Cable sheath fault signal analysis method and system based on full digital processing

Abstract

This invention relates to the field of fault prediction technology and proposes a method and system for analyzing cable sheath fault signals based on all-digital processing. The method includes: performing band-stop timing differential joint purification on the raw signal set acquired by the receiver to obtain a preprocessed signal set of the cable sheath under test; dynamically configuring the operating mode of the receiver; matching the corresponding fault feature analysis strategy in the preset strategy library to obtain the multi-dimensional fault feature vector of the cable sheath under test; calculating the comprehensive confidence score of the cable sheath under test using a self-calibration confidence evaluation formula; calculating a dynamic threshold based on the historical composite error dataset of the cable sheath under test using the self-calibration confidence evaluation formula to analyze the fault location of the cable sheath under test; generating a location result report and dynamically updating the preset strategy library and the self-calibration confidence evaluation formula, thereby effectively improving the accuracy of fault location through quantitative evaluation.

Description

Technical Field

[0001] This invention relates to the field of fault prediction technology, and in particular to a method and system for analyzing cable sheath fault signals based on fully digital processing. Background Technology

[0002] As a crucial component of power cables, the integrity of the cable sheath directly impacts the operational safety of cable lines. During long-term operation, cables are susceptible to grounding faults due to external forces, chemical corrosion, and construction defects. This can lead to circulating currents within the metallic sheath, accelerating insulation aging and even causing power outages. Currently, fault location in cable sheaths primarily relies on methods such as step voltage analysis, fault direction identification, and differential voltage comparison. In practical applications, receivers are used to collect electrical signals from the sheath, and operators manually determine the fault location based on changes in signal waveform color or amplitude. However, cable laying sites often experience complex electromagnetic interference from power frequency currents and traction currents. Combined with the DC drift component introduced by the sensors themselves, the acquired raw signals are contaminated with significant noise.

[0003] Existing methods rely excessively on operator experience in the fault location decision-making process. Operators must infer fault locations based on observed waveform morphology, deflection direction, and their subjective understanding of the current line characteristics. Different operators may arrive at significantly different conclusions in the same scenario. When the fault point is located in special laying environments such as high-resistance grounding or under support structures, the energy attenuation of the fault signal is rapid and its characteristics become blurred. Operators struggle to accurately extract robust indicators of the fault state and quantify the reliability of their current judgments. Furthermore, existing methods lack effective reuse of historical test data. Each fault investigation must start from scratch, failing to calibrate current judgments using accumulated testing experience. This compromises the consistency and accuracy of location results, impacting overall troubleshooting efficiency. Summary of the Invention

[0004] This invention provides a method and system for analyzing cable sheath fault signals based on fully digital processing, in order to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a cable sheath fault signal analysis method based on all-digital processing, comprising: S1: Perform band-stop timing differential joint purification on the raw signal set acquired by the receiver to obtain the pre-processed signal set of the cable sheath under test, automatically identify the type of the connected sensor, and dynamically configure the working mode of the receiver. S2: Based on the working mode, match the corresponding fault feature analysis strategy in the preset strategy library, perform three-dimensional coupling analysis on the preprocessed signal set, and obtain the multi-dimensional fault feature vector of the cable sheath to be tested. S3: Based on the multidimensional fault feature vector, the comprehensive confidence score of the cable sheath under test is calculated using the self-calibration confidence evaluation formula; S4: Based on the historical composite error dataset of the cable sheath to be tested, calculate the dynamic threshold using the self-calibration confidence evaluation formula. When the comprehensive confidence score exceeds the dynamic threshold, analyze the location of the fault point in the cable sheath to be tested. S5: Generate a location result report based on the fault location, working mode and comprehensive confidence score, and dynamically update the preset strategy library and the self-calibration confidence evaluation formula.

[0006] In a preferred embodiment, the step of performing band-stop timing differential joint purification on the raw signal set acquired by the receiver to obtain a preprocessed signal set of the cable sheath under test, automatically identifying the type of the connected sensor, and dynamically configuring the operating mode of the receiver includes: Acquire the raw signal set collected by the receiver, the raw signal set including multiple sampling points arranged sequentially along the time axis; A band-stop filter bank is applied to the original signal set to suppress the power frequency fundamental wave and integer harmonic interference components, resulting in a filtered signal set. The filtered signal set is then subjected to time-series differential processing, and the DC drift component is eliminated by utilizing the amplitude difference between adjacent sampling points, resulting in a preprocessed dataset. Based on the periodic fluctuation characteristics and amplitude envelope characteristics in the preprocessed dataset, the sensor type is automatically identified and a sensor type identifier is obtained by matching them with a preset sensor feature library. Based on the sensor type identifier, the operating mode parameter set corresponding to the sensor type identifier is matched, and the operating mode parameter set is written into the signal processing channel of the receiver to complete the configuration of the receiver's operating mode.

[0007] In a preferred embodiment, based on the working mode, matching the corresponding fault feature analysis strategy in the preset strategy library, and performing three-dimensional coupled analysis on the preprocessed signal set to obtain the multi-dimensional fault feature vector of the cable sheath under test includes: In the preset strategy library, fault feature analysis strategies corresponding to the working mode are retrieved. The fault feature analysis strategies include time-domain feature types, frequency-domain feature types, and mode feature types. Based on the fault feature analysis strategy, three-dimensional feature extraction is performed on the preprocessed signal set to obtain the time-domain feature components, frequency-domain feature components, and mode feature components of the cable layer under test. The feature vectors of the three feature components are concatenated to obtain the multidimensional fault feature vector of the cable sheath under test.

[0008] In a preferred embodiment, the step of calculating the comprehensive confidence score of the cable sheath under test based on the multidimensional fault feature vector using a self-calibrating confidence assessment formula includes: The multidimensional fault feature vector is normalized, and the processing result is compared with the historical normal feature vector template corresponding to the working mode to obtain the time domain feature deviation component, frequency domain feature deviation component and mode feature deviation component. Based on the multi-dimensional weighting coefficients and the feature deviation components in the fault feature analysis strategy, the comprehensive confidence score of the cable sheath under test is calculated using the self-calibration confidence evaluation formula.

[0009] In a preferred embodiment, the calibration confidence assessment formula is expressed as: ; In the formula, This represents the overall confidence score of the cable sheath under test. This indicates the index number corresponding to the characteristic deviation component. , This indicates the number of characteristic deviation components of the cable sheath under test. Indicates the first layer of the cable sheath under test Each characteristic deviation component Indicates the first The dimension weight coefficients corresponding to the feature deviation components This represents the historical error fluctuation value of the cable sheath under test. This represents the self-calibration adjustment factor of the cable sheath under test.

[0010] In a preferred embodiment, the step of calculating a dynamic threshold based on the historical composite error dataset of the cable sheath under test using the self-calibration confidence assessment formula, and analyzing the fault location of the cable sheath under test when the comprehensive confidence score exceeds the dynamic threshold, includes: Obtain the historical composite error dataset of the cable sheath to be tested, and filter the historical evaluation records associated with the current working mode from the historical composite error dataset to obtain a subset of historical records of the same mode; Based on the historical confidence scores corresponding to each historical evaluation in the same pattern historical record subset, the dynamic threshold benchmark value of the cable sheath to be tested is determined. Based on the historical composite error dataset, the confidence trend feature closest to the current evaluation time is extracted, and the trend correction amount corresponding to the confidence trend feature is determined; The dynamic threshold reference value and the trend correction amount are corrected and superimposed to obtain the dynamic threshold of the cable sheath to be tested. When the overall confidence score exceeds the dynamic threshold, the location of the fault point in the sheath of the cable under test is analyzed.

[0011] In a preferred embodiment, the step of analyzing the location of the fault point in the cable sheath under test when the overall confidence score exceeds the dynamic threshold includes: The comprehensive confidence score is compared with the dynamic threshold. When the comprehensive confidence score is greater than the dynamic threshold, it is determined that the sheath of the cable under test has a fault, and the corresponding multidimensional fault feature vector is obtained. Based on the polarity reversal feature, direction indication feature and voltage gradient feature in the multidimensional fault feature vector, the multidimensional fault orientation feature of the fault point in the sheath of the cable under test relative to the current measurement position is determined. The location of the fault point in the cable sheath under test is obtained by comprehensively weighting the orientation in the multidimensional fault orientation feature.

[0012] In a preferred embodiment, the step of generating a location result report based on the fault location, operating mode, and comprehensive confidence score, and dynamically updating the preset strategy library and the self-calibration confidence assessment formula, includes: The location of the fault point, the working mode, and the comprehensive confidence score are obtained. The location of the fault point is mapped to a preset cable path coordinate reference system to obtain the relative position coordinates. Based on the comparison between the comprehensive confidence score and the preset confidence level threshold, the confidence level corresponding to the comprehensive confidence score is obtained; Based on the relative position coordinates, the mode identifier corresponding to the working mode, and the confidence level, a positioning result report of the cable sheath to be tested is generated; The preset strategy library and the self-calibration confidence evaluation formula are dynamically updated.

[0013] In a preferred embodiment, the dynamic updating of the preset strategy library and the self-calibration confidence evaluation formula includes: The deviation between the comprehensive confidence score and the dynamic threshold is used as the error feedback amount of the cable sheath under test. Based on the error feedback amount, the historical normal feature vector template corresponding to the working mode is retrieved from the preset strategy library and benchmarked. The correction result replaces the original historical normal feature vector template in the preset strategy library. Based on the error feedback, the self-calibration adjustment factor in the self-calibration confidence assessment formula is corrected, and the updated self-calibration adjustment factor is written into the self-calibration confidence assessment formula.

[0014] To address the aforementioned problems, this invention also provides a cable sheath fault signal analysis system based on all-digital processing. The system includes a purification identification and configuration module, a feature extraction module, a confidence assessment module, a fault analysis module, and a location update module, wherein: The purification identification and configuration module is used to perform band-stop timing differential joint purification on the original signal set collected by the receiver to obtain the pre-processed signal set of the cable sheath under test, automatically identify the type of sensor connected, and dynamically configure the working mode of the receiver. The feature extraction module is used to match the corresponding fault feature analysis strategy in the preset strategy library based on the working mode, perform three-dimensional coupling analysis on the preprocessed signal set, and obtain the multi-dimensional fault feature vector of the cable sheath to be tested. The confidence assessment module is used to calculate the comprehensive confidence score of the cable sheath under test based on the multidimensional fault feature vector using a self-calibrating confidence assessment formula. The fault analysis module is used to calculate a dynamic threshold based on the historical composite error dataset of the cable sheath under test using the self-calibration confidence evaluation formula. When the comprehensive confidence score exceeds the dynamic threshold, the fault location of the cable sheath under test is analyzed. The location update module is used to generate a location result report based on the fault location, working mode and comprehensive confidence score, and dynamically update the preset strategy library and the self-calibration confidence evaluation formula.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention performs band-stop timing differential joint purification on the original signal set, which can suppress power frequency interference and DC drift while preserving fault characteristics. This results in a higher signal-to-noise ratio for the preprocessed signal set entering the subsequent analysis stage. Based on automatic sensor type identification and dynamic configuration of operating modes, the receiver automatically switches to the corresponding signal processing path when step voltage, intelligent direction sensors, and intelligent differential pressure detection devices are connected, eliminating the interference of manual setting errors on subsequent analysis and ensuring the consistency of multi-dimensional fault feature vector extraction. In the feature extraction stage, time-domain feature extraction, frequency-domain feature extraction, and mode feature extraction are performed simultaneously on the preprocessed signal set, and the feature components of the three dimensions are concatenated. This enhances fault features that are difficult to distinguish under a single dimension under the joint representation of the three dimensions, improving the completeness of the multi-dimensional fault feature vector in characterizing the sheathing state.

[0016] 2. This invention maps multidimensional fault feature vectors into a comprehensive confidence score using a self-calibrated confidence assessment formula, transforming fault determination from qualitative observation into a quantifiable evaluation indicator. During fault location analysis, a dynamic threshold is generated using a historical composite error dataset. Fault determination is only triggered when the comprehensive confidence score exceeds the dynamic threshold, reducing the risk of misjudgment under different operating conditions with fixed thresholds. After generating the location result report, error feedback is used to update the templates in the preset strategy library and the adjustment factors in the self-calibrated confidence assessment formula. This allows the fault determination criteria to adaptively optimize with the accumulation of test data, improving the consistency of location results and troubleshooting efficiency. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating a cable sheath fault signal analysis method based on all-digital processing, provided in an embodiment of the present invention. Figure 2 This is a functional block diagram of a cable sheath fault signal analysis system based on all-digital processing provided in an embodiment of the present invention; The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0018] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0019] This application provides a method for analyzing cable sheath fault signals based on all-digital processing. The executing entity of this method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the method for analyzing cable sheath fault signals based on all-digital processing can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms.

[0020] Reference Figure 1 The diagram shown is a flowchart illustrating a cable sheath fault signal analysis method based on all-digital processing according to an embodiment of the present invention. In this embodiment, the cable sheath fault signal analysis method based on all-digital processing includes: S1: Perform band-stop timing differential joint purification on the raw signal set acquired by the receiver to obtain the pre-processed signal set of the cable sheath under test, automatically identify the type of the connected sensor, and dynamically configure the working mode of the receiver. In this embodiment of the invention, the step of performing band-stop timing differential joint purification on the raw signal set acquired by the receiver to obtain a preprocessed signal set of the cable sheath under test, automatically identifying the type of the connected sensor, and dynamically configuring the operating mode of the receiver includes: Acquire the raw signal set collected by the receiver, the raw signal set including multiple sampling points arranged sequentially along the time axis; A band-stop filter bank is applied to the original signal set to suppress the power frequency fundamental wave and integer harmonic interference components, resulting in a filtered signal set. The filtered signal set is then subjected to time-series differential processing, and the DC drift component is eliminated by utilizing the amplitude difference between adjacent sampling points, resulting in a preprocessed dataset. Based on the periodic fluctuation characteristics and amplitude envelope characteristics in the preprocessed dataset, the sensor type is automatically identified and a sensor type identifier is obtained by matching them with a preset sensor feature library. Based on the sensor type identifier, the operating mode parameter set corresponding to the sensor type identifier is matched, and the operating mode parameter set is written into the signal processing channel of the receiver to complete the configuration of the receiver's operating mode.

[0021] It should be specifically noted that the receiver acquires the raw signal set through its sensor interface. This raw signal set contains multiple sampling points arranged sequentially along the time axis, with each sampling point corresponding to the amplitude of the sheath electrical signal at a given moment.

[0022] A band-stop filter bank is applied to the original signal set. The band-stop filter bank is composed of multiple notch filters with different center frequencies cascaded together. The center frequency of each notch filter corresponds to the frequency position of the fundamental frequency wave and its integer harmonics. Each sampling point in the original signal set is passed through the notch filter bank in sequence for attenuation processing, generating a filtered signal set with the power frequency interference component removed.

[0023] The filtered signal set is processed by time-series differential processing. The amplitudes of the filtered signals corresponding to two adjacent sampling points on the time axis are subtracted. The difference between the amplitude of the steady-state DC drift component, which is close to zero in a very small time interval, and the amplitude of the fault characteristic signal, which is significantly different, is used to cancel the DC drift component by the amplitude difference between adjacent sampling points, thus generating a preprocessed signal set.

[0024] Periodic fluctuation features and amplitude envelope features are extracted from the preprocessed signal set. Periodic fluctuation features are obtained by detecting the time interval between the alternation of peak and valley values ​​in the preprocessed signal set. Amplitude envelope features are obtained by extracting the contour line formed by continuous peak points in the preprocessed signal set. The periodic fluctuation features and amplitude envelope features are compared one by one with the reference periodic features and reference envelope features corresponding to each sensor type in the preset sensor feature library. The sensor type with the highest feature matching degree is selected as the sensor type identifier.

[0025] Furthermore, the preset sensor feature library is a set of sensor electrical signal characteristic parameters pre-established and stored inside the receiver, containing a family of reference feature vectors corresponding to all sensor types supported by the receiver interface. Each sensor type in this sensor feature library corresponds to a set of reference feature vectors, which are composed of reference periodic fluctuation features and reference amplitude envelope features.

[0026] The baseline periodic fluctuation feature is the time interval between the alternating peak and trough values ​​extracted from a known normal sheath signal acquired under standard test conditions using a corresponding type of sensor. The baseline amplitude envelope feature is the contour line baseline shape formed by consecutive peak points extracted from the acquired signal under the same standard test conditions. The process of establishing the sensor feature library involves sequentially connecting each supported sensor type to the receiver interface, inputting a standard excitation signal to the receiver, acquiring the sensor's response signal under standard excitation, extracting the periodic fluctuation feature and amplitude envelope feature from the response signal, associating and storing the extracted periodic fluctuation feature and amplitude envelope feature with the sensor type, and repeating this process until all sensor types have completed the feature extraction and association storage of the standard response signal.

[0027] Each set of reference feature vectors stored in the sensor feature library corresponds one-to-one with a sensor type identifier. When automatically identifying the sensor type of a preprocessed signal set, the periodic fluctuation features and amplitude envelope features extracted from the preprocessed signal set are compared one by one with the reference periodic fluctuation features and reference amplitude envelope features of each set of reference feature vectors in the sensor feature library. The sensor type identifier corresponding to the set of reference feature vectors with the highest feature matching degree is selected as the identification result.

[0028] The sensor type identifier is automatically generated by the receiver by extracting the feature quantities that characterize the physical properties of the preprocessed signal and matching them with the reference features in the preset sensor feature library. The value of the sensor type identifier corresponds one-to-one with the types of sensors supported by the receiver interface, including the intelligent orientation sensor identifier, the step voltage probe identifier, and the intelligent differential pressure detection device identifier.

[0029] When the receiver detects that the matching degree between the periodic fluctuation characteristics and amplitude envelope characteristics of the preprocessed signal set and the reference periodic characteristics and reference envelope characteristics corresponding to the intelligent orientation sensor reaches its maximum, an intelligent orientation sensor identifier is generated. When the receiver detects that the matching degree between the periodic fluctuation characteristics and amplitude envelope characteristics of the preprocessed signal set and the reference periodic characteristics and reference envelope characteristics corresponding to the step voltage probe reaches its maximum, a step voltage probe identifier is generated. When the receiver detects that the matching degree between the periodic fluctuation characteristics and amplitude envelope characteristics of the preprocessed signal set and the reference periodic characteristics and reference envelope characteristics corresponding to the intelligent differential pressure detection device reaches its maximum, an intelligent differential pressure detection device identifier is generated.

[0030] The preset mode configuration table is queried based on the sensor type identifier. The mode configuration table stores the correspondence between the sensor type identifier and the working mode parameter set. The working mode parameter set bound to the sensor type identifier is retrieved from the mode configuration table. The gain coefficient and signal conditioning parameters in the working mode parameter set are written into the register corresponding to the digital signal processing link of the receiver to complete the dynamic configuration of the receiver's working mode.

[0031] By using band-stop timing differential joint purification to simultaneously suppress power frequency interference and eliminate DC drift, the preprocessed signal set improves the signal-to-noise ratio while retaining the characteristic components of the sheath fault, providing a high-quality signal foundation for subsequent feature extraction. Combined with automatic sensor type identification and dynamic writing of the operating mode parameter set, it avoids operational errors introduced by manual settings, ensuring that the receiver signal processing link and the currently accessed sensor always remain matched, thus guaranteeing the consistency and reliability of the input conditions throughout the entire analysis process from the source.

[0032] S2: Based on the working mode, match the corresponding fault feature analysis strategy in the preset strategy library, perform three-dimensional coupling analysis on the preprocessed signal set, and obtain the multi-dimensional fault feature vector of the cable sheath to be tested. In this embodiment of the invention, the step of matching the corresponding fault feature analysis strategy in the preset strategy library based on the working mode, and performing three-dimensional coupled analysis on the preprocessed signal set to obtain the multi-dimensional fault feature vector of the cable sheath under test includes: In the preset strategy library, fault feature analysis strategies corresponding to the working mode are retrieved. The fault feature analysis strategies include time-domain feature types, frequency-domain feature types, and mode feature types. Based on the fault feature analysis strategy, three-dimensional feature extraction is performed on the preprocessed signal set to obtain the time-domain feature components, frequency-domain feature components, and mode feature components of the cable layer under test. The feature vectors of the three feature components are concatenated to obtain the multidimensional fault feature vector of the cable sheath under test.

[0033] It should be specifically explained that, based on the working mode identifier, the fault feature analysis strategy corresponding to the working mode identifier is retrieved from the preset strategy library. The preset strategy library stores a correspondence table between working mode identifiers and fault feature analysis strategies. The working mode identifiers are step voltage mode identifier, intelligent direction sensor mode identifier, and intelligent differential pressure detection mode identifier. The retrieval method is to input the working mode identifier of the current working mode as the query condition into the correspondence table, and retrieve the fault feature analysis strategy bound to the working mode identifier from the correspondence table. The fault feature analysis strategy defines the time domain feature type, frequency domain feature type, and mode feature type to be executed in this three-dimensional feature extraction.

[0034] Furthermore, the preset strategy library is a collection of fault feature analysis strategies and feature benchmark data that are pre-built and stored inside the receiver. It consists of three parts: a fault feature analysis strategy set, a historical normal feature vector template set, and a fault feature benchmark value set.

[0035] The fault feature analysis strategy set stores the fault feature analysis strategies corresponding to each working mode. Each fault feature analysis strategy is stored with the working mode identifier as an index. The strategy content includes the corresponding time domain feature type, frequency domain feature type and mode feature type.

[0036] When the operating mode is identified as a step voltage mode, the corresponding fault feature analysis strategy defines the time-domain feature types as signal period characteristics, duty cycle characteristics, and amplitude envelope characteristics; the frequency-domain feature types as power frequency interference residual measurement and signal main frequency energy distribution characteristics; and the mode feature type as polarity reversal characteristics. When the operating mode is identified as a smart direction sensor mode, the corresponding fault feature analysis strategy defines the time-domain feature types as signal period characteristics, duty cycle characteristics, and amplitude envelope characteristics; the frequency-domain feature types as power frequency interference residual measurement and signal main frequency energy distribution characteristics; and the mode feature type as direction indication characteristics. When the operating mode is identified as a smart differential pressure detection mode, the corresponding fault feature analysis strategy defines the time-domain feature types as signal period characteristics, duty cycle characteristics, and amplitude envelope characteristics; the frequency-domain feature types as power frequency interference residual measurement and signal main frequency energy distribution characteristics; and the mode feature type as voltage gradient characteristics. The fault feature analysis strategy set is established by sequentially recording the feature analysis strategies corresponding to each operating mode into the receiver's internal memory, and establishing a retrieval relationship using the operating mode identifier as an index.

[0037] The historical normal feature vector template set stores historical normal feature vector templates corresponding to each operating mode. Each historical normal feature vector template contains a normal time-domain reference component, a normal frequency-domain reference component, and a normal mode reference component. These reference components are obtained by extracting signals from known normal sheaths using the connected sensors under the corresponding operating mode and storing the resulting feature components. The fault feature reference value set stores fault feature reference values ​​corresponding to each operating mode. These fault feature reference values ​​are comprehensive reference values ​​obtained by fusing the feature components after extracting signals from known faulty sheaths using the connected sensors under the corresponding operating mode and performing three-dimensional feature extraction.

[0038] Based on the time-domain feature types defined in the fault feature analysis strategy, time-domain feature extraction is performed on the preprocessed signal set. The time-domain feature extraction method is to detect the alternation period of the peak and valley values ​​of the signal in the preprocessed signal set along the time axis to obtain the signal periodic feature, identify the time position of the peak in half a period to obtain the duty cycle feature, and extract the contour line formed by connecting the peak points in each period of the signal to obtain the amplitude envelope feature. The above signal periodic feature, duty cycle feature and amplitude envelope feature are combined into time-domain feature components.

[0039] Based on the frequency domain feature types defined in the fault feature analysis strategy, frequency domain feature extraction is performed on the preprocessed signal set. The frequency domain feature extraction method is to perform frequency band energy detection on the preprocessed signal set through a multi-channel filter group. The multi-channel filter group consists of multiple bandpass filters covering the power frequency fundamental band and each preset harmonic band. The preprocessed signal set is input into each bandpass filter respectively, and the energy amplitude of each frequency band is extracted. The energy amplitude of the power frequency fundamental band and the energy amplitude of each harmonic band are used as the power frequency interference residual measure. At the same time, the main frequency energy distribution of the preprocessed signal set is detected to obtain the signal main frequency energy distribution characteristics. The power frequency interference residual measure and the signal main frequency energy distribution characteristics are combined into frequency domain feature components.

[0040] Based on the mode feature types defined in the fault feature analysis strategy, mode feature extraction is performed on the preprocessed signal set. When the working mode is step voltage mode, polarity reversal features representing the positive and negative deflection directions of the signal are extracted from the preprocessed signal set. When the working mode is intelligent direction sensor mode, direction indication features representing the transition of the signal from peak to valley value are extracted from the preprocessed signal set. When the working mode is intelligent differential pressure detection mode, voltage gradient features representing the rate of change of the signal amplitude along the line are extracted from the preprocessed signal set. The extracted polarity reversal features, direction indication features, or voltage gradient features are used as mode feature components.

[0041] The time-domain feature components, frequency-domain feature components, and mode feature components are concatenated and combined according to a preset splicing order to generate a multi-dimensional feature vector. This multi-dimensional feature vector contains feature information from three dimensions: time-domain feature components, frequency-domain feature components, and mode feature components. Finally, a multi-dimensional fault feature vector of the cable sheath under test is obtained.

[0042] By retrieving the corresponding fault feature analysis strategy from the preset strategy library using the working mode identifier, the selection of time-domain, frequency-domain, and mode feature types remains consistent with the currently accessed sensor types, avoiding extraction bias caused by feature type mismatch. The time-domain, frequency-domain, and mode feature components are extracted simultaneously from the preprocessed signal set, and these three feature components are concatenated. This enhances the fault characterization that is difficult to identify under a single dimension through a three-dimensional joint expression, improving the completeness of the multi-dimensional fault feature vector in depicting the sheath fault state and providing more comprehensive input for subsequent confidence assessment.

[0043] S3: Based on the multidimensional fault feature vector, the comprehensive confidence score of the cable sheath under test is calculated using the self-calibration confidence evaluation formula; In this embodiment of the invention, the step of calculating the comprehensive confidence score of the cable sheath under test based on the multidimensional fault feature vector using a self-calibrating confidence assessment formula includes: The multidimensional fault feature vector is normalized, and the processing result is compared with the historical normal feature vector template corresponding to the working mode to obtain the time domain feature deviation component, frequency domain feature deviation component and mode feature deviation component. Based on the multi-dimensional weighting coefficients and the feature deviation components in the fault feature analysis strategy, the comprehensive confidence score of the cable sheath under test is calculated using the self-calibration confidence evaluation formula.

[0044] The calibration confidence assessment formula is expressed as follows: ; In the formula, This represents the overall confidence score of the cable sheath under test. This represents the index number corresponding to the characteristic deviation component. , This indicates the number of characteristic deviation components of the cable sheath under test. Indicates the first layer of the cable sheath under test Each characteristic deviation component Indicates the first The dimension weight coefficients corresponding to the feature deviation components. This represents the historical error fluctuation value of the cable sheath under test. This represents the self-calibration adjustment factor of the cable sheath under test.

[0045] Specifically, time-domain feature components, frequency-domain feature components, and mode feature components are extracted from the multi-dimensional fault feature vector. The signal periodicity, duty cycle, and amplitude envelope features in the time-domain feature components are normalized by scaling each feature value with the reference range of the corresponding feature type, so that each feature value is mapped to a uniform numerical range, thus obtaining the normalized time-domain component. The power frequency interference residual measurement and signal main frequency energy distribution features in the frequency-domain feature components are normalized to obtain the normalized frequency-domain component. The polarity reversal feature, direction indication feature, or voltage gradient feature in the mode feature components are normalized to obtain the normalized mode component.

[0046] Obtain the historical normal feature vector template corresponding to the working mode. The historical normal feature vector template includes normal time domain reference components, normal frequency domain reference components, and normal mode reference components. Calculate the difference between each feature value in the normalized time domain component and the corresponding reference value in the normal time domain reference component, and take the absolute value to obtain the time domain feature deviation component. Calculate the difference between each feature value in the normalized frequency domain component and the corresponding reference value in the normal frequency domain reference component, and take the absolute value to obtain the frequency domain feature deviation component. Calculate the difference between each feature value in the normalized mode component and the corresponding reference value in the normal mode reference component, and take the absolute value to obtain the mode feature deviation component.

[0047] The time-domain dimension weight coefficients corresponding to the time-domain feature types, the frequency-domain dimension weight coefficients corresponding to the frequency-domain feature types, and the mode-domain dimension weight coefficients corresponding to the mode feature types are obtained from the fault feature analysis strategy. Each dimension weight coefficient is retrieved from the weight allocation table stored in the fault feature analysis strategy. Each feature type in the weight allocation table corresponds to a pre-set dimension weight coefficient.

[0048] Extract the historical confidence score sequence corresponding to the working mode from the historical composite error dataset, calculate the standard deviation of the historical confidence score sequence, and obtain the historical error fluctuation value. The standard deviation is calculated by taking the arithmetic mean of the confidence scores of each historical assessment in the historical confidence score sequence, then calculating the square of the difference between each confidence score and the arithmetic mean, summing the square values ​​and dividing by the number of historical assessments to obtain the variance, and taking the square root of the variance to obtain the historical error fluctuation value.

[0049] The self-calibration adjustment factor is retrieved from the preset calibration parameter table based on the sensor type identifier and the operating mode identifier. The preset calibration parameter table stores the correspondence between the combination of sensor type identifier and operating mode identifier and the self-calibration adjustment factor. The query method is to input the current sensor type identifier and operating mode identifier as joint query conditions into the preset calibration parameter table, and retrieve the corresponding self-calibration adjustment factor from the table.

[0050] Furthermore, the preset calibration parameter table is a set of calibration parameter queries that is pre-established and stored inside the receiver. It uses a combination of sensor type identifier and operating mode identifier as the query index and stores the self-calibration adjustment factor corresponding to each combination.

[0051] Each entry in the preset calibration parameter table includes a sensor type identifier field, an operating mode identifier field, and a self-calibration adjustment factor field. The sensor type identifier field takes the value of one of the following: intelligent direction sensor identifier, step voltage probe identifier, and intelligent differential pressure detection device identifier. The operating mode identifier field takes the operating mode identifier corresponding to the sensor type identifier. The self-calibration adjustment factor field takes the suppression intensity value obtained by measuring the confidence score fluctuation amplitude of multiple known normal protective layer tests under standard calibration conditions and back-calculating based on the fluctuation amplitude.

[0052] The overall confidence score is calculated using the self-calibration confidence assessment formula. The calculation method involves multiplying each of the time-domain, frequency-domain, and mode-domain bias components by its corresponding dimension weight coefficient, then adding one to obtain the weighted value for each bias dimension. These weighted values ​​are then multiplied together to obtain the combined product. The historical error fluctuation value is added one, and the natural logarithm is taken. This natural logarithm is then multiplied by the self-calibration adjustment factor, and the result is added one to obtain the denominator. The combined product is divided by the denominator to obtain the overall confidence score. The overall confidence score is a dimensionless scalar. When the individual bias components are small, the overall confidence score approaches a baseline level. As any bias component increases, the overall confidence score rises accordingly. The smaller the historical error fluctuation value, the more sensitively the overall confidence score reflects the current bias status.

[0053] Furthermore, in the self-calibration confidence assessment formula, the overall confidence score is determined jointly by the numerator and denominator. The characteristic deviation component in the numerator is derived from the absolute value of the difference between the characteristic components of each dimension in the multidimensional fault characteristic vector and the corresponding benchmark component in the historical normal characteristic vector template. The characteristic deviation component reflects the degree of deviation between the current sheath signal characteristics and the normal state benchmark; the greater the deviation, the larger the value of the characteristic deviation component.

[0054] The dimension weight coefficients are derived from the weight allocation table stored in the fault feature analysis strategy. The dimension weight coefficients corresponding to each feature type in the weight allocation table are pre-set according to the degree of contribution of that feature type to the fault characterization. The feature type with a greater degree of contribution has a larger dimension weight coefficient.

[0055] The historical error fluctuation value is derived from the standard deviation of the historical confidence score sequence corresponding to the current working mode in the historical composite error dataset. The standard deviation is obtained by taking the arithmetic mean of each confidence score in the sequence, calculating the square of the difference, summing the results, dividing by the number of times, and taking the square root. The greater the dispersion of each evaluation result in the historical confidence score sequence, the greater the value of the historical error fluctuation value.

[0056] The self-calibration adjustment factor is derived from the preset calibration parameter table. The preset calibration parameter table stores the correspondence between the combination of sensor type identifier and working mode identifier and the self-calibration adjustment factor. Different combinations of sensor types and working modes are assigned different self-calibration adjustment factors based on the fluctuation range of their confidence scores in historical tests.

[0057] The self-calibration confidence assessment formula quantifies the deviation between the feature components of each dimension in the multidimensional fault feature vector and the historical normal benchmark into a comprehensive confidence score. The numerator is constructed by multiplying the deviation components of each dimension with the dimension weight coefficients, and the denominator is constructed by using the historical error fluctuation value and the self-calibration adjustment factor. This allows the deviation of the current sheath signal characteristics from the normal state to be effectively captured under multidimensional joint characterization. At the same time, the dispersion of historical assessment results and the calibration characteristics of the current sensor and working mode can dynamically adjust the scoring results. When the deviation increases, the score increases accordingly, and when the historical fluctuation increases, the score is moderately suppressed. This transforms fault judgment from qualitative judgment that relies on the subjective observation of operators into quantitative assessment with self-calibration capabilities, providing a consistent and reliable basis for subsequent fault point analysis.

[0058] S4: Based on the historical composite error dataset of the cable sheath to be tested, calculate the dynamic threshold using the self-calibration confidence evaluation formula. When the comprehensive confidence score exceeds the dynamic threshold, analyze the location of the fault point in the cable sheath to be tested. In this embodiment of the invention, the step of calculating a dynamic threshold based on the historical composite error dataset of the cable sheath under test using the self-calibration confidence assessment formula, and analyzing the fault location of the cable sheath under test when the comprehensive confidence score exceeds the dynamic threshold, includes: Obtain the historical composite error dataset of the cable sheath to be tested, and filter the historical evaluation records associated with the current working mode from the historical composite error dataset to obtain a subset of historical records of the same mode; Based on the historical confidence scores corresponding to each historical evaluation in the same pattern historical record subset, the dynamic threshold benchmark value of the cable sheath to be tested is determined. Based on the historical composite error dataset, the confidence trend feature closest to the current evaluation time is extracted, and the trend correction amount corresponding to the confidence trend feature is determined; The dynamic threshold reference value and the trend correction amount are corrected and superimposed to obtain the dynamic threshold of the cable sheath to be tested. When the overall confidence score exceeds the dynamic threshold, the location of the fault point in the sheath of the cable under test is analyzed.

[0059] When the overall confidence score exceeds the dynamic threshold, the location of the fault point in the cable sheath under test is analyzed, including: The comprehensive confidence score is compared with the dynamic threshold. When the comprehensive confidence score is greater than the dynamic threshold, it is determined that the sheath of the cable under test has a fault, and the corresponding multidimensional fault feature vector is obtained. Based on the polarity reversal feature, direction indication feature and voltage gradient feature in the multidimensional fault feature vector, the multidimensional fault orientation feature of the fault point in the sheath of the cable under test relative to the current measurement position is determined. The location of the fault point in the cable sheath under test is obtained by comprehensively weighting the orientation in the multidimensional fault orientation feature.

[0060] It needs to be specifically explained that historical evaluation records associated with the current working mode are selected from the historical composite error dataset of the cable sheath to be tested. The selection method is to use the working mode identifier of the current working mode as the query condition, and search for all historical evaluation records in the historical composite error dataset whose working mode identifier field is the same as the identifier. All the retrieved historical evaluation records are arranged in chronological order to generate a subset of historical records of the same mode. Each record in the subset of historical records of the same mode contains the confidence score of a historical evaluation and the timestamp of that evaluation.

[0061] Extract the historical confidence scores corresponding to each historical assessment from the historical history subset of the same pattern, count the total number of historical assessments in the historical history subset of the same pattern, add up all the historical confidence scores one by one and divide by the total number of assessments to obtain the historical baseline mean, subtract each historical confidence score from the historical baseline mean to obtain the difference, square each difference and add them up one by one, divide the sum of squares by the total number of assessments to obtain the variance, take the square root of the variance to obtain the historical fluctuation range, and superimpose the historical baseline mean and the historical fluctuation range to obtain the dynamic threshold baseline value.

[0062] The confidence scores corresponding to the most recent assessments at the current assessment time are extracted from the historical composite error dataset and arranged in chronological order to form a recent confidence sequence. The direction of rise and fall of two adjacent scores in the recent confidence sequence is compared. If the subsequent score is consistently higher than the previous score, it is determined to be an upward trend. If the subsequent score is consistently lower than the previous score, it is determined to be a downward trend. If the scores alternate between rise and fall, it is determined to be a stable trend. The positive trend correction amount is determined based on the upward trend, the negative trend correction amount is determined based on the downward trend, and the zero trend correction amount is determined based on the stable trend, thus obtaining the trend correction amount.

[0063] The dynamic threshold is obtained by algebraically adding the dynamic threshold baseline value and the trend correction amount. When the trend is upward, the dynamic threshold is adjusted upward according to the baseline value. When the trend is downward, the dynamic threshold is adjusted downward according to the baseline value. When the trend is stable, the dynamic threshold remains consistent with the baseline value.

[0064] The comprehensive confidence score is numerically compared with the dynamic threshold. When the comprehensive confidence score is greater than the dynamic threshold, it is determined that there is a fault in the sheath of the cable under test, and the mode feature components are extracted from the multidimensional fault feature vector.

[0065] The first orientation of the fault point of the cable sheath under test relative to the current measurement position is determined based on the polarity reversal feature contained in the mode feature component. The polarity reversal feature indicates whether the fault point is in front of or behind the current measurement position by the positive or negative deflection direction of the step voltage signal.

[0066] The second orientation of the fault point of the cable sheath under test relative to the current measurement position is determined based on the directional indication feature contained in the mode feature component. The directional indication feature indicates whether the fault point is upstream or downstream of the current measurement position by the transition from peak to valley value of the current signal of the intelligent directional sensor.

[0067] The fault point of the cable sheath under test is determined relative to the current measurement location based on the voltage gradient features contained in the mode feature components. The voltage gradient features indicate whether the fault point is located at the far end or near end of the current measurement location through the rate of change of the voltage signal rising or falling along the line.

[0068] The first, second, and third orientations together constitute a multi-dimensional fault location feature. A comprehensive weighted judgment is applied to the first, second, and third orientations within this feature. The weighting method involves determining the weighting coefficient for each orientation feature based on the current operating mode. In step voltage mode, the polarity reversal feature has the highest weighting coefficient; in intelligent direction sensor mode, the direction indication feature has the highest weighting coefficient; and in intelligent differential pressure detection mode, the voltage gradient feature has the highest weighting coefficient. The final orientation is determined by the sum of the products of the indicated direction of each orientation feature and its corresponding weighting coefficient, thus obtaining the location of the fault point in the cable sheath under test.

[0069] A dynamic threshold benchmark is determined by filtering historical records of the same pattern from the historical composite error dataset and calculating the mean and fluctuation range of historical confidence scores. Simultaneously, recent confidence trend features are extracted to generate trend correction values, enabling the dynamic threshold to adaptively adjust according to the statistical distribution of historical evaluation data and recent trends. Fault determination is only triggered when the comprehensive confidence score exceeds this dynamic threshold, avoiding the problem of misjudgment or missed judgment that can easily occur with fixed thresholds under different operating conditions and signal environments. During fault location analysis, polarity reversal features, direction indication features, and voltage gradient features from the multi-dimensional fault feature vector are used to determine the fault orientation in different dimensions. A comprehensive weighted judgment is then applied to each orientation, integrating the directional information of multiple signal dimensions into the fault location result, improving the accuracy and consistency of fault location analysis.

[0070] S5: Generate a location result report based on the fault location, working mode and comprehensive confidence score, and dynamically update the preset strategy library and the self-calibration confidence evaluation formula.

[0071] In this embodiment of the invention, the step of generating a location result report based on the fault location, operating mode, and comprehensive confidence score, and dynamically updating the preset strategy library and the self-calibration confidence assessment formula, includes: The location of the fault point, the working mode, and the comprehensive confidence score are obtained. The location of the fault point is mapped to a preset cable path coordinate reference system to obtain the relative position coordinates. Based on the comparison between the comprehensive confidence score and the preset confidence level threshold, the confidence level corresponding to the comprehensive confidence score is obtained; Based on the relative position coordinates, the mode identifier corresponding to the working mode, and the confidence level, a positioning result report of the cable sheath to be tested is generated; The preset strategy library and the self-calibration confidence evaluation formula are dynamically updated.

[0072] The dynamic updating of the preset strategy library and the self-calibration confidence evaluation formula includes: The deviation between the comprehensive confidence score and the dynamic threshold is used as the error feedback amount of the cable sheath under test. Based on the error feedback amount, the historical normal feature vector template corresponding to the working mode is retrieved from the preset strategy library and benchmarked. The correction result replaces the original historical normal feature vector template in the preset strategy library. Based on the error feedback, the self-calibration adjustment factor in the self-calibration confidence assessment formula is corrected, and the updated self-calibration adjustment factor is written into the self-calibration confidence assessment formula.

[0073] Specifically, it's necessary to explain the acquisition of fault location, operating mode, and overall confidence score. Fault location includes the distance and direction information of the fault point relative to the current measurement location. The operating mode is one of the following: the step voltage mode identifier, the intelligent direction sensor mode identifier, or the intelligent differential pressure detection mode identifier corresponding to the currently connected sensor. The overall confidence score is a dimensionless evaluation value. The fault location is mapped to a preset cable path coordinate reference system. This preset cable path coordinate reference system is a coordinate system established along the cable extension direction with the origin at the starting point of the measured cable path. The distance and direction of the fault point relative to the current measurement location are converted into mileage along the cable path, obtaining the relative position coordinates.

[0074] The overall confidence score is compared step-by-step with a pre-set set of confidence level thresholds. This set includes a high-confidence threshold and a medium-confidence threshold, with the high-confidence threshold being greater than the medium-confidence threshold. When the overall confidence score is greater than the high-confidence threshold, the corresponding confidence level is determined to be high-confidence. When the overall confidence score is between the medium-confidence threshold and the high-confidence threshold, the corresponding confidence level is determined to be medium-confidence. When the overall confidence score is less than the medium-confidence threshold, the corresponding confidence level is determined to be low-confidence.

[0075] The relative position coordinates, the mode identifier corresponding to the working mode, and the confidence level are combined to generate a positioning result report. The positioning result report includes the mileage position of the fault point along the cable path, the name of the working mode used to perform this positioning, and the confidence level of this positioning result. The positioning result report is stored in text format in the data recording storage area of ​​the receiver.

[0076] The deviation between the overall confidence score and the dynamic threshold is calculated by subtracting the dynamic threshold from the overall confidence score to obtain the error feedback. The error feedback is positive when the overall confidence score is greater than the dynamic threshold, and zero when the overall confidence score is less than or equal to the dynamic threshold.

[0077] Based on the error feedback, the historical normal feature vector template corresponding to the working mode in the preset strategy library is corrected. The correction method is as follows: if the error feedback is positive, the normal time domain reference component, normal frequency domain reference component, and normal mode reference component in the historical normal feature vector template are weighted and superimposed with the error feedback. The weights of the superposition are set according to the fluctuation characteristics of each component. The superimposed result replaces the original reference component. If the error feedback is zero, the reference components in the historical normal feature vector template remain unchanged, and the corrected historical normal feature vector template replaces the original historical normal feature vector template in the preset strategy library.

[0078] The self-calibration adjustment factor in the self-calibration confidence assessment formula is modified based on the error feedback amount. The modification method is as follows: if the error feedback amount is positive, the self-calibration adjustment factor is increased by a step value that is positively correlated with the error feedback amount; if the error feedback amount is zero, the self-calibration adjustment factor is decreased by a fixed step value. The updated self-calibration adjustment factor is then written into the self-calibration confidence assessment formula to complete the dynamic update of the self-calibration confidence assessment formula.

[0079] By mapping the fault location to a preset cable path coordinate reference system, a location result report containing relative position coordinates, operating mode identifier, and confidence level is generated. This ensures that the output results of each fault location have a unified spatial reference benchmark and confidence level, facilitating operators to directly obtain the mileage location of the fault point. The error feedback between the comprehensive confidence score and the dynamic threshold is used to continuously correct the historical normal feature vector templates in the preset strategy library and the self-calibration adjustment factor in the self-calibration confidence evaluation formula. This allows the normal benchmark and calibration parameters to adaptively optimize with the accumulation of detection data. As the number of tests increases, the dynamic threshold and confidence evaluation results for subsequent fault determination increasingly align with the actual characteristics of the current line and sensors, achieving continuous self-calibration and performance improvement throughout the entire location process.

[0080] In a specific embodiment of the present invention, the cable sheath fault signal analysis method based on all-digital processing is deeply integrated into the receiver of the HC-10 cable sheath fault rapid location system. The method of the present invention will now be further described in conjunction with the specific functions and operation of this device: The transmitter outputs a DC pulsating signal with a duty cycle of 0.5:1.0 to the faulty cable. The operator connects the intelligent direction sensor to the receiver interface ⑨. After the receiver acquires the raw signal, it sequentially passes it through a 50Hz, 100Hz, and 150Hz band-stop filter bank to suppress power frequency interference. Then, it performs time-series differential analysis on adjacent sampling points to eliminate DC drift, obtaining a preprocessed signal set. The periodic fluctuation characteristics and amplitude envelope characteristics in the signal are extracted. The periodic fluctuation characteristics refer to a square wave with a period of approximately 2 seconds. These are matched with a preset sensor feature library, automatically identifying it as an intelligent direction sensor, and completing the dynamic configuration of the working mode.

[0081] The receiver performs three-dimensional feature extraction on the preprocessed signal set, including time domain, frequency domain, and mode. The time domain components include signal period, duty cycle, and current amplitude; the frequency domain components include power frequency interference residual and main frequency energy distribution; and the mode components include direction indication features. The three components are then concatenated into a multi-dimensional fault feature vector.

[0082] Deviation analysis was performed between the multidimensional fault feature vector and the historical normal feature vector template to obtain the feature deviation components and their weight coefficients for each dimension. These components, combined with historical error fluctuation values ​​and self-calibration adjustment factors, were then substituted into the self-calibration confidence assessment formula to calculate the comprehensive confidence score. The score measured before the fault point was 1.32, which is within the normal range.

[0083] The dynamic threshold baseline value of 1.50 was calculated from historical records of the same pattern. Combined with an upward trend correction of +0.15, a dynamic threshold of 1.65 was obtained. When the sensor was moved along the cable for measurement, after passing the fault point, the current value sharply decreased, the waveform became erratic, and the overall confidence score rose to 2.85, exceeding the dynamic threshold. The system determined the fault location by comprehensively considering polarity reversal characteristics, direction indication characteristics, and voltage gradient characteristics.

[0084] The fault point is mapped to the cable path coordinate reference system, generating a location report of "K0+085m, intelligent direction sensor mode, high confidence". The deviation of 1.20 between the current score and the dynamic threshold is used as the error feedback quantity to correct the historical normal feature vector template and self-calibration adjustment factor, enabling the system to complete a self-calibration optimization.

[0085] This embodiment achieves full automation of the process from signal purification, feature extraction, quantitative evaluation to adaptive updates, effectively improving the accuracy and consistency of fault location.

[0086] like Figure 2 The diagram shown is a functional block diagram of a cable sheath fault signal analysis system based on fully digital processing provided in an embodiment of the present invention.

[0087] The cable sheath fault signal analysis system 100 based on all-digital processing described in this invention can be installed in an electronic device. Depending on the functions implemented, the cable sheath fault signal analysis system 100 based on all-digital processing may include a purification and identification configuration module 101, a feature extraction module 102, a confidence assessment module 103, a fault analysis module 104, and a location update module 105. The module described in this invention can also be called a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, and which are stored in the memory of the electronic device.

[0088] In this embodiment, the functions of each module / unit are as follows: The purification identification and configuration module 101 is used to perform band-stop timing differential joint purification on the original signal set collected by the receiver to obtain the pre-processed signal set of the cable sheath under test, automatically identify the type of sensor connected, and dynamically configure the working mode of the receiver. The feature extraction module 102 is used to match the corresponding fault feature analysis strategy in the preset strategy library based on the working mode, perform three-dimensional coupling analysis on the preprocessed signal set, and obtain the multi-dimensional fault feature vector of the cable sheath to be tested. The confidence assessment module 103 is used to calculate the comprehensive confidence score of the cable sheath under test based on the multidimensional fault feature vector and through a self-calibrating confidence assessment formula. The fault analysis module 104 is used to calculate a dynamic threshold based on the historical composite error dataset of the cable sheath under test using the self-calibration confidence evaluation formula. When the comprehensive confidence score exceeds the dynamic threshold, the fault location of the cable sheath under test is analyzed. The positioning update module 105 is used to generate a positioning result report based on the fault location, working mode and comprehensive confidence score, and dynamically update the preset strategy library and the self-calibration confidence evaluation formula.

[0089] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0090] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0091] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0092] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0093] The embodiments of this application can acquire and process relevant data based on an artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0094] Finally, 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 it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for analyzing cable sheath fault signals based on fully digital processing, characterized in that, The method includes: S1: Perform band-stop timing differential joint purification on the raw signal set acquired by the receiver to obtain a preprocessed signal set of the cable sheath under test, automatically identify the type of the connected sensor, and dynamically configure the operating mode of the receiver, including: Acquire the raw signal set collected by the receiver, the raw signal set including multiple sampling points arranged sequentially along the time axis; A band-stop filter bank is applied to the original signal set to suppress the power frequency fundamental wave and integer harmonic interference components, resulting in a filtered signal set. The filtered signal set is then subjected to time-series differential processing, and the DC drift component is eliminated by utilizing the amplitude difference between adjacent sampling points, resulting in a preprocessed dataset. Based on the periodic fluctuation characteristics and amplitude envelope characteristics in the preprocessed dataset, the sensor type is automatically identified and a sensor type identifier is obtained by matching them with a preset sensor feature library. Based on the sensor type identifier, the operating mode parameter set corresponding to the sensor type identifier is matched, and the operating mode parameter set is written into the signal processing channel of the receiver to complete the configuration of the receiver's operating mode. S2: Based on the aforementioned working mode, the corresponding fault feature analysis strategy in the preset strategy library is matched, and the preprocessed signal set is subjected to three-dimensional coupled analysis to obtain the multi-dimensional fault feature vector of the cable sheath under test, including: In the preset strategy library, fault feature analysis strategies corresponding to the working mode are retrieved. The fault feature analysis strategies include time-domain feature types, frequency-domain feature types, and mode feature types. Based on the fault feature analysis strategy, three-dimensional feature extraction is performed on the preprocessed signal set to obtain the time-domain feature components, frequency-domain feature components, and mode feature components of the cable layer under test. The feature vectors of the three feature components are concatenated to obtain the multidimensional fault feature vector of the cable sheath under test. S3: Based on the multidimensional fault feature vector, the comprehensive confidence score of the cable sheath under test is calculated using the self-calibration confidence evaluation formula; S4: Based on the historical composite error dataset of the cable sheath to be tested, calculate the dynamic threshold using the self-calibration confidence evaluation formula. When the comprehensive confidence score exceeds the dynamic threshold, analyze the location of the fault point in the cable sheath to be tested. S5: Generate a location result report based on the fault location, working mode and comprehensive confidence score, and dynamically update the preset strategy library and the self-calibration confidence evaluation formula.

2. The cable sheath fault signal analysis method based on all-digital processing as described in claim 1, characterized in that, The comprehensive confidence score of the cable sheath under test is calculated based on the multidimensional fault feature vector using a self-calibrated confidence assessment formula, including: The multidimensional fault feature vector is normalized, and the processing result is compared with the historical normal feature vector template corresponding to the working mode to obtain the time domain feature deviation component, frequency domain feature deviation component and mode feature deviation component. Based on the multi-dimensional weighting coefficients and the feature deviation components in the fault feature analysis strategy, the comprehensive confidence score of the cable sheath under test is calculated using the self-calibration confidence evaluation formula.

3. The cable sheath fault signal analysis method based on all-digital processing as described in claim 2, characterized in that, The calibration confidence assessment formula is expressed as follows: ； In the formula, This represents the overall confidence score of the cable sheath under test. This represents the index number corresponding to the characteristic deviation component. , This indicates the number of characteristic deviation components of the cable sheath under test. Indicates the first layer of the cable sheath under test Each characteristic deviation component Indicates the first The dimension weight coefficients corresponding to the feature deviation components. This represents the historical error fluctuation value of the cable sheath under test. This represents the self-calibration adjustment factor of the cable sheath under test.

4. The cable sheath fault signal analysis method based on all-digital processing as described in claim 1, characterized in that, Based on the historical composite error dataset of the cable sheath under test, a dynamic threshold is calculated using the self-calibration confidence assessment formula. When the comprehensive confidence score exceeds the dynamic threshold, the location of the fault point in the cable sheath under test is analyzed, including: Obtain the historical composite error dataset of the cable sheath to be tested, and filter the historical evaluation records associated with the current working mode from the historical composite error dataset to obtain a subset of historical records of the same mode; Based on the historical confidence scores corresponding to each historical evaluation in the same pattern historical record subset, the dynamic threshold benchmark value of the cable sheath to be tested is determined. Based on the historical composite error dataset, the confidence trend feature closest to the current evaluation time is extracted, and the trend correction amount corresponding to the confidence trend feature is determined; The dynamic threshold reference value and the trend correction amount are corrected and superimposed to obtain the dynamic threshold of the cable sheath to be tested. When the overall confidence score exceeds the dynamic threshold, the location of the fault point in the sheath of the cable under test is analyzed.

5. The cable sheath fault signal analysis method based on all-digital processing as described in claim 4, characterized in that, When the overall confidence score exceeds the dynamic threshold, the location of the fault point in the cable sheath under test is analyzed, including: The comprehensive confidence score is compared with the dynamic threshold. When the comprehensive confidence score is greater than the dynamic threshold, it is determined that the sheath of the cable under test has a fault, and the corresponding multidimensional fault feature vector is obtained. Based on the polarity reversal feature, direction indication feature and voltage gradient feature in the multidimensional fault feature vector, the multidimensional fault orientation feature of the fault point in the sheath of the cable under test relative to the current measurement position is determined. The location of the fault point in the cable sheath under test is obtained by comprehensively weighting the orientation in the multidimensional fault orientation feature.

6. The cable sheath fault signal analysis method based on all-digital processing as described in claim 1, characterized in that, The process of generating a location result report based on the fault location, operating mode, and comprehensive confidence score, and dynamically updating the preset strategy library and the self-calibration confidence assessment formula, includes: The location of the fault point, the working mode, and the comprehensive confidence score are obtained. The location of the fault point is mapped to a preset cable path coordinate reference system to obtain the relative position coordinates. Based on the comparison between the comprehensive confidence score and the preset confidence level threshold, the confidence level corresponding to the comprehensive confidence score is obtained; Based on the relative position coordinates, the mode identifier corresponding to the working mode, and the confidence level, a positioning result report of the cable sheath to be tested is generated; The preset strategy library and the self-calibration confidence evaluation formula are dynamically updated.

7. The cable sheath fault signal analysis method based on all-digital processing as described in claim 6, characterized in that, The dynamic updating of the preset strategy library and the self-calibration confidence evaluation formula includes: The deviation between the comprehensive confidence score and the dynamic threshold is used as the error feedback amount of the cable sheath under test. Based on the error feedback amount, the historical normal feature vector template corresponding to the working mode is retrieved from the preset strategy library and benchmarked. The correction result replaces the original historical normal feature vector template in the preset strategy library. Based on the error feedback, the self-calibration adjustment factor in the self-calibration confidence assessment formula is corrected, and the updated self-calibration adjustment factor is written into the self-calibration confidence assessment formula.

8. A cable sheath fault signal analysis system based on fully digital processing, characterized in that, To implement the cable sheath fault signal analysis method based on all-digital processing as described in claim 1, the system includes a purification identification configuration module, a feature extraction module, a confidence assessment module, a fault analysis module, and a location update module, wherein: The purification and identification configuration module is used to perform band-stop timing differential joint purification on the raw signal set collected by the receiver to obtain a pre-processed signal set of the cable sheath under test, automatically identify the type of the connected sensor, and dynamically configure the operating mode of the receiver, including: Acquire the raw signal set collected by the receiver, the raw signal set including multiple sampling points arranged sequentially along the time axis; A band-stop filter bank is applied to the original signal set to suppress the power frequency fundamental wave and integer harmonic interference components, resulting in a filtered signal set. The filtered signal set is then subjected to time-series differential processing, and the DC drift component is eliminated by utilizing the amplitude difference between adjacent sampling points, resulting in a preprocessed dataset. Based on the periodic fluctuation characteristics and amplitude envelope characteristics in the preprocessed dataset, the sensor type is automatically identified and a sensor type identifier is obtained by matching them with a preset sensor feature library. Based on the sensor type identifier, the operating mode parameter set corresponding to the sensor type identifier is matched, and the operating mode parameter set is written into the signal processing channel of the receiver to complete the configuration of the receiver's operating mode. The feature extraction module is used to match the corresponding fault feature analysis strategy in the preset strategy library based on the working mode, and perform three-dimensional coupled analysis on the preprocessed signal set to obtain the multi-dimensional fault feature vector of the cable sheath under test, including: In the preset strategy library, fault feature analysis strategies corresponding to the working mode are retrieved. The fault feature analysis strategies include time-domain feature types, frequency-domain feature types, and mode feature types. Based on the fault feature analysis strategy, three-dimensional feature extraction is performed on the preprocessed signal set to obtain the time-domain feature components, frequency-domain feature components, and mode feature components of the cable layer under test. The feature vectors of the three feature components are concatenated to obtain the multidimensional fault feature vector of the cable sheath under test. The confidence assessment module is used to calculate the comprehensive confidence score of the cable sheath under test based on the multidimensional fault feature vector using a self-calibrating confidence assessment formula. The fault analysis module is used to calculate a dynamic threshold based on the historical composite error dataset of the cable sheath under test using the self-calibration confidence evaluation formula. When the comprehensive confidence score exceeds the dynamic threshold, the fault location of the cable sheath under test is analyzed. The location update module is used to generate a location result report based on the fault location, working mode and comprehensive confidence score, and dynamically update the preset strategy library and the self-calibration confidence evaluation formula.

Images