A big data early warning system for cable faults

Abstract

The present invention belongs to the technical field of cable fault early warning, and specifically provides a big data early warning system for cable faults. The system mainly includes: an operating data acquisition module: acquiring the operating data of the cable, where the operating data includes the cable length, the conductor currents of phase A and phase C, the cable surface temperature, and the ambient temperature; a characteristic factor generation module: generating an eddy current index and a sheath circulating current correction factor according to the operating data of the cable; an insulation characteristic analysis module: processing the eddy current index and the sheath circulating current correction factor by using a cable insulation characteristic analysis model; and an early warning result generation module: generating an early warning level and a potential hazard location according to the partial discharge intensity and the air gap defect index. The present invention can achieve accurate decoupling and quantification of the insulation deterioration degree and the potential hazard location, improve the sensitivity of early defect detection and the anti-interference ability, support the real-time positioning and hierarchical early warning of potential hazards during the entire life cycle of the cable, and provide multi-dimensional decision-making basis for the active protection of the power grid.

Description

Technical Field

[0001] The present invention belongs to the technical field of cable fault early warning, and in particular relates to a cable fault big data early warning system. Background Art

[0002] During cable operation, the electromagnetic-thermal coupling effect induces distortion of conductor current distribution and microscopic defects in the insulation layer, resulting in irreversible degradation accumulation. The accumulation of local discharge energy and distortion of the air gap electric field accelerate insulation aging, significantly shortening the cable life and causing hidden faults.

[0003] Traditional monitoring methods rely only on a single temperature or current threshold, which makes it difficult to effectively distinguish environmental interference from early cable defects in practical applications. In addition, for short cables, thermal hysteresis is more obvious, and long cables are easily disturbed by noise. The combination of the two can easily lead to misjudgment of monitoring results. Existing cable early warning methods are difficult to adapt to differences in parameters such as cable length and cross-sectional area, and cannot adapt well to these parameter changes. They often tend to ignore the nonlinear relationship between current phase fluctuations and temperature conduction, making it difficult to accurately quantify the degree of insulation degradation of the cable and determine the location of hidden dangers, ultimately resulting in delayed alarms and unclear positioning. Summary of the invention

[0004] The present invention provides a cable fault big data early warning system, which effectively solves the problem that the cable early warning method in the prior art is difficult to adapt to the differences in parameters such as cable length and cross-sectional area, and cannot adapt well to these parameter changes. It is often easy to ignore the nonlinear correlation between current phase fluctuation and temperature conduction, which makes it difficult to accurately quantify the degree of insulation degradation of the cable and determine the location of hidden dangers. The system achieves accurate decoupling and quantification of the degree of insulation degradation and the location of hidden dangers, improves the sensitivity and anti-interference ability of early defect detection, supports real-time positioning and graded early warning of hidden dangers throughout the life cycle of the cable, and provides a multi-dimensional decision-making basis for active protection of the power grid.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] The present invention provides a cable fault big data early warning system, comprising: an operation data acquisition module: acquiring the operation data of the cable, the operation data including the cable length, A-phase and C-phase conductor currents, and the cable skin temperature and ambient temperature.

[0007] Characteristic factor generation module: Generates eddy current index and sheath circulation correction factor based on the cable operation data.

[0008] Insulation Feature Analysis Module: It uses a cable insulation feature analysis model to process the eddy current index and sheath circulating current correction factor. The cable insulation feature analysis model is based on the Wide&Deep architecture. On the Wide side, the eddy current index is input and connected to the cable length to generate linear features. On the Deep side, a fully connected network is used for processing, and the number of neurons is determined according to the cable length, and the partial discharge intensity and air gap defect index are output.

[0009] Early Warning Result Generation Module: Generate the early warning level and potential hazard location according to the partial discharge intensity and air gap defect index.

[0010] Furthermore, generate the eddy current index according to the operating data of the cable, including: performing skin effect compensation on the conductor currents of phase A and phase C to generate the compensated first current value and second current value respectively; taking a specified power frequency period as the window, calculating the sample entropy value of the phase difference between the first current value and the second current value; calculating the conductor eddy current index according to the first current value and the second current value.

[0011] Furthermore, generate the sheath circulating current correction factor according to the operating data of the cable, including: setting a cross-sectional threshold; obtaining the conductor cross-sectional area of the cable; when the conductor cross-sectional area is greater than the cross-sectional threshold, generating the sheath circulating current correction factor according to the conductor cross-sectional area and the sample entropy value.

[0012] Furthermore, generating the sheath circulating current correction factor according to the conductor cross-sectional area and the sample entropy value includes: performing adaptive wavelet packet decomposition on the first current value and the second current value, and extracting the high-frequency band energy ratio of a specified number; setting an entropy threshold, if the sample entropy value is greater than the entropy threshold, activating the high-frequency energy strengthening flag; setting a basic correction factor, if the high-frequency strengthening flag is activated, using the product of the strengthening coefficient and the basic correction factor as the enhanced correction factor; setting the first period window and the second period window, if the high-frequency energy strengthening flag is activated, performing time-domain smoothing on the basic correction factor or the enhanced correction factor using the first period window mean value, otherwise performing time-domain smoothing on the basic correction factor or the enhanced correction factor using the second period window mean value; where the second period window mean value is greater than the first period window mean value.

[0013] Furthermore, generating the early warning level and potential hazard location according to the partial discharge intensity and air gap defect index includes: performing working condition coding on the partial discharge intensity and air gap defect index to generate coded data; generating the true temperature rise and temperature rise acceleration characteristics according to the cable surface temperature; using a fault decision model to process the true temperature rise, temperature rise acceleration characteristics and coded data. The fault decision model is based on the gradient boosting tree architecture, and determines the splitting features according to the coded data to construct a decision tree, and outputs the early warning level and potential hazard location.

[0014] Further, perform condition coding on the partial discharge intensity and the air gap defect index to generate coded data, including: obtaining the operation years of the cable, setting a first trigger threshold according to the operation years, and generating a first coding result when the partial discharge intensity is greater than the first trigger threshold; setting a second trigger threshold and a third trigger threshold, and generating a second coding result when the air gap defect index is greater than the second trigger threshold and the sample entropy value is greater than the third trigger threshold.

[0015] Further, generate the true temperature rise according to the cable surface temperature, including: obtaining the cable insulation layer thickness; performing thermal resistance compensation on the cable surface temperature according to the cable surface temperature, the ambient temperature, and the cable insulation layer thickness to generate the true temperature rise.

[0016] Further, generate the temperature rise acceleration feature, including: performing piecewise polynomial fitting on the true temperature rise to generate a baseline temperature rise curve; marking the difference between the true temperature rise and the baseline temperature rise curve as a dynamic residual sequence; extracting the extreme point distribution density of the dynamic residual sequence; setting a density threshold; when the extreme point distribution density is greater than the density threshold: performing empirical mode decomposition on the dynamic residual sequence, and extracting the energy proportion of the first specified order intrinsic mode component; calculating the permutation entropy value of the dynamic residual sequence, setting an entropy threshold, and activating an abnormal transient mark when the permutation entropy value is greater than the entropy threshold; fusing the energy proportion and the permutation entropy value according to the abnormal transient mark to generate the temperature rise acceleration feature.

[0017] Further, fusing the energy proportion and the permutation entropy value according to the abnormal transient mark to generate the temperature rise acceleration feature, including: when the abnormal transient mark is activated, performing a multiplication operation on the energy proportion of the specified order intrinsic mode component and the permutation entropy value to generate the temperature rise acceleration feature, otherwise taking the cumulative sum of the energy proportion as the temperature rise acceleration feature.

[0018] Further, determine the splitting feature according to the coded data to construct a decision tree, including: when the first coding result is generated, constructing a decision tree with the temperature rise acceleration feature as the preferred splitting feature; when the second coding result is generated, using the product of the true temperature rise and the temperature rise acceleration feature as the splitting basis.

[0019] Advantages of the present invention:

[0020] The present invention constructs an insulation feature analysis model based on the Wide&Deep architecture. On the Wide side, the eddy current index and the cable length are fused, and on the Deep side, a fully connected network processes the sheath circulating current correction factor to dynamically adapt to the cable parameters, effectively solving the problems in the prior art that the cable warning method is difficult to adapt to the parameter differences such as the cable length and cross-sectional area, cannot well adapt to these parameter changes, and often easily ignores the non-linear correlation between the current phase fluctuation and the temperature conduction, resulting in difficulty in accurately quantifying the insulation deterioration degree of the cable and determining the hidden danger location. It realizes the accurate decoupling and quantification of the insulation deterioration degree and the hidden danger location, improves the sensitivity of early defect detection and the anti-interference ability, supports the real-time location and hierarchical warning of hidden dangers during the whole life cycle of the cable, and provides a multi-dimensional decision-making basis for the active protection of the power grid.

[0021] Other features and advantages of the present invention will be described in the following specification, and part of them will become obvious from the specification or be understood by implementing the present invention. The objectives and other advantages of the present invention can be realized and obtained by the structures pointed out in the specification and the drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0023] Figure 1 The figure shows a schematic diagram of a cable fault big data warning system of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0024] To solve the problems proposed in the background art, the present invention constructs an insulation feature analysis model based on the Wide&Deep architecture. On the Wide side, the eddy current index and the cable length are fused, and on the Deep side, a fully connected network processes the sheath circulating current correction factor to dynamically adapt to the cable parameters, realizes the accurate decoupling and quantification of the insulation deterioration degree and the hidden danger location, improves the sensitivity of early defect detection and the anti-interference ability, supports the real-time location and hierarchical warning of hidden dangers during the whole life cycle of the cable, and provides a multi-dimensional decision-making basis for the active protection of the power grid.

[0025] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0026] In some embodiments, as Figure 1 shown, the present invention provides a cable fault big data early warning system, including: an operation data acquisition module, a characteristic factor generation module, an insulation characteristic analysis module, and an early warning result generation module.

[0027] The method for the operation of the cable fault big data early warning system includes: the method for the operation of the operation data acquisition module, the method for the operation of the characteristic factor generation module, the method for the operation of the insulation characteristic analysis module, and the method for the operation of the early warning result generation module.

[0028] The method for the operation of the operation data acquisition module is: S100. Acquire the operation data of the cable, where the operation data includes the cable length, the conductor currents of phase A and phase C, and the cable surface temperature and the ambient temperature.

[0029] The method for the operation of the characteristic factor generation module is: S200. Generate an eddy current index and a sheath circulating current correction factor according to the operation data of the cable.

[0030] The method for the operation of the insulation characteristic analysis module is: S300. Process the eddy current index and the sheath circulating current correction factor by using a cable insulation characteristic analysis model. The cable insulation characteristic analysis model is based on the Wide&Deep architecture. At the Wide side, the eddy current index is input and connected to the cable length to generate linear features. At the Deep side, a fully connected network is used for processing, and the number of neurons is determined according to the cable length, and the partial discharge intensity and the air gap defect index are output.

[0031] The method for the operation of the early warning result generation module is: S400. Generate an early warning level and a potential hazard location according to the partial discharge intensity and the air gap defect index.

[0032] In some embodiments, generating an eddy current index according to the operation data of the cable includes:

[0033] Sa210. Perform skin effect compensation on the conductor currents of phase A and phase C, and respectively generate a compensated first current value and a compensated second current value.

[0034] The cable conductor has a skin effect under high-frequency current, resulting in an underestimation of the actual effective values of the currents of phase A and phase C. A frequency-domain compensation algorithm can be used to correct the currents of phase A and phase C.

[0035] Original phase A current I A and phase C current I c , and the sampling frequency can be 10 kHz.

[0036] Calculate the skin depth δ based on the conductor material and cross-sectional shape. For example, the formula for calculating the skin depth of a copper conductor at 50 Hz power frequency is: where ρ represents the conductor resistivity, ω represents the angular frequency (2π × 50 Hz), and μ represents the magnetic permeability.

[0037] Perform frequency-domain compensation on the phase A and phase C conductor currents to generate the compensated first current value and second current value. The compensation formula can be: where I ' A and I ' C represent the compensated first current value and second current value respectively, and d represents the conductor diameter of the cable.

[0038] Sa220. Taking the specified power-frequency period as the window, calculate the sample entropy value of the phase difference between the first current value and the second current value.

[0039] The sample entropy value can quantify the phase fluctuation characteristics of the phase A and phase C currents.

[0040] Taking 20 ms as the time window, extract the phase difference sequence Δφ(t) of the compensated first current value I ' A and the second current value I ' C .

[0041] Calculate the sample entropy value S entropy of the phase difference sequence, which is used to characterize the complexity of the phase difference. The specific method is:

[0042] Set the embedding dimension m = 2 and the tolerance threshold r = 0.2 × std(Δφ), where std(Δφ) represents the standard deviation of the calculated phase difference sequence Δφ(t).

[0043] Statistical the vector matching probabilities of lengths m and m + 1 in the phase difference sequence Δφ(t). The formula is: where S entropy represents the sample entropy value, B m (r) and B m+1 (r) represent the number of matches of vectors of lengths m and m + 1 within the tolerance threshold r in the phase difference sequence respectively.

[0044] Sa230. Calculate the conductor eddy current index based on the first current value and the second current value.

[0045] Calculate the first current value I' A and the ratio of the root mean square values of the second current value I ' C : wherein, RMS(I' A ) and RMS(I' C ) respectively represent the root mean square of the first current value I' A and the second current value I' C .

[0046] According to the sample entropy value S entropy and the ratio R of the root mean square of the current, a eddy current index is generated by linear weighting: Eeddy = k1×R + k2×S entropy ; wherein, k1 and k2 respectively represent the empirical weights of the ratio R of the root mean square of the current and the sample entropy value S entropy , such as 0.7 and 0.3 respectively, and can be determined by regression of historical fault data.

[0047] In some embodiments, generating a sheath circulating current correction factor according to the operating data of the cable includes:

[0048] Sb210. Set a cross-sectional threshold; obtain the conductor cross-sectional area of the cable.

[0049] Set the cross-sectional threshold according to the cable type (such as cross-linked polyethylene cable) and historical operating data. For example, for medium-voltage cables, the cross-sectional threshold is set to 300 mm 2 . The conductor cross-sectional area can be obtained through cable design parameters or real-time measurement.

[0050] Sb220. When the conductor cross-sectional area is greater than the cross-sectional threshold, generate a sheath circulating current correction factor according to the conductor cross-sectional area and the sample entropy value.

[0051] In some embodiments, generating a sheath circulating current correction factor according to the conductor cross-sectional area and the sample entropy value includes:

[0052] Sb221. Perform adaptive wavelet packet decomposition on the first current value and the second current value, and extract the high-frequency band energy ratio of a specified number.

[0053] The wavelet decomposition can use the db4 wavelet basis function.

[0054] The high-frequency band (6 - 8 sub-bands) energy ratio E high after the 3rd layer decomposition can be extracted, and the calculation formula is: wherein, E k represents the energy value of the kth sub-band.

[0055] Sb222. Set an entropy threshold, if the sample entropy value is greater than the entropy threshold, activate the high-frequency energy enhancement mark.

[0056] Collect the phase difference sample entropy values during the normal operation of the collection device, calculate their mean and standard deviation, and set the mean plus several times the standard deviation as the entropy threshold. For example, the entropy threshold S can be set. th = 1.2.

[0057] Sb223. Set the basic correction factor. If the high-frequency enhancement mark is activated, use the product of the enhancement coefficient and the basic correction factor as the enhanced correction factor.

[0058] Set the basic correction factor C base The calculation formula is: C base = 0.5×S + 0.5×S entropy ; where S represents the conductor cross-sectional area, and the coefficient 0.5 represents the empirical weight.

[0059] If the high-frequency energy enhancement mark is activated, then use the enhancement coefficient γ = 1.2 to generate the enhanced correction factor C enhanced : C enhanced = γ×C base The enhancement coefficient γ can be obtained by fitting, theoretical derivation, etc.

[0060] Sb224. Set the first cycle window and the second cycle window. If the high-frequency energy enhancement mark is activated, perform time-domain smoothing on the basic correction factor or the enhanced correction factor using the mean value of the first cycle window; otherwise, perform time-domain smoothing on the basic correction factor or the enhanced correction factor using the mean value of the second cycle window; among them, the mean value of the second cycle window is greater than the mean value of the first cycle window.

[0061] Set the first cycle window as a short window, such as 10 minutes, and set the second cycle window as a long window, such as 60 minutes. When the mark is activated, perform moving average smoothing on the basic correction factor C base or the enhanced correction factor C enhanced using the moving average value of the first cycle window. When the mark is not activated, perform moving average smoothing using the moving average value of the second cycle window.

[0062] The short window can quickly respond to high-frequency abnormal fluctuations, while the long window ensures stability under steady-state working conditions.

[0063] By accurately separating high-frequency interference components through adaptive wavelet packet decomposition and dynamically adjusting the correction factor in combination with the sample entropy value, it is possible to solve the misjudgment problem of sheath circulating current caused by changes in conductor cross-sectional area and high-frequency noise.

[0064] For example, when a large cross-section cable encounters arc discharge, the abnormal characteristics of the sheath circulating current are amplified in time through the enhanced correction factor, improving the accuracy of hidden danger location.

[0065] In some embodiments, S300 is specifically implemented as follows:

[0066] The inputs on the Wide side are the eddy current index Eeddy and the cable length L.

[0067] Perform a linear combination of the eddy current index and the cable length to generate the composite feature F. wide . The formula is: F wide = ω1 × Eeddy + ω2 × L + b; where, ω1 and ω2 represent weight coefficients, which can be obtained through gradient descent, and b represents the bias term.

[0068] The inputs on the Deep side are the sheath circulating current correction factor, and the number of neurons in the fully connected network structure is dynamically adjusted according to the cable length L.

[0069] For example, when L ≤ 1 km, a 3-layer fully connected network is adopted, and the number of neurons is 16, 16, 8; when L > 1 km, it is extended to a 4-layer fully connected network, and the number of neurons is 32, 32, 16, 8.

[0070] Use the ReLU activation function in the hidden layer and the Sigmoid activation function in the output layer.

[0071] The output results of the cable insulation feature analysis model are the partial discharge intensity P discharge and the air gap defect index D. gap , the partial discharge intensity P discharge can range from 0 to 100 pC, and the larger the value, the more serious the insulation deterioration. The air gap defect index D gap can range from 0 to 1, and the closer the value is to 1, the higher the risk of air gap defects.

[0072] When training the cable insulation feature analysis model, the mean square error function is used for the partial discharge intensity, and the binary cross entropy is used for the air gap defect index. The total loss is the weighted sum of the partial discharge intensity and the air gap defect index.

[0073] For example: Loss = 0.7 × Loss1 + 0.3 × Loss2; where, Loss represents the total loss function, Loss1 represents the mean square error loss function, Loss2 represents the binary cross entropy loss function, and 0.7 and 0.3 respectively represent the weights of the mean square error loss function and the binary cross entropy loss function, which can be determined by the method of grid search combined with validation set evaluation.

[0074] The Adam optimizer can be selected, with an initial learning rate of 0.001, decaying by 10% every 50 epochs, which can effectively adjust the model parameters and avoid too large a step size in the later stage; the batch size is set to 64, which can balance the calculation efficiency and memory occupancy and make the training efficient and stable; it is planned to train for 300 epochs, and the training will terminate if the validation set loss does not decrease for 10 consecutive epochs, which can limit the number of learning times and prevent overfitting.

[0075] In some embodiments, generating a warning level and a potential hazard location based on the partial discharge intensity and the air gap defect index includes:

[0076] S410. Perform condition encoding on the partial discharge intensity and the air gap defect index to generate encoded data; generate the true temperature rise and the temperature rise acceleration characteristics based on the cable surface temperature.

[0077] S420. Process the true temperature rise, the temperature rise acceleration characteristics, and the encoded data using a fault decision model. The fault decision model is based on the gradient boosting tree and constructs a decision tree by determining the splitting characteristics according to the encoded data, and outputs the warning level and the potential hazard location.

[0078] In some embodiments, performing condition encoding on the partial discharge intensity and the air gap defect index to generate encoded data includes:

[0079] S411. Obtain the operation years of the cable, set a first trigger threshold according to the operation years, and generate a first encoding result when the partial discharge intensity is greater than the first trigger threshold.

[0080] Set the first trigger threshold T1 = 10×Y. Exemplarily, when the operation years Y = 8 years, the first trigger threshold T1 = 80 pC.

[0081] If the partial discharge intensity P discharge = 90 pC, then P discharge > T1, trigger the first encoding result Code1, such as the first encoding result Code1 = 1, otherwise Code1 = 0.

[0082] S412. Set a second trigger threshold and a third trigger threshold, and generate a second encoding result when the air gap defect index is greater than the second trigger threshold and the sample entropy value is greater than the third trigger threshold.

[0083] The second trigger threshold T2 = 0.8 and the third trigger threshold T3 = 1.5 can be set.

[0084] When the air gap defect index D gap > T2 and the sample entropy value S entropy > T3, generate the second encoding result Code2, such as Code2 = 1, otherwise Code2 = 0.

[0085] In some embodiments, generating the true temperature rise based on the cable surface temperature includes:

[0086] Sa413. Obtain the thickness of the cable insulation layer.

[0087] The thickness of the cable insulation layer H can be obtained through cable design parameters or real-time measurement.

[0088] Sa414. Compensate for the thermal resistance of the cable surface temperature based on the cable surface temperature, ambient temperature, and cable insulation layer thickness to generate the true temperature rise.

[0089] Calculate the original temperature difference ΔTraw, ΔTraw = T surface -T ambient ; where T surface represents the cable surface temperature, and T ambient represents the ambient temperature.

[0090] Compensate for the original temperature difference ΔTraw based on the insulation layer thickness H to generate the true temperature rise ΔT: For every 10 mm increase in the insulation layer thickness, the true temperature rise compensation coefficient doubles to correct the blocking effect of the heat conduction path.

[0091] Using the thermal resistance compensation formula with the insulation layer thickness as the key parameter makes the true temperature rise closer to the actual heating state of the conductor.

[0092] In some embodiments, generate the temperature rise acceleration feature, including:

[0093] Sb413. Perform piecewise polynomial fitting on the true temperature rise to generate the baseline temperature rise curve.

[0094] The piecewise rule can be a 10-minute time window, and quadratic polynomial fitting is used within each window.

[0095] For example, the temperature rise data of a certain cable from 10:00 to 10:10 is: 50°C, 52°C, 55°C, 58°C, 60°C, then the fitting curve ΔT base (t) is: ΔT base (t) = -0.1t 2 + 2.5t + 50; where t represents the time variable.

[0096] Piecewise polynomial fitting can eliminate baseline drift, and the dynamic residual sequence can highlight transient fluctuations.

[0097] Sb414. Mark the difference between the true temperature rise and the baseline temperature rise curve as the dynamic residual sequence; extract the extreme point distribution density of the dynamic residual sequence.

[0098] Calculate the dynamic residual sequence R(t), R(t) = ΔT(t) - ΔTraw(t); ΔT(t) and ΔTraw(t) respectively represent the true temperature rise ΔT and the fitting curve ΔT changing with time t base .

[0099] Extract the extreme point distribution density ρ of the dynamic residual sequence R(t), that is, the number of maximum / minimum values per unit time.

[0100] Exemplarily, if 8 extreme points are detected within 5 minutes, the distribution density of extreme points

[0101] The dual criteria of extreme point density and permutation entropy can accurately identify abnormal events. For example, when ρ > 5 per minute and PE > 0.7, it is determined as arc discharge interference.

[0102] Sb415. Set the density threshold; when the distribution density of extreme points is greater than the density threshold: perform empirical mode decomposition on the dynamic residual sequence, and extract the energy proportion of the specified order of the first few intrinsic mode components; calculate the permutation entropy value of the dynamic residual sequence, set the entropy threshold, and activate the abnormal transient mark when the permutation entropy value is greater than the entropy threshold.

[0103] Set the density threshold ρ th = 5 per minute, when ρ > ρ th , perform the following operations:

[0104] Decompose the dynamic residual sequence R(t) into multiple intrinsic mode components, and extract the specified order of components, such as the 3rd order.

[0105] If it is the 3rd order, calculate the energy proportions E1, E2, and E3 of IMF1 - IMF3.

[0106] Set the embedding dimension m = 3, calculate the permutation entropy value PE of the dynamic residual sequence R(t), set the entropy threshold PE th , if PE > PE th , then activate the abnormal transient mark, such as Flag = 1, otherwise Flag = 0.

[0107] Sb416. Generate the temperature rise acceleration feature by fusing the energy proportion and the permutation entropy value according to the abnormal transient mark.

[0108] In some embodiments, generating the temperature rise acceleration feature by fusing the energy proportion and the permutation entropy value according to the abnormal transient mark includes:

[0109] When the abnormal transient mark is activated, select the energy proportion of the specified order of the intrinsic mode components and the permutation entropy value for multiplication operation to generate the temperature rise acceleration feature, otherwise use the cumulative sum of the energy proportions as the temperature rise acceleration feature.

[0110] Exemplarily, when Flag = 1, select the product of the energy proportion E2 of the 2nd order intrinsic mode component and the permutation entropy value PE as the temperature rise acceleration feature A temp : A temp = E2 × PE. When Flag = 0, accumulate the energy proportions of IMF1 - IMF3 as the temperature rise acceleration feature.

[0111] The fusion of the energy proportion and the entropy value can quantify the severity of the temperature rise change.

[0112] In some embodiments, determining a splitting feature according to encoded data to construct a decision tree includes:

[0113] When generating the first encoded result, construct a decision tree with the temperature rise acceleration feature as the priority splitting feature.

[0114] Exemplarily, if the temperature rise acceleration feature A temp > 0.3, it is determined as a high-risk branch, and after splitting, it enters the sub-node with a warning level of 3 or 4.

[0115] The first encoded result focuses on transient temperature rise fluctuations and adapts to the insulation degradation mode dominated by partial discharge.

[0116] When generating the second encoded result, use the product of the true temperature rise and the temperature rise acceleration feature as the splitting basis.

[0117] Exemplarily, if ΔT × A temp > 30, it is determined as a composite fault branch, and after splitting, it enters the sub-node where the hidden danger location is more than 20% away from the starting end.

[0118] The second encoded result synthesizes the steady-state temperature rise and transient fluctuations and adapts to the coupled fault of air gap defects and circulation disturbances.

[0119] The output warning level can be the risk levels of 1 to 4 manually marked, and the hidden danger location is marked as the percentage from the starting end of the cable. For example, 25% means at the quarter position of the total length.

[0120] The training process is carried out in batches. Each batch inputs 64 groups of data including true temperature rise, temperature rise acceleration, and encoded data, and outputs the warning level and the hidden danger location.

[0121] In terms of loss calculation, the classification loss is for the warning level. For example, the error weight of the high-risk level (levels 3 to 4) is increased by 3 times to let the model give priority to learning serious faults; the regression loss is for the hidden danger location, and the absolute difference between the prediction and the true location is calculated.

[0122] Through multiple experiments, compare the comprehensive performance indicators of the model on the validation set for warning level prediction and hidden danger location prediction under different weight ratios, and select the ratio with the optimal comprehensive performance. For example, the total loss is optimized by weighted summation with the classification loss accounting for 70% and the regression loss accounting for 30%.

[0123] For dynamic parameter update, use the Adam optimizer with an initial learning rate of 0.001, which decays by 10% every 50 rounds. Set an early stopping mechanism. If the validation set loss does not decrease for 10 consecutive rounds, stop training to prevent overfitting.

[0124] Through dynamic splitting rules and combined loss optimization, the fault decision-making model can quickly adapt to different fault modes, accurately output the risk level and potential hazard location, and guide the maintenance personnel to conduct efficient troubleshooting.

[0125] In some embodiments, the present invention provides a device, which includes a memory and a processor; the memory is used to store a computer program; the processor is used to implement the steps of a method for running a cable fault big data warning system when executing the computer program.

[0126] In some embodiments, the present invention provides a readable storage medium, in which computer program instructions are stored. When the computer program instructions are read and run by a processor, the steps of a method for running a cable fault big data warning system are executed.

[0127] Among them, any reference to a memory, storage, database, or other medium used in the embodiments provided by the present invention may include non-volatile and / or volatile memories. The non-volatile memory may include a read-only memory (ROM), a programmable ROM (PROM), an electrically programmable ROM (EPROM), an electrically erasable programmable ROM (EEPROM), or a flash memory. The volatile memory may include a random access memory (RAM) or an external cache memory.

[0128] It should be noted that in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising a..." does not exclude the existence of additional identical elements in the process, method, article or device including the element.

[0129] Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A big data early warning system for cable faults, characterized in that, Including: Operation data acquisition module: acquiring the operation data of the cable, where the operation data includes cable length, conductor currents of phase A and phase C, cable surface temperature, and ambient temperature; Characteristic factor generation module: generating an eddy current index and a sheath circulating current correction factor according to the operation data of the cable; Insulation characteristic analysis module: processing the eddy current index and the sheath circulating current correction factor by using a cable insulation characteristic analysis model. The cable insulation characteristic analysis model is based on the Wide&Deep architecture. On the Wide side, the eddy current index is input and connected with the cable length to generate linear features. On the Deep side, a fully connected network is used for processing, and the number of neurons is determined according to the cable length, and the partial discharge intensity and the air gap defect index are output; Early warning result generation module: generating an early warning level and a potential hazard location according to the partial discharge intensity and the air gap defect index; Among them, generating the eddy current index according to the operation data of the cable includes: Perform frequency-domain compensation on the conductor currents of phase A and phase C to generate the compensated first current value I' A and the second current value I ' C , and the compensation formula is: where I A and I c represent the original phase A current and phase C current respectively, d represents the conductor diameter of the cable, and δ represents the skin depth of the conductor; Taking the specified power frequency period as a window, calculate the sample entropy value S of the phase difference between the first current value and the second current value entropy ; Calculate the first current value I' A and the root mean square value ratio R of the second current value I ' C and generate the eddy current index Eeddy through linear weighting: Eeddy = k1×R + k2×S entropy ; where k1 and k2 respectively represent the empirical weights of the root mean square ratio R of the current and the sample entropy value S entropy ​ 2. The cable fault big data early warning system according to claim 1, characterized in that, Generating the sheath circulating current correction factor according to the operation data of the cable includes: Setting a cross-sectional threshold; acquiring the conductor cross-sectional area of the cable; When the conductor cross-sectional area is greater than the cross-sectional threshold, generating the sheath circulating current correction factor according to the conductor cross-sectional area and the sample entropy value.

3. The cable fault big data early warning system according to claim 1, characterized in that Generating the sheath circulating current correction factor according to the conductor cross-sectional area and the sample entropy value includes: Performing adaptive wavelet packet decomposition on the first current value and the second current value, and extracting the high-frequency band energy ratio of a specified number; Setting an entropy threshold. If the sample entropy value is greater than the entropy threshold, activating the high-frequency energy enhancement mark; Setting a basic correction factor. If the high-frequency enhancement mark is activated, using the product of the enhancement coefficient and the basic correction factor as the enhanced correction factor; Setting a first period window and a second period window. If the high-frequency energy enhancement mark is activated, performing time-domain smoothing on the basic correction factor or the enhanced correction factor by using the mean value of the first period window, otherwise performing time-domain smoothing on the basic correction factor or the enhanced correction factor by using the mean value of the second period window; where the mean value of the second period window is greater than the mean value of the first period window.

4. The cable fault big data early warning system according to claim 1, characterized in that, Generating the early warning level and the potential hazard location according to the partial discharge intensity and the air gap defect index includes: Performing condition coding on the partial discharge intensity and the air gap defect index to generate coded data; generating the true temperature rise and the temperature rise acceleration characteristics according to the cable surface temperature; Processing the true temperature rise, the temperature rise acceleration characteristics, and the coded data by using a fault decision model. The fault decision model is based on the gradient boosting tree architecture, and a decision tree is constructed according to the coded data to determine the splitting features, and the early warning level and the potential hazard location are output.

5. The cable fault big data early warning system according to claim 3, wherein, Performing condition coding on the partial discharge intensity and the air gap defect index to generate coded data includes: Acquiring the operation years of the cable, setting a first trigger threshold according to the operation years. When the partial discharge intensity is greater than the first trigger threshold, generating a first coding result; Setting a second trigger threshold and a third trigger threshold. When the air gap defect index is greater than the second trigger threshold and the sample entropy value is greater than the third trigger threshold, generating a second coding result.

6. The cable fault big data early warning system according to claim 3, characterized in that Generating the true temperature rise according to the cable surface temperature includes: Acquiring the cable insulation layer thickness; Performing thermal resistance compensation on the cable surface temperature according to the cable surface temperature, the ambient temperature, and the cable insulation layer thickness to generate the true temperature rise.

7. The cable fault big data early warning system according to claim 5, wherein Generating the temperature rise acceleration characteristics includes: Perform piecewise polynomial fitting on the true temperature rise to generate a baseline temperature rise curve; Mark the difference between the true temperature rise and the baseline temperature rise curve as a dynamic residual sequence; extract the extreme point distribution density of the dynamic residual sequence; Set a density threshold; when the extreme point distribution density is greater than the density threshold: Perform empirical mode decomposition on the dynamic residual sequence, extract the energy proportion of the specified order of intrinsic mode components in the front; calculate the permutation entropy value of the dynamic residual sequence, set an entropy threshold, and activate the abnormal transient mark when the permutation entropy value is greater than the entropy threshold; Generate a temperature rise acceleration feature by fusing the energy proportion and the permutation entropy value according to the abnormal transient mark.

8. The cable fault big data early warning system according to claim 7, characterized in that Generate a temperature rise acceleration feature by fusing the energy proportion and the permutation entropy value according to the abnormal transient mark, including: When the abnormal transient mark is activated, select the energy proportion and the permutation entropy value of the specified order of intrinsic mode components for multiplication operation to generate a temperature rise acceleration feature, otherwise accumulate the total sum of the energy proportion as the temperature rise acceleration feature.

9. The cable fault big data early warning system according to claim 4, characterized in that, Determine a splitting feature according to the encoded data to construct a decision tree, including: When generating the first encoded result, construct a decision tree with the temperature rise acceleration feature as the priority splitting feature; When generating the second encoded result, use the product of the true temperature rise and the temperature rise acceleration feature as the splitting basis.

Images