Power cable partial discharge diagnosis and analysis method based on slope behavior analysis

Abstract

The invention discloses a power cable partial discharge diagnosis and analysis method based on slope behavior analysis, and relates to the field of power system detection and calculation, and the method comprises the steps: collecting power data through a sensor, and carrying out the preprocessing; identifying a discharge event from the signal sequence using a ramp behavior analysis algorithm; based on the phase distribution, the polarity distribution and the time interval parameter under each voltage level, performing visual display and analysis; and fitting characteristic change rules under different voltage levels by using a smooth spline function. Through a slope behavior analysis algorithm, noise and real partial discharge signals can be effectively distinguished, and the problem that the signals are difficult to identify in a complex environment in a traditional method is solved. The algorithm is combined with dynamic threshold adjustment, and partial discharge characteristics under different voltage levels can be accurately captured.

Description

Technical Field

[0001] The present invention relates to the field of power system detection and calculation, and in particular to a power cable partial discharge diagnosis and analysis method based on slope behavior analysis. Background Art

[0002] Power cables are a vital component of the power system, and their operational status is crucial to the stable operation of the power grid. However, cables may experience abnormal conditions during operation. Partial discharge (PD) is a common fault phenomenon in cables and one of the main causes of cable performance degradation. Partial discharge (PD) is typically caused by factors such as internal defects in the insulation material, external contamination, mechanical stress, or aging. It manifests as localized electrical discharges due to increased electric field strength. This phenomenon not only accelerates insulation aging but can also lead to equipment failures and even system short circuits, seriously impacting power system reliability. Therefore, diagnosing and analyzing PD patterns and trends in operating cables to promptly detect potential insulation issues is crucial for ensuring stable power system operation.

[0003] Although existing partial discharge detection technology can diagnose and analyze partial discharge phenomena to a certain extent, it still has limitations: mainly reflected in the following aspects:

[0004] Signal-to-noise ratio problem. Traditional cable partial discharge detection technology faces the severe challenge of difficulty in separating noise from effective signals. In the on-site detection environment, various interference signals from multiple sources, such as electromagnetic interference, communication equipment radiation, switching noise and thermal noise, are similar to or overlap with the spectral characteristics of the PD signal, making it difficult to determine the filtering boundary. Especially in low signal-to-noise ratio environments, such as heavy industrial areas, around substations or during peak load periods of power systems, the PD signal energy is often submerged in the background noise. Traditional methods such as frequency domain filtering, wavelet transform and empirical mode decomposition have limited ability to suppress non-stationary noise, and the parameter optimization process requires a lot of manual intervention, which reduces the robustness of the detection system. The high sensitivity of the detection results to noise leads to a significant increase in the false detection rate and missed detection rate, which seriously affects the reliability of the cable insulation status assessment and brings greater uncertainty to preventive maintenance decisions.

[0005] Insufficient trend analysis. PD trend analysis faces multiple challenges: First, traditional analysis methods struggle to effectively distinguish random fluctuations in discharge activity from actual degradation trends. Second, phase-amplitude-number maps overlap between different types of PDs, and as non-stationary random signals, the time-frequency characteristics of PD signals are difficult to accurately describe using simple parameters. Furthermore, PD activity exhibits complex periodic variations influenced by external factors such as load cycles, and existing methods often fail to effectively separate these interfering factors. Furthermore, research on PD characteristics under voltage gradient variations is severely insufficient. Existing studies have mostly employed static testing methods, ignoring the modulation effect of voltage fluctuations on PD activity during actual operation. This results in deviations between predicted PD behavior under multiple voltage conditions and actual conditions. These issues collectively restrict the accuracy and reliability of PD trend analysis.

[0006] It's difficult to strike a balance between real-time performance and computational efficiency. Partial discharge detection systems face a severe contradiction between real-time monitoring and computational efficiency, hindering their practical application. Partial discharge detection typically requires a high sampling rate of over 10 MHz, generating massive amounts of data in the megabytes per second, placing enormous pressure on real-time processing. Advanced signal processing algorithms such as continuous wavelet transforms and deep neural networks offer high detection accuracy, but their higher computational complexity leads to processing delays, making it difficult to meet the requirements of real-time monitoring and early warning. Field monitoring equipment is often limited by power consumption and size, making it impossible to accommodate high-performance computing platforms. However, long-term continuous monitoring requires algorithm stability to avoid memory leaks or the accumulation of computing resources. Large-scale equipment requires multi-point simultaneous monitoring to improve detection reliability, which increases the complexity of data fusion and parallel computing. These factors collectively restrict the application of high-precision algorithms in practical monitoring systems, creating a technical bottleneck that makes it difficult to strike a balance between detection accuracy and real-time performance.

[0007] Analysis of these deficiencies reveals numerous limitations in existing partial discharge analysis technologies, including difficulty identifying signals in low signal-to-noise ratio environments, inadequate feature extraction and trend analysis, and the difficulty balancing high-precision algorithms with real-time performance. Consequently, a new, more efficient, integrated solution is urgently needed that can monitor and accurately evaluate partial discharge characteristics in complex environments and under varying voltage conditions in real time. Summary of the Invention

[0008] To solve the above technical problems, the present invention provides the following technical solutions: a method for diagnosing and analyzing partial discharge of power cables based on slope behavior analysis, which comprises the following steps: using sensors to collect power data and perform preprocessing;

[0009] Based on the preprocessing results, a ramp behavior analysis algorithm is used to identify discharge events from the signal sequence;

[0010] Based on the statistics of discharge events, the phase distribution, polarity distribution and time interval parameters at each voltage level are visualized and analyzed;

[0011] Use smoothing spline function to fit the characteristic variation law under different voltage levels;

[0012] The slope behavior analysis algorithm based on the preprocessing results includes setting a dynamic threshold to distinguish significant events from stable events, detecting the amplitude difference and slope angle of significant events to quantify the steepness;

[0013] The visualization and analysis include correlating the number of significant events with the PD peak distribution through a dual Y-axis graph, establishing a frequency correlation between voltage, phase angle, and PD amplitude using a three-dimensional histogram, and analyzing the variation of PD pulse time difference with voltage level based on a two-dimensional bar graph.

[0014] As a preferred embodiment of the method for diagnosing and analyzing partial discharge in power cables based on ramp behavior analysis described in the present invention, the method comprises: collecting power data using sensors, sequentially applying 50 Hz AC voltages of different voltage levels to XLPE cable samples through a programmable AC power supply, measuring the output of the HFCT using a digital storage oscilloscope and recording the data, recording 32 cycles of data at each voltage level, including actual partial discharge events and noise generated by the measurement;

[0015] While recording the signal, the voltage waveform is also recorded synchronously, and the collected signal data is stored in a time series format, including timestamps, partial discharge signal amplitudes, and corresponding voltage phase information.

[0016] As a preferred solution of the power cable partial discharge diagnostic analysis method based on slope behavior analysis described in the present invention, the preprocessing is to use wavelet transform to perform multi-resolution analysis on the signal to remove high-frequency noise, and perform normalization processing to map the original amplitude of the measurement data to the [-1,1] interval.

[0017] As a preferred embodiment of the method for diagnosing and analyzing partial discharge of power cables based on slope behavior analysis described in the present invention, the method comprises: using a slope behavior analysis algorithm based on the preprocessing results to identify discharge events from a signal sequence by setting a threshold, traversing the normalized signal data, and distinguishing significant events from stationary events in the time series data set;

[0018] Detect the starting and ending points where the signal amplitude exceeds the threshold. If it exceeds the threshold, the time series data is classified as a significant event, otherwise it is classified as a stationary event.

[0019] The significant time includes significant rising events and significant falling events;

[0020] The changing moments of significant events include the start time and end time of the event. The amplitude difference corresponding to the start time and end time and the inclination angle of the slope formed by the difference relative to the horizontal direction are calculated to quantify the steepness of the change and output the slope behavior analysis diagram.

[0021] As a preferred solution of the power cable partial discharge diagnosis and analysis method based on slope behavior analysis described in the present invention, the threshold is iteratively analyzed by manual peak noise amplitude, and the threshold is dynamically adjusted according to the signal-to-noise ratio and partial discharge characteristics of the experimental data.

[0022] As a preferred solution of the power cable partial discharge diagnosis and analysis method based on slope behavior analysis described in the present invention, wherein: the phase distribution, polarity distribution and time interval parameters at each voltage level based on the discharge event statistics are visualized and analyzed, including:

[0023] Based on the slope behavior analysis diagram, the characteristics of significant events are reflected and extracted for analysis;

[0024] Statistical analysis was performed on significant events at each voltage level. The number of significant events within each voltage cycle, the distribution of PD peak values of significant events, and the distribution frequency of PD peaks at different phase angles within significant events were calculated. The main phase intervals of partial discharge events and the time difference changes of each PD pulse measured at different voltages were determined.

[0025] Based on the number of significant events and the distribution of PD peaks of significant events within each voltage cycle, a dual Y-axis graph is drawn at different voltage levels. The X-axis represents the AC voltage cycle, the left Y-axis represents the number of various PD peaks, and the right Y-axis represents the number of significant events. PD peak data and significant event data are displayed in different ways to analyze the changing trends between PD peak characteristics and the number of significant events in each cycle.

[0026] By comparing data at different voltage levels, the relationship between voltage level and the number of significant events was studied, and the characteristics of positive and negative polarity PD peaks at all voltage levels were analyzed to reveal the influence of voltage level on PD characteristics.

[0027] Based on the distribution frequency of PD peaks at different phase angles in significant events, a three-dimensional histogram is plotted to visualize partial discharge data at different voltage levels. The phase angle range and normalized amplitude range are divided into several intervals, where the X-axis represents the phase angle, the Y-axis represents the normalized amplitude, and the Z-axis represents the occurrence frequency. The height of each cube represents the number of discharge events at a specific phase angle and amplitude combination.

[0028] By studying the frequency distribution of PD events within a specific voltage phase angle range, the correlation between voltage level, phase angle and PD amplitude is established.

[0029] Based on the time difference change of each PD pulse measured at different voltages, a two-dimensional bar graph is used to plot the time difference change of each PD pulse measured at different voltages, where the X-axis represents the AC voltage period and the Y-axis represents the time difference of each PD pulse;

[0030] By comparing data at different voltage levels, the changing trend of the overall range of the time difference of each PD pulse at different voltage levels is studied, and the correlation between voltage level and PD pulse time characteristics is established.

[0031] As a preferred solution of the power cable partial discharge diagnosis and analysis method based on slope behavior analysis described in the present invention, wherein: the use of smoothing spline function to fit the characteristic change law under different voltage levels includes:

[0032] According to the time difference and amplitude change of each PD pulse measured at different voltages, linear polynomial fitting and smoothing spline fitting were performed respectively. The goodness of fit value and root mean square error of the fitting curve were calculated to evaluate the fitting effect. The goodness of fit is proportional to the explanatory power of the data, and the root mean square error is proportional to the deviation of the actual data.

[0033] A computer device includes a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the power cable partial discharge diagnosis and analysis method based on slope behavior analysis are implemented.

[0034] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the power cable partial discharge diagnosis and analysis method based on slope behavior analysis as described above.

[0035] Beneficial effects of the present invention: Through a ramp behavior analysis algorithm, the present invention can effectively distinguish between noise and true partial discharge signals, solving the problem of traditional methods having difficulty identifying signals in complex environments. Combined with dynamic threshold adjustment, the algorithm can accurately capture the characteristics of partial discharge at different voltage levels. The present invention not only focuses on the instantaneous characteristics of partial discharge but also provides the behavioral patterns and changing trends of partial discharge at different voltage levels through statistical analysis and trend extraction. Through characteristic statistical analysis and polynomial fitting, the number of partial discharge events, peak distribution, phase angle distribution, and time difference changes can be clearly displayed, providing a more comprehensive basis for cable status assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0037] Figure 1 An overall flow chart of a method for diagnosing and analyzing partial discharge of power cables based on slope behavior analysis provided by the first embodiment of the present invention;

[0038] Figure 2 A slope behavior analysis diagram of a power cable partial discharge diagnosis and analysis method based on slope behavior analysis provided by the first embodiment of the present invention;

[0039] Figure 3 A graph showing discharge peak values and total number of events within 10 cycles of a power cable partial discharge diagnostic analysis method based on slope behavior analysis provided by the first embodiment of the present invention;

[0040] Figure 4 A graph showing the frequency of occurrence of PD peak amplitudes at different phase angles in a power cable partial discharge diagnosis and analysis method based on slope behavior analysis provided by the first embodiment of the present invention;

[0041] Figure 5 This is a PD pulse time difference diagram in a power cable partial discharge diagnosis and analysis method based on slope behavior analysis provided by the first embodiment of the present invention.

[0042] Figure 6 This is a fitting curve diagram of a power cable partial discharge diagnosis and analysis method based on slope behavior analysis provided in the third embodiment of the present invention. DETAILED DESCRIPTION

[0043] To make the above-mentioned objects, features, and advantages of the present invention more clearly understood, the following detailed description of the specific embodiments of the present invention is given in conjunction with the accompanying drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in this field without creative work should fall within the scope of protection of the present invention.

[0044] Example 1, reference Figures 1 to 5 According to one embodiment of the present invention, a method for diagnosing and analyzing partial discharge of power cables based on slope behavior analysis is provided, comprising:

[0045] Traditional cable partial discharge analysis technology faces the severe challenge of difficulty in separating noise from valid signals. Therefore, a method is needed to effectively distinguish noise from true partial discharge signals.

[0046] Existing feature extraction methods mostly rely on statistical parameters (such as mean, standard deviation, and skewness). These parameters are not stable across different voltage levels and device states, making it difficult to accurately reflect the essential characteristics of partial discharge. Existing detection methods mostly focus on the instantaneous characteristics of partial discharge, but lack analysis of the behavioral patterns and trends of partial discharge at different voltage levels, making it difficult to provide comprehensive analysis results. Therefore, a method is needed to accurately extract and classify partial discharge features while providing a comprehensive trend analysis.

[0047] Traditional methods rely on complex data analysis and high computing resources, and have poor real-time processing capabilities for large-scale data, which affects application efficiency. The present invention implements the ramp behavior algorithm in the form of parallel and recursive programming, thereby improving data processing efficiency.

[0048] Therefore, the present invention proposes a power cable partial discharge diagnosis and analysis method based on slope behavior analysis.

[0049] Use sensors to collect power data and perform pre-processing.

[0050] Based on the preprocessing results, a ramp behavior analysis algorithm is used to identify discharge events from the signal sequence.

[0051] Based on the discharge event statistics, the phase distribution, polarity distribution and time interval parameters at each voltage level are visualized and analyzed.

[0052] The smoothing spline function is used to fit the characteristic variation law under different voltage levels.

[0053] A programmable AC power supply sequentially applied 50Hz AC voltages at different voltage levels (e.g., 6.4kV, 7.4kV, 9.4kV, and 11.3kV) to XLPE cable samples. A digital storage oscilloscope was used to measure and record the HFCT output, recording 32 cycles of data at each voltage level. This dataset included actual partial discharge events and measurement noise. Simultaneously with the signal recording, the voltage waveform was also recorded to facilitate subsequent analysis of the relationship between partial discharge events and voltage phase.

[0054] The HFCT has a bandwidth of 0.5-80 MHz and a transmission ratio of 1:10, while utilizing a coupling capacitor to stabilize the voltage and provide charging during PD events.

[0055] Store the collected signal data in a time series format, including timestamps, partial discharge signal amplitudes, and corresponding voltage phase information. Ensure a consistent data format for subsequent processing and analysis.

[0056] Depending on the background noise level, recorded PD data may need to be denoised in some cases to remove high-frequency noise and interfering signals. The filter design should be optimized based on the frequency characteristics of the PD signal to preserve the key signal features. Common filtering methods include fast Fourier transform (FFT), short-time Fourier transform (STFT), low-pass and notch filtering, and wavelet-based filtering. For example, wavelet transforms can be used to perform multi-resolution analysis of the signal, removing high-frequency noise while preserving the details of the PD pulse.

[0057] The raw amplitude of the measured data is mapped to the interval [-1, 1], for example, 30 mV corresponds to 1, eliminating dimensionality. The normalized signal is then split into multiple time windows, each containing a complete voltage cycle. This facilitates the subsequent analysis of partial discharge events within each cycle.

[0058] A threshold is set to distinguish significant events from stationary events in the time series dataset. The threshold is determined iteratively by manually analyzing the peak noise amplitude. The threshold is dynamically adjusted based on the signal-to-noise ratio and partial discharge characteristics of the experimental data. At low voltage levels, the threshold can be set low to capture weak partial discharge signals; at high voltage levels, the threshold can be appropriately increased to filter out noise.

[0059] Traverse the normalized signal data and detect the starting and ending points where the signal amplitude exceeds the threshold. If the threshold is exceeded, the time series data is classified as a significant event; otherwise, it is classified as a stationary event. At the same time, significant events can be further classified according to the trend of the data signal amplitude change into:

[0060] (1) Significant rising event: The signal amplitude shows a clear increasing trend.

[0061] (2) Significant decrease event: The signal amplitude shows a significant decreasing trend.

[0062] The moment of change for a significant event is recorded as the event's start time (t1) and end time (t2), with Δt being the time interval. The initial PD size W1 at t1 and the final size W2 at t2 are recorded, and the difference between the two, ΔWs, is calculated. The system also calculates the angle θ of ΔWs relative to the horizontal to quantitatively describe the steepness of the change.

[0063] Slope behavior analysis algorithm specific applications such as Figure 2As shown in the figure: First, significant events and stable events are distinguished by setting an appropriate threshold, and the threshold is dynamically adjusted according to different voltage levels; then, the normalized signal data is traversed to detect points exceeding the threshold and the trend of data signal amplitude change to further classify the time series data; finally, the significant event features are extracted, including the start time t1 and end time t2 of the event, the time interval Δt, the initial size W1 and final size W2, the amplitude change ΔW, and the change slope angle θ. Figure 2 The top panel shows the original signal curve, with the threshold line (dashed blue line), slope angle θ, amplitude change ΔW, and time interval Δt marked. The bottom panel breaks down the same curve by event type: blue indicates a stationary event (the signal does not exceed the threshold), green indicates a significant decrease in signal amplitude (a significant decrease in signal amplitude), and orange indicates a significant increase in signal amplitude (a significant increase in signal amplitude).

[0064] The algorithm described above produces two output data sets: significant events, which capture the occurrence of PD peaks, and stationary events, which account for noise. Therefore, for further evaluation, stationary events are not considered. Significant events, however, encompass not only PD peaks but also the oscillations associated with them. Oscillations are a byproduct of the measurement process, and the initial PD pulse and its waveform should be of primary interest. The first PD peak that occurs during an event represents the actual partial discharge.

[0065] Statistical analysis was performed on significant events at each voltage level to calculate the following parameters:

[0066] Number of significant events: Counts the number of significant events in each voltage cycle.

[0067] Peak distribution: This function measures the distribution of statistically significant PD peaks, including the number of PD peaks, the number of positive PD peaks, and the number of negative PD peaks. The polarity of the voltage cycle affects the polarity of the PD peak. Positive PD peaks are observed in positive voltage cycles, and vice versa.

[0068] Phase angle distribution: Calculate the distribution frequency of PD peaks at different phase angles in statistically significant events to determine the main phase intervals of partial discharge events.

[0069] Time difference variation: Statistics on the time difference variation of each PD pulse measured at different voltages.

[0070] In order to fully understand the occurrence mechanism and evolution law of partial discharge, it is necessary to systematically analyze and extract its characteristic trends from multiple dimensions.

[0071] A. A dual Y-axis chart structure is used for data visualization. The X-axis represents 10 AC voltage cycles, the left Y-axis is used to display the number of various PD peaks, and the right Y-axis is used to display the number of significant events. Data is displayed using a combination of stacked bar charts and line charts. PD peak data is displayed in the form of a stacked bar chart (light green represents the sum of PD peaks, blue represents negative PD peaks, and red represents positive PD peaks), and significant event data is displayed in the form of a black line chart. Figure 3 shown.

[0072] First, a systematic analysis of the data at each voltage level was conducted. By statistically analyzing the variation range of the total number of significant events and the variation range of the total number of PD peaks, a correlation between the number of PD peaks and the number of significant events was established. Furthermore, the distribution characteristics of positive and negative PD peaks within each voltage cycle were specifically examined. Furthermore, by comparing data at different voltage levels, the relationship between voltage level and the number of significant events was studied, and the characteristics of positive and negative PD peaks at all voltage levels were analyzed, thereby revealing the influence of voltage level on PD characteristics.

[0073] Since at low voltage levels, there are no positive peaks, only the negative voltage cycles are analyzed further.

[0074] The PD test data is visualized using a 3D histogram, which divides the phase angle range (150-300 degrees) and the normalized amplitude range (-3 to 0) into several bins. The horizontal x-axis represents the phase angle (degrees), the horizontal y-axis represents the normalized PD amplitude, and the vertical z-axis represents the frequency of occurrence. The height of each cube represents the number of discharge events under a specific phase angle and amplitude combination. Figure 4 shown.

[0075] First, a systematic analysis of the data at each voltage level was conducted to analyze the peak frequency and distribution of PD pulses within each normalized PD amplitude range. This systematic analysis of data at different voltage levels focused on the distribution of PD event frequencies within a specific voltage phase angle range, thereby establishing a correlation between voltage level, phase angle, and PD amplitude.

[0076] Use a two-dimensional bar graph to plot the time difference (Δt) of each PD pulse measured at different voltages. The X-axis represents the AC voltage period, and the Y-axis represents the time difference of each PD pulse. Figure 5 As shown:

[0077] First, a systematic analysis of the data at each voltage level was conducted, with the variation range of Δt recorded and analyzed in detail, and the distribution characteristics of the time difference of PD pulses were statistically analyzed. Furthermore, by comparing data at different voltage levels, the variation trends of the overall Δt range at different voltage levels were studied. The analysis focused on the concentrated Δt intervals of most PD pulses, revealing the influence of voltage level on PD pulse timing characteristics. Through this systematic analysis, a correlation between voltage level and PD pulse timing characteristics was established.

[0078] At the same time, in order to determine the changing trend of PD peak value under different voltage levels, linear polynomial fitting and smoothing spline fitting were performed on the variables Δt and ΔW, respectively. Figure 6 As shown, the mathematical equations for the two fits are as follows:

[0079] Δw(Δt)=p1((Δt) 2 +p2Δt+p3

[0080]

[0081] The smoothing parameter λ controls the smoothness of the fitted curve. A larger λ results in a smoother curve but may reduce fitting accuracy; a smaller λ results in a closer fit to the data points but may lead to overfitting. f(Δt) is a smoothing function used to fit the relationship between ΔW and Δt. This optimization problem is solved using numerical optimization methods (such as the least squares method) to obtain the smoothing spline function f(Δt).

[0082] Calculate the goodness of fit of the fitting curve (R 2 ) value and root mean square error (RMSE) to evaluate the fitting effect. 2 A higher value indicates that the fitting method has a better ability to explain the data. A lower RMSE value indicates that the deviation between the fitted curve and the actual data is smaller, and the fitting effect is better.

[0083]

[0084] Among them, S res is the residual sum of squares, SS tot is the total sum of squares, ΔW i is the actual observed value. is the model prediction value. n is the total number of data points. is the average of the actual observations.

[0085] Furthermore, the present invention utilizes an innovative Ramp Behavior Analysis (RBA) algorithm to effectively distinguish between noise and true partial discharge signals, resolving the difficulty traditional methods face in signal identification in complex environments. This algorithm, combined with dynamic threshold adjustment, accurately captures partial discharge characteristics at different voltage levels. This invention not only focuses on the instantaneous characteristics of partial discharge but also, through statistical analysis and trend extraction, provides behavioral patterns and trends of partial discharge at different voltage levels. Through feature statistical analysis and polynomial fitting, it clearly displays the number of partial discharge events, peak distribution, phase angle distribution, and time difference variations, providing a more comprehensive basis for cable condition assessment. This invention overcomes the limitations of existing technologies in insufficient behavioral pattern analysis at multiple voltage levels. By dynamically monitoring and analyzing partial discharge characteristics at different voltage levels, it can truly reflect the dynamic characteristics of the cable under actual operating conditions, improving the accuracy of cable condition assessment and the practicality of the prediction model. The present invention utilizes parallel and recursive programming to implement the ramp behavior algorithm, significantly improving data processing efficiency. Compared to traditional methods, this algorithm achieves high accuracy while meeting the requirements of real-time monitoring, resolving the technical bottleneck of balancing high-precision algorithms with real-time performance.

[0086] Example 2 is an embodiment of the present invention.

[0087] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0088] The logic and / or steps represented in the flowcharts or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing the logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (e.g., a computer-based system, a system including a processor, or other system that can fetch and execute instructions from an instruction execution system, apparatus, or device). For purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0089] More specific examples (a non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic devices), a portable computer disk cartridge (magnetic devices), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), a fiber optic device, and a portable compact disc read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering, or processing in another suitable manner as necessary, and then stored in a computer memory.

[0090] It should be understood that various parts of the present invention can be implemented using hardware, software, firmware, or a combination thereof. In the above-described embodiments, multiple steps or methods can be implemented using software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented using hardware, as in another embodiment, any one of the following technologies known in the art or a combination thereof can be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application-specific integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0091] Example 3, reference Figure 6 This is the third embodiment of the present invention. In order to verify the beneficial effects of the present invention, economic benefit calculation and simulation experiments are used to conduct scientific demonstration. This embodiment experiments on the existing traditional method and the method of this embodiment.

[0092] In order to determine the changing trend of PD peak under different voltage levels, linear polynomial fitting and smoothing spline fitting were performed on the variables Δt and ΔW, respectively. Figure 6 As shown, the mathematical equations for the two fits are as follows:

[0093] Δw(Δt)=p1((Δt) 2 +p2Δt+p3

[0094]

[0095] The smoothing parameter λ controls the smoothness of the fitted curve. A larger λ results in a smoother curve but may reduce fitting accuracy; a smaller λ results in a closer fit to the data points but may lead to overfitting. f(Δt) is a smoothing function used to fit the relationship between ΔW and Δt. This optimization problem is solved using numerical optimization methods (such as the least squares method) to obtain the smoothing spline function f(Δt).

[0096] Calculate the goodness of fit of the fitting curve (R 2 ) value and root mean square error (RMSE) to evaluate the fitting effect. 2 A higher value indicates that the fitting method has a better ability to explain the data. A lower RMSE value indicates that the deviation between the fitted curve and the actual data is smaller, and the fitting effect is better.

[0097]

[0098]

[0099]

[0100]

[0101] Among them, S res is the residual sum of squares, SS tot is the total sum of squares, ΔW i is the actual observed value. is the model prediction value. n is the total number of data points. is the average of the actual observations.

[0102] Through calculation, we can know that the RMSE value increases with the increase of voltage level, R 2 The value decreases as the voltage level increases.

[0103] At the same time, the smoothing spline produced a better fit compared to the quadratic polynomial, with higher R at all voltage levels. 2 Therefore, when fitting the changing trend of PD peak, a smooth line fitting should be used.

[0104] Through a multi-dimensional analysis of partial discharge characteristics, its operating patterns and characteristic patterns are systematically revealed. This multi-dimensional characteristic analysis provides an important theoretical foundation for a deeper understanding of partial discharge mechanisms and for assessing equipment operating conditions. This systematic analysis of these characteristics not only reveals the fundamental operating patterns of partial discharge but also provides a scientific basis for predicting and preventing partial discharge.

[0105] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for diagnosing and analyzing partial discharge of power cables based on slope behavior analysis, characterized in that: include: Use sensors to collect power data and perform pre-processing; Based on the preprocessing results, a ramp behavior analysis algorithm is used to identify discharge events from the signal sequence; Based on the statistics of discharge events, the phase distribution, polarity distribution and time interval parameters at each voltage level are visualized and analyzed; Use smoothing spline function to fit the characteristic variation law under different voltage levels; The slope behavior analysis algorithm based on the preprocessing results includes setting a dynamic threshold to distinguish significant events from stable events, detecting the amplitude difference and slope angle of significant events to quantify the steepness; The visualization and analysis include correlating the number of significant events with the PD peak distribution through a dual Y-axis graph, establishing a frequency correlation between voltage, phase angle, and PD amplitude using a three-dimensional histogram, and analyzing the variation of PD pulse time difference with voltage level based on a two-dimensional bar graph.

2. The method for diagnosing and analyzing partial discharge of power cables based on slope behavior analysis according to claim 1, characterized in that: The use of sensors to collect power data includes sequentially applying 50 Hz AC voltages of different voltage levels to XLPE cable samples through a programmable AC power supply, measuring the output of the HFCT and recording the data using a digital storage oscilloscope, recording 32 cycles of data at each voltage level, including actual partial discharge events and noise generated by the measurement; While recording the signal, the voltage waveform is also recorded synchronously, and the collected signal data is stored in a time series format, including timestamps, partial discharge signal amplitudes, and corresponding voltage phase information.

3. The method for diagnosing and analyzing partial discharge of power cables based on slope behavior analysis according to claim 2, characterized in that: The preprocessing is to perform multi-resolution analysis on the signal using wavelet transform to remove high-frequency noise and perform normalization processing to map the original amplitude of the measurement data to the interval [-1,1].

4. The method for diagnosing and analyzing partial discharge of power cables based on slope behavior analysis according to claim 3, characterized in that: The method of using a ramp behavior analysis algorithm based on the preprocessing results to identify discharge events from the signal sequence is to set a threshold, traverse the normalized signal data, and distinguish significant events from stable events in the time series data set; Detect the starting and ending points where the signal amplitude exceeds the threshold. If it exceeds the threshold, the time series data is classified as a significant event, otherwise it is classified as a stationary event. The significant time includes significant rising events and significant falling events; The changing moments of significant events include the start time and end time of the event. The amplitude difference corresponding to the start time and end time and the inclination angle of the slope formed by the difference relative to the horizontal direction are calculated to quantify the steepness of the change and output the slope behavior analysis diagram.

5. The method for diagnosing and analyzing partial discharge of power cables based on slope behavior analysis according to claim 4, characterized in that: The threshold is iteratively adjusted by manually analyzing the peak noise amplitude and dynamically according to the signal-to-noise ratio and partial discharge characteristics of the experimental data.

6. The method for diagnosing and analyzing partial discharge of power cables based on slope behavior analysis according to claim 5, characterized in that: The phase distribution, polarity distribution and time interval parameters at each voltage level based on the discharge event statistics are visualized and analyzed, including: Based on the slope behavior analysis diagram, the characteristics of significant events are reflected and extracted for analysis; Statistical analysis was performed on significant events at each voltage level. The number of significant events within each voltage cycle, the distribution of PD peak values of significant events, and the distribution frequency of PD peaks at different phase angles within significant events were calculated. The main phase intervals of partial discharge events and the time difference changes of each PD pulse measured at different voltages were determined. Based on the number of significant events and the distribution of PD peaks of significant events within each voltage cycle, a dual Y-axis graph is drawn at different voltage levels. The X-axis represents the AC voltage cycle, the left Y-axis represents the number of various PD peaks, and the right Y-axis represents the number of significant events. PD peak data and significant event data are displayed in different ways to analyze the changing trends between PD peak characteristics and the number of significant events in each cycle. By comparing data at different voltage levels, the relationship between voltage level and the number of significant events was studied, and the characteristics of positive and negative polarity PD peaks at all voltage levels were analyzed to reveal the influence of voltage level on PD characteristics. Based on the distribution frequency of PD peaks at different phase angles in significant events, a three-dimensional histogram is plotted to visualize partial discharge data at different voltage levels. The phase angle range and normalized amplitude range are divided into several intervals, where the X-axis represents the phase angle, the Y-axis represents the normalized amplitude, and the Z-axis represents the occurrence frequency. The height of each cube represents the number of discharge events at a specific phase angle and amplitude combination. By studying the frequency distribution of PD events within a specific voltage phase angle range, the correlation between voltage level, phase angle and PD amplitude is established. Based on the time difference change of each PD pulse measured at different voltages, a two-dimensional bar graph is used to plot the time difference change of each PD pulse measured at different voltages, where the X-axis represents the AC voltage period and the Y-axis represents the time difference of each PD pulse; By comparing data at different voltage levels, the changing trend of the overall range of the time difference of each PD pulse at different voltage levels is studied, and the correlation between voltage level and PD pulse time characteristics is established.

7. The method for diagnosing and analyzing partial discharge of power cables based on slope behavior analysis according to claim 6, characterized in that: The method of fitting the characteristic variation rules under different voltage levels using the smoothing spline function includes: According to the time difference and amplitude change of each PD pulse measured at different voltages, linear polynomial fitting and smoothing spline fitting were performed respectively. The goodness of fit value and root mean square error of the fitting curve were calculated to evaluate the fitting effect. The goodness of fit is proportional to the explanatory power of the data, and the root mean square error is proportional to the deviation of the actual data.

8. 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 a power cable partial discharge diagnosis and analysis method based on slope behavior analysis according to any one of claims 1 to 7 are implemented.

9. 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 a power cable partial discharge diagnosis and analysis method based on slope behavior analysis according to any one of claims 1 to 7 are implemented.

Images