0.1 to obtain the environmental interference value 0.3℃, and correct to obtain the target joint temperature 84.7℃. The corrected target joint temperature is stored in the feature database for subsequent fusion feature construction and fault judgment.

[0027] S102: Calculate the partial discharge effective value based on the partial discharge pulse amplitude data of the cable joint, and calculate the contact resistance change rate based on the contact resistance data of the cable joint.

[0028] In this embodiment, the partial discharge pulse amplitude data of the cable joint refers to the amplitude size information of the pulse signal when the cable joint generates partial discharge due to insulation defects, which provides raw data for calculating the partial discharge effective value and reflects the basic parameters of the joint insulation state. The partial discharge effective value refers to a quantitative parameter representing the discharge intensity calculated based on the partial discharge pulse amplitude data. The contact resistance data of the cable joint refers to the resistance value of the contact part of the cable joint, which reflects the basic parameters of the joint contact state. The contact resistance change rate refers to a parameter representing the degree of resistance change calculated based on the contact resistance data.

[0029] In this embodiment, considering that the partial discharge intensity is directly related to the insulation state of the cable joint, insulation aging will cause the discharge effective value to increase; the degree of contact resistance change is directly related to the contact state of the joint, and poor contact will cause the resistance change rate to increase. Both are key characterization parameters of faults, and it is difficult to quantify the fault degree only by the original pulse amplitude or resistance value. Therefore, by calculating the effective value and the change rate, the original data can be converted into a quantitative index that can be directly used for fault judgment, laying a foundation for subsequent fusion feature construction and ensuring the accuracy of fault judgment.

[0030] For example, the embodiment can be applied to parameter calculation of 10kV distribution network cable joints, and the specific steps are as follows: (1) Obtain partial discharge pulse amplitude data of the cable joint. In this embodiment, a high-frequency current sensor can be sleeved on the grounding wire of the cable joint, and the sensor continuously collects 1 million partial discharge pulse amplitude data per second. During the collection process, the sensor zero drift signal is filtered synchronously to obtain effective pulse amplitude data.

[0031] (2) Calculate the partial discharge effective value based on the partial discharge pulse amplitude data. In this embodiment, the pulse amplitude data of 1 second can be retrieved, and then wavelet threshold denoising is performed. Then, the square of all pulse amplitude data after denoising is calculated, and the average value is calculated. Finally, the average value is square rooted to obtain the partial discharge effective value of the 1-second period.

[0032] (3) Obtain the contact resistance data of the cable joint. In this embodiment, an online low resistance tester can be used, and the two test clamps of the tester are connected to the terminal clamps at both ends of the cable joint. The tester collects contact resistance data at a frequency of every 5 minutes, and records the resistance data collected for the first time when the joint is installed as the initial contact resistance. In this embodiment, the difference between the current resistance and the initial resistance can be calculated, and then the difference is divided by the initial resistance to obtain the contact resistance change rate. The contact resistance change rate is associated with the partial discharge effective value calculated at the same period to prepare for the subsequent construction of fusion features.

[0033] S103: Construct fusion features based on the target joint temperature, the partial discharge effective value, and the contact resistance change rate, and determine the fault determination result based on the fusion features.

[0034] In this embodiment, the fusion features are comprehensive features formed by integrating the target joint temperature, the partial discharge effective value, and the contact resistance change rate, which are used to provide multi-dimensional basis for fault determination and avoid single parameter deviation. The fault determination result is the conclusion that the joint is normal or faulty based on the fusion features, which provides a premise for subsequent fault type determination.

[0035] In this embodiment, considering that three different types of cable joint faults, i.e., local overheating, insulation aging, and poor contact, will cause temperature, partial discharge, and contact resistance parameters to be abnormal, respectively, a single parameter is easily disturbed to cause misjudgment. This embodiment constructs fusion features by integrating three types of core parameters, which can comprehensively reflect the multi-dimensional state of joint heat, insulation, and contact. Then, based on the fusion features, the fault can be determined, which can avoid the limitations of a single parameter and improve the accuracy and reliability of fault determination, thereby laying a foundation for subsequent accurate positioning of fault types.

[0036] For example, this embodiment can be applied to 10kV urban distribution network cable joint fault determination, and the specific steps are as follows: Suppose that the target joint temperature (e.g., 75°C), the partial discharge effective value (e.g., 0.4V), and the contact resistance change rate (e.g., 6%) are calculated, and the rated parameter threshold of the cable joint of this model is: the rated temperature rise 70°C, the partial discharge effective value threshold 0.3V, and the contact resistance change rate threshold 5%.

[0037] In this embodiment, the target joint temperature, the partial discharge effective value, and the contact resistance change rate are combined in the order of temperature-discharge-resistance to form a fusion feature vector, such as [75°C, 0.4V, 6%], and the three types of parameters are standardized, such as the difference between the target temperature and the rated temperature rise, the ratio of the discharge effective value to the threshold, and the ratio of the resistance change rate to the threshold, to ensure the uniformity of the parameter magnitude and improve the effectiveness of the feature.

[0038] In this embodiment, a pre-trained lightweight random forest model (containing 8 decision trees) is used. The standardized fusion feature vector is input into the model, each decision tree outputs a normal or fault result based on the comparison between the feature and the threshold, and the outputs of the 8 trees are counted according to the rule of majority. If 6 or more trees determine a fault, the final fault determination result is a fault, and a fault warning signal is generated; if 6 or more trees determine normal, the result is normal, and the current fusion feature is stored in the historical database for subsequent time series analysis.

[0039] S104: If the fault determination result is a fault, determine the fault type determination result based on the fusion feature, the historical cable joint temperature data, the historical cable environment temperature data, the historical partial discharge pulse amplitude data of the cable joint, and the historical contact resistance data of the cable joint.

[0040] In this embodiment, the historical cable joint temperature data refers to the temperature records of the cable joint in the past period of time, which provides the basis for the temperature change trend for fault type determination and assists in identifying progressive heat-related faults. The historical cable environment temperature data refers to the temperature records of the environment where the cable joint is located in the past period of time, which analyzes the influence of the environment on the temperature in combination with the historical joint temperature to exclude the misleading of environmental interference on fault type determination. The historical partial discharge pulse amplitude data of the cable joint refers to the records of the partial discharge pulse amplitude of the cable joint in the past period of time, which provides the basis for the discharge intensity change trend for fault type determination and assists in identifying insulation-related faults. The historical contact resistance data of the cable joint refers to the records of the contact resistance of the cable joint in the past period of time, which provides the basis for the resistance change trend for fault type determination and assists in identifying contact state-related faults. The fault type determination result refers to the specific fault category conclusion (such as local overheating, insulation aging, and poor contact) based on real-time and historical data, which clarifies the nature of the fault and provides direct guidance for targeted maintenance.

[0041] In this embodiment, considering that the real-time features of different fault types of cable joints (local overheating, insulation aging, poor contact) may exist in cross, such as temperature rise caused by both local overheating and poor contact, it is difficult to accurately distinguish by real-time fusion features alone. The historical data can reflect the fault development trend. Insulation aging will show a trend of continuous increase in local discharge pulse amplitude, poor contact will show a trend of gradually increasing contact resistance, and local overheating will show a trend of continuous expansion of the deviation between joint temperature and environmental temperature.

[0042] Based on this logic, this embodiment combines fusion features (current state) and multi-dimensional historical data (trend), which can analyze fault types from both current state and development process, avoid type misjudgment caused by single real-time data, ensure that the fault type determination result can directly match the maintenance demand, improve the maintenance efficiency and accuracy, and solve the problem that the prior art cannot accurately classify faults by relying on real-time data alone.

[0043] For example, this embodiment can be applied to fault type determination of 10kV distribution network cable joints, and the specific steps are as follows: (1) This embodiment can first obtain the fusion features corresponding to the fault determination result when the fault is determined, including the target joint temperature, the local discharge effective value, and the contact resistance change rate. Then, the historical data of the cable joint in the past 30 days is obtained, wherein the historical cable joint temperature data is 144 groups of temperature values collected every 10 minutes per day, the historical cable environmental temperature data is the highest, lowest and average environmental temperature per day, the historical local discharge pulse amplitude data of the cable joint is the maximum pulse amplitude within 1 second per day, and the historical contact resistance data of the cable joint is 3 groups of contact resistance values collected per day (1 group in the morning, 1 group in the afternoon, and 1 group in the evening).

[0044] (2) For the historical cable joint temperature data, this embodiment can calculate the average temperature per day, draw a trend curve with the number of days as the horizontal axis and the daily average temperature as the vertical axis, and analyze whether the temperature shows a continuous rising trend. For the historical cable environmental temperature data, the deviation value of the environmental temperature from 25°C per day is calculated, and the correlation between the deviation and the historical joint temperature deviation is compared to determine whether the joint temperature change is dominated by the environment. For the historical local discharge pulse amplitude data, this embodiment can calculate the ratio of the maximum pulse amplitude per day to the local discharge effective value, and observe whether the ratio shows a continuous increasing trend. For the historical contact resistance data, this embodiment can calculate the difference between the resistance value per day and the initial contact resistance, and observe whether the difference gradually expands.

[0045] (3) If the target joint temperature in the fusion feature is significantly higher than the rated value, the historical joint temperature shows a sustained upward trend, and the correlation between the historical environmental temperature deviation and the joint temperature deviation is weak, and the historical partial discharge pulse amplitude and the historical contact resistance have no obvious abnormalities, then the fault type is determined to be local overheating; if the fusion feature shows that the partial discharge effective value is significantly exceeded, the historical partial discharge pulse amplitude is continuously increasing and the ratio of the maximum amplitude to the effective value is increasing, and other data have no obvious abnormalities, then the fault type is determined to be insulation aging; if the fusion feature shows that the contact resistance change rate is exceeded, the difference between the historical contact resistance and the initial value gradually expands, the historical joint temperature shows a slow upward trend with the increase of the resistance, and other data have no obvious abnormalities, then the fault type is determined to be poor contact.

[0046] From the above, the embodiment of the present application first determines the actual scene according to the voltage level and installation method of the cable joint, and then matches the corresponding temperature correction value to eliminate the interference of the environmental temperature on the joint temperature. The final joint temperature data obtained can better fit the real operating state, avoiding misjudgment of faults due to environmental cold and heat. The embodiment of the present application no longer relies on a single data, but combines the three key information of the corrected joint temperature, the partial discharge intensity and the contact resistance change to determine whether there is a fault. If it is determined that there is a fault, the historical operation data is also referred to, and the change trend reflected by the historical data is used to accurately identify whether it is local overheating, insulation aging or poor contact, so that the problem can be directly found out during maintenance without blind troubleshooting.

[0047] In summary, the embodiment of the present application can reduce fault misjudgment and omission, accurately locate the fault type, reduce the risk of power line tripping, speed up the maintenance, reduce the power outage time, and ensure the stable supply of power.

[0048] In an embodiment of the present application, the discharge effective value is calculated based on the partial discharge pulse amplitude data of the cable joint, comprising: The partial discharge pulse amplitude data of the cable joint is adaptively denoised by a wavelet threshold to obtain a denoised partial discharge signal. The discharge effective value is calculated based on the denoised partial discharge signal.

[0049] In the embodiment, the partial discharge pulse amplitude data of the cable joint is adaptively denoised by a wavelet threshold to obtain a denoised partial discharge signal, specifically comprising: The partial discharge pulse amplitude data of the cable joint is wavelet decomposed based on a target decomposition layer to obtain a target approximation coefficient and a detail coefficient corresponding to each target decomposition layer; A first operation is performed on the detail coefficient corresponding to each target decomposition layer to obtain a noise-filtered detail coefficient of each layer; The filtered detail coefficient of each layer and the target approximation coefficient are reconstructed by wavelet inverse transform to obtain the denoised partial discharge signal. wherein the first operation comprises: extracting background noise data in the detail coefficients corresponding to the target decomposition layer; calculating a standard deviation of the background noise data, and determining a noise threshold corresponding to the target decomposition layer based on the standard deviation; performing noise filtering on the detail coefficients corresponding to the target decomposition layer based on the noise threshold.

[0050] In the embodiment, the adaptive wavelet threshold denoising refers to a wavelet denoising method that dynamically adjusts the threshold according to the noise characteristics of the partial discharge pulse amplitude data. The denoised partial discharge signal refers to a signal that retains the effective discharge components after noise filtering. The target decomposition layer number refers to the preset decomposition level during wavelet decomposition, which can be determined in combination with the sampling frequency. The target approximation coefficient refers to the coefficient reflecting the overall trend of the signal after wavelet decomposition, which can include parameters such as signal amplitude mean, trend slope, etc. The detail coefficient corresponding to the target decomposition layer refers to the coefficient reflecting the high-frequency components (including noise and effective pulses) of the signal at each decomposition layer, which can include parameters such as high-frequency amplitude, high-frequency fluctuation frequency, etc. The first operation refers to a series of steps for noise filtering of the detail coefficient, which can include operation parameters such as noise extraction, threshold calculation, and coefficient filtering, etc. The background noise data refers to the noise segment without effective discharge in the detail coefficient. The standard deviation refers to the dispersion degree index of the background noise data. The noise threshold refers to the critical value that distinguishes noise from effective pulses, which can be a multiple coefficient of the standard deviation, such as 1.2 times, 1.5 times the standard deviation, etc. The wavelet inverse transform refers to a transformation method that reconstructs the processed coefficient into a signal.

[0051] The consideration behind this embodiment is that the partial discharge pulse amplitude data is susceptible to electromagnetic noise interference, and fixed threshold denoising is prone to lose effective pulses or retain noise, resulting in deviation in subsequent effective value calculation. The adaptive wavelet threshold denoising can dynamically adjust the threshold according to the decomposition layer, and adapt to the noise characteristics of different levels. Therefore, this embodiment first separates the signal trend and high-frequency components (including noise) through wavelet decomposition, then extracts the background noise and calculates the dedicated threshold for each layer of detail coefficient, avoiding the limitations of a single threshold; finally, the signal is reconstructed to retain the effective components. This design can accurately filter noise and ensure that the denoised signal truly reflects the discharge state, providing reliable data for subsequent discharge effective value calculation, solving the problem of insufficient precision of fixed threshold denoising.

[0052] For example, the embodiment can be applied to 10kV cable joint partial discharge signal processing, and the specific steps are as follows: (1) For the partial discharge pulse amplitude data with a sampling frequency of 1 MHz, the embodiment can set the target decomposition layer number to be 3 layers in combination with field test experience; select db4 wavelet base as the decomposition base function, and perform wavelet decomposition on the collected continuous 1-second pulse amplitude data (100 million data points in total) to obtain one group of target approximation coefficients and three groups of detail coefficients. The target approximation coefficients can reflect the overall trend of the signal, and there are about 250,000 data points. The three groups of detail coefficients correspond to the first to third decomposition layers respectively, and each group has about 250,000 data points, containing high-frequency noise and effective pulses.

[0053] (2) Taking the first decomposition layer as an example, the embodiment can first extract the first 1000 data points of the detail coefficients of this layer (verified by historical data, this segment has no effective discharge and is background noise data); the embodiment can calculate the standard deviation of the background noise data, and if the standard deviation is 0.8V, the threshold coefficient is set to 1.2, and the noise threshold of this layer is determined based on the standard deviation, which is 0.96V (1.2 times the standard deviation); the embodiment can set the data (judged as noise) in the detail coefficients of this layer with an absolute value less than 0.96V to 0, and retain the data (judged as effective pulses) with an absolute value greater than or equal to 0.96V, to obtain the first layer of detail coefficients after noise filtering. The second and third decomposition layers of detail coefficients are processed in the same way to obtain the corresponding noise-filtered detail coefficients.

[0054] (3) The embodiment can input the three layers of filtered detail coefficients and the untreated target approximation coefficients into the wavelet inverse transform process, use the same db4 wavelet base as the decomposition, perform inverse transform operation, and reconstruct to obtain the 1-second long partial discharge signal after noise reduction. In the reconstructed signal, the electromagnetic noise amplitude is reduced from the original 2V to below 0.3V, and the waveform distortion of the effective discharge pulse is controlled within 5%.

[0055] (4) For the partial discharge signal after noise reduction, the embodiment can extract all pulse amplitude data within 1 second, square each data point first, calculate the average of all square values, and then take the square root of the average to obtain the discharge effective value of this period, which is used for subsequent fusion feature construction and fault judgment.

[0056] The embodiment can accurately filter electromagnetic noise in the partial discharge pulse amplitude data, adapt to the noise characteristics of each decomposition layer through adaptive threshold, avoid loss of effective signals or residual noise caused by fixed threshold, and ensure that the partial discharge signal after noise reduction is real and reliable. The discharge effective value calculated based on the signal has higher accuracy and can accurately reflect the insulation state of the cable joint, reducing the false judgment or missed judgment of faults caused by noise interference. At the same time, the wavelet decomposition and inverse transform process adapts to the characteristics of the engineering actual data, does not require complex hardware, is easy to popularize in existing detection scenes, and provides high-quality data support for subsequent fault judgment.

[0057] In an embodiment of the present application, the fault determination result is determined based on the fusion feature, comprising: The fault determination result is determined by a light random forest model based on the fusion feature; the fault determination result includes no fault or fault; The light random forest model is trained based on a plurality of historical fusion feature samples and a plurality of historical fusion feature samples respectively corresponding to historical fault labels.

[0058] In the embodiment, the light random forest model refers to a random forest model with simplified calculation complexity, which may include parameters such as the number of decision trees (e.g. 8-15), the maximum depth of each tree (e.g. 3-6 layers), and the number of random feature selection, and is suitable for low-power scenarios. No fault refers to the conclusion that the cable joint is in normal operating state in the fault determination result. The historical fusion feature sample refers to a set of fusion feature data collected in the past, which may include historical target joint temperature, historical partial discharge effective value, historical contact resistance change rate, etc., and is stored in time sequence. The historical fault label refers to the no fault or fault label corresponding to the historical fusion feature sample, which is used for model training.

[0059] In the embodiment, considering that existing fault determination is mostly dependent on a single parameter threshold and is easily misjudged by interference, while the random forest model can fuse multiple features to achieve accurate classification, but the traditional model has large calculation amount and is difficult to deploy on site detection equipment. Logically, the light random forest model is used to reduce the number of decision trees, control the tree depth to reduce the calculation load, and balance the model accuracy and deployment feasibility; the model is trained with historical fusion feature samples and corresponding historical fault labels, so that the model can learn the feature law under different states, output no fault or fault result based on real-time fusion feature, avoid single parameter limitation, ensure accurate fault determination and adapt to on-site hardware, and solve the contradiction between low accuracy and difficult deployment of traditional model.

[0060] For example, the embodiment can be applied to 10kV distribution network cable joint fault determination, and the specific steps are as follows: (1) The embodiment can collect historical data of such cable joints, extract 3000 groups of historical fusion feature samples, of which 1800 groups are no fault samples and 1200 groups are fault samples. The no fault sample corresponds to the historical fault label of no fault, the target joint temperature is within the rated temperature rise ±5℃, the partial discharge effective value is ≤0.3V, and the contact resistance change rate is ≤5%. The fault sample corresponds to the historical fault label of fault, at least one parameter exceeds the above range, and all samples are attached with basic information such as collection time and joint model.

[0061] (2) The embodiment can set the model parameters: the number of decision trees is 8, the maximum depth of each tree is 5 layers, and 2 fusion features are selected from the target joint temperature, the partial discharge effective value and the contact resistance change rate each time the feature is split; the embodiment can divide 3000 groups of samples into a training set and a validation set in a ratio of 7:3, train the model with the training set, and adjust the parameters through the validation set. If the validation set determines that the accuracy is less than 90%, 2 more decision trees are added for retraining until the accuracy reaches more than 92%, and the lightweight random forest model training is completed.

[0062] (3) The embodiment can collect the cable joint temperature, environmental temperature, partial discharge pulse amplitude data and contact resistance data according to the previous steps, calculate the target joint temperature, partial discharge effective value and contact resistance change rate, and combine to form a group of real-time fusion features.

[0063] (4) The embodiment can input the real-time fusion features into the trained lightweight random forest model, and each decision tree outputs an independent result based on its own feature splitting rule (such as determining a fault if the partial discharge effective value is greater than 0.3V); the outputs of the 8 decision trees are counted, and if the number of no-fault determinations is greater than or equal to 5, the fault determination result is no fault, otherwise it is determined as a fault.

[0064] The embodiment can realize fault determination through a lightweight random forest model, which not only retains the high accuracy of random forest multi-feature fusion, but also reduces the calculation amount by simplifying the model parameters, and can be deployed without high-performance hardware, adapting to the on-site detection scene. The model trained based on historical labeled samples can learn the feature rules of different fault states, avoid misjudgment problems of single parameter threshold, and improve the fault determination accuracy. At the same time, the model outputs clear results of no fault or fault, providing a clear premise for subsequent fault type determination, and improving the efficiency and reliability of the overall detection process.

[0065] In an embodiment of the present application, the fault type determination result is determined based on the fusion features, historical cable joint temperature data, historical cable environmental temperature data, historical partial discharge pulse amplitude data of the cable joint and historical contact resistance data of the cable joint, comprising: The historical contact resistance data of the cable joint is divided based on the first time step to obtain a plurality of historical contact resistance window data, the historical contact resistance change rate of each historical contact resistance window data is calculated to obtain a historical contact resistance change rate sequence, and the historical contact resistance change trend feature is extracted based on the historical contact resistance change rate sequence; The historical joint temperature mean sequence is obtained based on the historical cable joint temperature data, and the historical joint temperature change trend feature is extracted based on the historical joint temperature mean sequence; The historical partial discharge effective value peak sequence is obtained based on historical partial discharge pulse amplitude data of the cable joint, and the historical partial discharge change trend feature is extracted based on the historical partial discharge effective value peak sequence; The historical temperature deviation value sequence is obtained based on historical cable joint temperature data and historical cable environment temperature data. The historical cable joint temperature data includes a plurality of historical cable joint temperatures obtained at a plurality of historical sampling time points, the historical cable environment temperature data includes a plurality of historical cable environment temperatures obtained at a plurality of historical sampling time points, and the historical temperature deviation value sequence includes a difference value between the historical cable joint temperature and the historical cable environment temperature corresponding to each historical sampling time point. The correlation coefficient of the historical contact resistance change rate sequence and the historical temperature deviation value sequence is calculated. Based on the fusion feature, the historical contact resistance change trend feature, the historical joint temperature change trend feature, the historical partial discharge change trend feature, and the correlation coefficient, a fault type determination result is output by a weighted K nearest neighbor algorithm.

[0066] In this embodiment, the first time step refers to a time interval for dividing historical contact resistance data, which can be determined in combination with a historical data collection frequency. The historical contact resistance window data refers to a historical contact resistance data segment divided according to the first time step. The historical contact resistance change rate sequence refers to a set of change rates corresponding to a plurality of historical contact resistance window data sorted by time. The historical contact resistance change trend feature refers to a trend index extracted from the historical contact resistance change rate sequence, which can include, for example, a trend slope, a change rate growth / drop amplitude, etc. The historical joint temperature mean value sequence refers to a set of historical cable joint temperature mean values calculated according to a time period. The historical joint temperature change trend feature refers to a trend index extracted from the historical joint temperature mean value sequence, which can include, for example, a temperature rise / drop rate, a mean value fluctuation range, etc.

[0067] The historical partial discharge effective value peak sequence refers to a set of historical partial discharge effective value maximum values calculated according to a time period. The historical partial discharge change trend feature refers to a trend index extracted from the historical partial discharge effective value peak sequence, which can include, for example, a peak value growth slope, a peak value fluctuation frequency, etc. The historical temperature deviation value sequence refers to a set of joint temperature and environment temperature difference values at each historical sampling time point, which can include, for example, a difference value maximum value / minimum value, a difference value change amplitude, etc. The correlation coefficient refers to an index representing the correlation degree between the historical contact resistance change rate sequence and the historical temperature deviation value sequence. The weighted K nearest neighbor algorithm refers to a K nearest neighbor classification algorithm that assigns different weights to neighboring samples.

[0068] In this embodiment, the consideration behind this embodiment is that the real-time features of different cable joint fault types may intersect, such as local overheating and poor contact both causing temperature rise, and it is difficult to accurately distinguish them by real-time fusion features alone; while the trend features and parameter correlation of historical data can reflect the essential differences of faults, such as poor contact which makes the contact resistance change rate and temperature deviation value strongly correlated.

[0069] This embodiment first extracts the historical trend features of contact resistance, temperature and partial discharge according to time steps, then calculates the correlation coefficient of resistance change rate and temperature deviation, and supplements the fault difference information from the trend + correlation dimension; finally, it fuses real-time fusion features and historical derived features through weighted K nearest neighbor algorithm, which not only retains real-time state information, but also uses historical rules to enhance classification ability, solving the problem that existing technologies cannot accurately distinguish fault types relying on real-time data alone.

[0070] For example, this embodiment can be applied to 10kV distribution network cable joint fault type determination, and the specific steps are as follows: (1) This embodiment can set the first time step to 1 day, divide the 1-day window from the nearly 30-day historical contact resistance data to obtain 30 historical contact resistance window data, each window containing 24 groups of resistance data on the same day; this embodiment can calculate the historical contact resistance change rate in each window, which is represented by the difference between the maximum resistance value in the window and the initial resistance value, and is sorted by date to form a historical contact resistance change rate sequence; this embodiment can linearly fit the sequence to extract the historical contact resistance change trend feature, such as a trend slope of 0.8% / day, indicating that the resistance change rate increases by 0.8% per day.

[0071] (2) This embodiment can calculate the daily average temperature from the nearly 30-day historical cable joint temperature data with a 1-day cycle, take the average of 24 groups of temperature data on the same day to form a historical joint temperature average sequence; linearly fit the sequence to extract the historical joint temperature change trend feature, such as a slope of 0.6℃ / day, indicating that the daily average temperature rises by 0.6℃ per day; at the same time, the nearly 30-day historical cable environmental temperature data is retrieved, 24 groups per day, the difference between the joint temperature and the environmental temperature at each historical sampling time point is calculated, and the historical temperature deviation value sequence is sorted by time.

[0072] (3) This embodiment can calculate the daily partial discharge effective value from the nearly 30-day cable joint historical partial discharge pulse amplitude data with a 1-day cycle, and take the maximum value of the daily effective value after denoising the pulse data on the same day to form a historical partial discharge effective value peak sequence; this embodiment can linearly fit the sequence to extract the historical partial discharge change trend feature, such as a slope of 0.1V / day, indicating that the peak value increases by 0.1V per day.

[0073] (4) The embodiment can calculate the correlation coefficient of the historical contact resistance change rate sequence and the historical temperature deviation value sequence based on the last 30 days, such as 0.8, indicating strong positive correlation. The embodiment can retrieve the current fusion feature, combine it with the historical contact resistance trend feature, the historical joint temperature trend feature, the historical partial discharge trend feature, and the correlation coefficient to form a classification input feature. The embodiment can load a pre-trained weighted K nearest neighbor algorithm, set K=5, assign a weight of 0.6 to the historical trend feature, and a weight of 0.4 to the correlation coefficient. The similarity between the input feature and the historical fault case feature in the algorithm is calculated, and the 5 most similar cases are selected. The case fault types are counted, such as 3 for poor contact, 1 for local overheating, and 1 for insulation aging. The embodiment can output the most frequent poor contact as the fault type determination result.

[0074] The embodiment extracts multi-dimensional historical trend features and correlation coefficients to make up for the defect that real-time fusion features alone cannot distinguish fault types, making fault type determination more in line with the nature of the fault. The historical trend features can reflect the gradual development process of the fault, and the correlation coefficient can strengthen the correlation between parameters. The combination of the two provides more abundant evidence for fault classification. At the same time, the weighted K nearest neighbor algorithm can highlight the influence of key features, improve classification accuracy, and avoid misjudgment caused by traditional algorithms treating all features equally. Ultimately, it improves the accuracy of distinguishing between local overheating, insulation aging, and poor contact, provides accurate guidance for maintenance, reduces the waste of time and cost caused by blind troubleshooting, and ensures the efficiency of power system operation and maintenance.

[0075] Corresponding to the cable joint fault detection method of the above embodiment, Figure 2 The structure block diagram of the cable joint fault detection device provided by an embodiment of the present application is shown. For ease of illustration, only the parts related to the embodiments of the present application are shown. For reference Figure 2 The cable joint fault detection device 20 includes a temperature correction module 21, an electrical data processing module 22, a fault determination module 23, and a fault type determination module 24.

[0076] The temperature correction module 21 is configured to obtain cable joint temperature and cable environment temperature data; determine the scene type based on the voltage level and laying method of the cable joint, and determine the temperature compensation coefficient based on the scene type; correct the cable joint temperature based on the temperature compensation coefficient and the cable environment temperature data to obtain the target joint temperature. The electrical data processing module 22 is configured to calculate the partial discharge effective value based on the partial discharge pulse amplitude data of the cable joint, and calculate the contact resistance change rate based on the contact resistance data of the cable joint. The fault determination module 23 is configured to construct a fusion feature based on the target joint temperature, the partial discharge effective value and the contact resistance change rate, and determine a fault determination result based on the fusion feature. The fault type determination module 24 is configured to, if the fault determination result is fault, determine a fault type determination result based on the fusion feature, historical cable joint temperature data, historical cable environment temperature data, historical partial discharge pulse amplitude data of the cable joint and historical contact resistance data of the cable joint.

[0077] In an embodiment of the present application, the scene type includes a high-voltage scene or a non-high-voltage scene, and a closed scene or an open scene; and the temperature correction module 21 is specifically configured to, when determining the scene type based on the voltage level of the cable joint and the laying mode, determine the scene type as the high-voltage scene or the non-high-voltage scene based on the voltage level of the cable joint, and determine the scene type as the closed scene or the open scene based on the laying mode of the cable joint.

[0078] In an embodiment of the present application, the temperature correction module 21 is specifically configured to, when determining the temperature compensation coefficient based on the scene type, if the scene type is the high-voltage scene and the closed scene, determine the temperature compensation coefficient as a first temperature compensation coefficient; if the scene type is the high-voltage scene and the open scene, determine the temperature compensation coefficient as a second temperature compensation coefficient; if the scene type is the medium-low-voltage scene and the closed scene, determine the temperature compensation coefficient as a third temperature compensation coefficient; if the scene type is the medium-low-voltage scene and the open scene, determine the temperature compensation coefficient as a fourth temperature compensation coefficient; the fourth temperature compensation coefficient is greater than the first temperature compensation coefficient and less than the third temperature compensation coefficient, and the first temperature compensation coefficient is greater than the second temperature compensation coefficient.

[0079] In an embodiment of the present application, the electrical data processing module 22 is specifically configured to, when calculating the discharge effective value based on the partial discharge pulse amplitude data of the cable joint, perform adaptive wavelet threshold denoising on the partial discharge pulse amplitude data of the cable joint to obtain a denoised partial discharge signal, and calculate the discharge effective value based on the denoised partial discharge signal.

[0080] In an embodiment of the present application, the electrical data processing module 22 is specifically configured to, when performing adaptive wavelet threshold denoising on the partial discharge pulse amplitude data of the cable joint to obtain a denoised partial discharge signal, perform wavelet decomposition on the partial discharge pulse amplitude data of the cable joint based on a target decomposition layer number to obtain a target approximation coefficient and a detail coefficient corresponding to each target decomposition layer; performing a first operation on the detail coefficients corresponding to each target decomposition layer to obtain noise-filtered detail coefficients of each layer; reconstructing the filtered detail coefficients of each layer and the target approximation coefficients through inverse wavelet transform to obtain a denoised partial discharge signal; The first operation includes: extracting background noise data in the detail coefficients corresponding to the target decomposition layer; calculating a standard deviation of the background noise data, and determining a noise threshold corresponding to the target decomposition layer based on the standard deviation; performing noise filtering on the detail coefficients corresponding to the target decomposition layer based on the noise threshold.

[0081] In an embodiment of the present application, the fault determination module 23 is specifically configured to determine the fault determination result based on the fusion feature through a light random forest model when determining the fault determination result based on the fusion feature. The fault determination result includes no fault or fault. The light random forest model is trained based on a plurality of historical fusion feature samples and a plurality of historical fusion feature samples respectively corresponding to historical fault labels.

[0082] In an embodiment of the present application, the fault type determination module 24 is specifically configured to: divide the historical contact resistance data of the cable joint based on a first time step to obtain a plurality of historical contact resistance window data, calculate a historical contact resistance change rate of each historical contact resistance window data to obtain a historical contact resistance change rate sequence, and extract a historical contact resistance change trend feature based on the historical contact resistance change rate sequence; obtain a historical joint temperature mean value sequence based on the historical cable joint temperature data, and extract a historical joint temperature change trend feature based on the historical joint temperature mean value sequence; obtain a historical partial discharge effective value peak sequence based on the historical partial discharge pulse amplitude data of the cable joint, and extract a historical partial discharge change trend feature based on the historical partial discharge effective value peak sequence; obtain a historical temperature deviation value sequence based on the historical cable joint temperature data and the historical cable environment temperature data. The historical cable joint temperature data includes a plurality of historical cable joint temperatures obtained at a plurality of historical sampling time points. The historical cable environment temperature data includes a plurality of historical cable environment temperatures obtained at a plurality of historical sampling time points. The historical temperature deviation value sequence includes a difference value between the historical cable joint temperature and the historical cable environment temperature corresponding to each historical sampling time point; calculate a correlation coefficient of the historical contact resistance change rate sequence and the historical temperature deviation value sequence; Based on fusion characteristics, historical contact resistance variation trends, historical joint temperature variation trends, historical partial discharge variation trends, and correlation coefficients, the fault type determination result is output through a weighted K-nearest neighbor algorithm.

[0083] See Figure 3 , Figure 3 This is a schematic block diagram of an electronic device provided according to an embodiment of this application. Figure 3 The electronic device 300 in this embodiment may include one or more processors 301, one or more input devices 302, one or more output devices 303, and one or more memories 304. The processors 301, input devices 302, output devices 303, and memories 304 communicate with each other via a communication bus 305. The memories 304 store computer programs, including program instructions. The processors 301 execute the program instructions stored in the memories 304. Specifically, the processors 301 are configured to invoke the program instructions to perform the functions of the modules in the aforementioned device embodiments, for example... Figure 2 The functions of the temperature correction module 21, electrical data processing module 22, fault determination module 23, and fault type determination module 24 are shown.

[0084] It should be understood that, in the embodiments of this application, the processor 301 may be a central processing unit (CPU), but it may also be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor.

[0085] Input device 302 may include a touchpad, a fingerprint sensor (for collecting the user's fingerprint information and fingerprint orientation information), a microphone, etc., and output device 303 may include a display (LCD, etc.), a speaker, etc.

[0086] The memory 304 may include read-only memory and random access memory, and provides instructions and data to the processor 301. A portion of the memory 304 may also include non-volatile random access memory. For example, the memory 304 may also store information about cable connectors.

[0087] In specific implementations, the processor 301, the input device 302, and the output device 303 described in the embodiments of the present application can perform the implementation manners described in the embodiments of the cable joint fault detection method provided by the embodiments of the present application, and can also perform the implementation manners of the electronic device 300 described in the embodiments of the present application, which will not be described here.

[0088] In another embodiment of the present application, a computer readable storage medium is provided, which stores a computer program. The computer program includes program instructions, which are executed by a processor to implement all or part of the processes of the above-mentioned embodiments. The computer program can also be used to instruct related hardware to complete the implementation. The computer program can be stored in a computer readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned method embodiments can be implemented. The computer program includes computer program code, which can be in the form of source code, object code, executable file, or some intermediate form. The computer readable medium can include any entity or device capable of carrying the computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc.

[0089] The computer readable storage medium can be an internal storage unit of the electronic device of any of the above-mentioned embodiments, such as a hard disk or a memory of the electronic device. The computer readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. Further, the computer readable storage medium can include both the internal storage unit and the external storage device of the electronic device. The computer readable storage medium is used to store the computer program and other programs and data required by the electronic device. The computer readable storage medium can also be used to temporarily store data that has been output or will be output.

[0090] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the electronic device and the unit described above can refer to the corresponding process in the foregoing method embodiment, and will not be repeated here.

[0091] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the electronic device and the unit described above can refer to the corresponding process in the foregoing method embodiment, and will not be repeated here.

[0092] In several embodiments provided in the present application, it should be understood that the disclosed electronic device and method can be implemented in other ways. For example, the above-described device embodiments are only schematic, for example, the division of the modules / units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of modules, units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the shown or discussed each other can be indirect coupling or communication connection through some interfaces or modules / units, and can also be electrical, mechanical or other form of connection.

[0093] The modules / units described as separate components can or can not be physically separated, and the components shown as modules / units can or can not be physical modules / units, that is, can be located in one place, or can be distributed to a plurality of network modules / units. Part or all of the modules / units can be selected according to actual needs to achieve the purpose of the embodiments of the present application.

[0094] In addition, each functional module / unit in each embodiment of the present application can be integrated in one processing module / unit, or each module / unit can exist physically, or two or more modules / units can be integrated in one module / unit. The integrated module / unit can be realized in the form of hardware or in the form of software functional module / unit.

[0095] The above merely provides the specific implementation of the present application, but the protection scope of the present application is not limited to this. Any person skilled in the art can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present application, and these modifications or replacements should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for detecting cable joint faults, characterized in that, include: Acquire cable joint temperature and cable ambient temperature data; The scenario type is determined based on the voltage level and laying method of the cable joint, and the temperature compensation coefficient is determined based on the scenario type. The cable joint temperature is corrected based on the temperature compensation coefficient and the cable ambient temperature data to obtain the target joint temperature; The effective value of partial discharge is calculated based on the partial discharge pulse amplitude data of the cable joint, and the contact resistance change rate is calculated based on the contact resistance data of the cable joint. A fusion feature is constructed based on the target joint temperature, the effective value of partial discharge, and the rate of change of contact resistance; and a fault determination result is determined based on the fusion feature. If the fault determination result is a fault, then the fault type determination result is determined based on the fusion characteristics, historical cable joint temperature data, historical cable ambient temperature data, historical partial discharge pulse amplitude data of the cable joint, and historical contact resistance data of the cable joint.

2. The cable joint fault detection method as described in claim 1, characterized in that, The scenario types include high-voltage scenarios or non-high-voltage scenarios, as well as closed scenarios or open scenarios; The determination of scenario type based on the voltage level and laying method of cable joints includes: The scenario type is determined as either a high-voltage scenario or a non-high-voltage scenario based on the voltage level of the cable connector. The scenario type is determined as either a closed scenario or an open scenario based on the cable joint laying method.

3. The cable joint fault detection method as described in claim 1, characterized in that, Determining the temperature compensation coefficient based on the scene type includes: If the scenario type is a high-pressure scenario or a closed scenario, then the temperature compensation coefficient is determined to be the first temperature compensation coefficient. If the scenario type is a high-pressure scenario or an open scenario, then the temperature compensation coefficient is determined to be the second temperature compensation coefficient. If the scenario type is a medium-low pressure scenario or a closed scenario, then the temperature compensation coefficient is determined to be the third temperature compensation coefficient. If the scenario type is a medium-low pressure scenario or an open scenario, then the temperature compensation coefficient is determined to be the fourth temperature compensation coefficient. The fourth temperature compensation coefficient is greater than the first temperature compensation coefficient and less than the third temperature compensation coefficient, and the first temperature compensation coefficient is greater than the second temperature compensation coefficient.

4. The cable joint fault detection method as described in claim 1, characterized in that, The calculation of the effective discharge value based on the partial discharge pulse amplitude data of the cable joint includes: Adaptive wavelet threshold denoising is performed on the partial discharge pulse amplitude data of the cable joint to obtain the denoised partial discharge signal; The effective value of the discharge is calculated based on the denoised partial discharge signal.

5. The cable joint fault detection method as described in claim 4, characterized in that, The adaptive wavelet threshold denoising of the partial discharge pulse amplitude data of the cable joint to obtain the denoised partial discharge signal includes: Wavelet decomposition is performed on the partial discharge pulse amplitude data of the cable joint based on the target decomposition level to obtain the target approximation coefficient and the detail coefficients corresponding to each target decomposition level. Perform the first operation on the detail coefficients corresponding to each target decomposition layer to obtain the detail coefficients of each layer after noise filtering; The filtered detail coefficients and target approximation coefficients are reconstructed by inverse wavelet transform to obtain the denoised partial discharge signal. The first operation includes: Extract the background noise data from the detail coefficients corresponding to the target decomposition layer; Calculate the standard deviation of the background noise data, and determine the noise threshold corresponding to the target decomposition layer based on the standard deviation; Noise filtering is performed on the detail coefficients corresponding to the target decomposition layer based on the noise threshold.

6. The cable joint fault detection method as described in claim 1, characterized in that, The process of determining the fault determination result based on the fusion features includes: Based on the fusion features, a lightweight random forest model is used to determine the fault determination result; the fault determination result includes no fault or fault. The lightweight random forest model is trained based on multiple historical fusion feature samples and the historical fault labels corresponding to the multiple historical fusion feature samples.

7. The cable joint fault detection method as described in claim 1, characterized in that, The fault type determination result based on the fusion features, historical cable joint temperature data, historical cable ambient temperature data, historical partial discharge pulse amplitude data of cable joints, and historical contact resistance data of cable joints includes: Based on the first time step, the historical contact resistance data of the cable joint is divided into multiple historical contact resistance window data. The historical contact resistance change rate of each historical contact resistance window data is calculated to obtain a historical contact resistance change rate sequence. Based on the historical contact resistance change rate sequence, the historical contact resistance change trend characteristics are extracted. A historical joint temperature mean sequence is obtained based on historical cable joint temperature data, and the historical joint temperature change trend characteristics are extracted based on the historical joint temperature mean sequence. Based on the historical partial discharge pulse amplitude data of the cable joint, a historical partial discharge effective value peak sequence is obtained, and the historical partial discharge change trend characteristics are extracted based on the historical partial discharge effective value peak sequence. A historical temperature deviation value sequence is obtained based on historical cable joint temperature data and historical cable ambient temperature data; the historical cable joint temperature data includes multiple historical cable joint temperatures obtained at multiple historical sampling time points, the historical cable ambient temperature data includes multiple historical cable ambient temperatures obtained at multiple historical sampling time points, and the historical temperature deviation value sequence includes the difference between the historical cable joint temperature and the historical cable ambient temperature corresponding to each historical sampling time point. Calculate the correlation coefficient between the historical contact resistance change rate sequence and the historical temperature deviation value sequence; Based on the fusion characteristics, the historical contact resistance variation trend characteristics, the historical joint temperature variation trend characteristics, the historical partial discharge variation trend characteristics, and the correlation coefficient, the fault type determination result is output through the weighted K-nearest neighbor algorithm.

8. A cable joint fault detection device, characterized in that, include: Temperature correction module is used to acquire cable joint temperature and cable ambient temperature data; The scenario type is determined based on the voltage level and laying method of the cable joint, and the temperature compensation coefficient is determined based on the scenario type. The temperature of the cable joint is corrected based on the temperature compensation coefficient and the cable ambient temperature data to obtain the target joint temperature. The electrical data processing module is used to calculate the effective value of partial discharge based on the partial discharge pulse amplitude data of the cable joint, and to calculate the contact resistance change rate based on the contact resistance data of the cable joint. The fault determination module is used to construct a fusion feature based on the target joint temperature, the effective value of partial discharge and the contact resistance change rate, and to determine the fault determination result based on the fusion feature; The fault type determination module is used to determine the fault type determination result based on the fusion features, historical cable joint temperature data, historical cable ambient temperature data, historical partial discharge pulse amplitude data of the cable joint, and historical contact resistance data of the cable joint if the fault determination result is a fault.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1 to 7.

Images