CN118130984B - Cable partial discharge fault real-time monitoring method based on data driving - Google Patents
Cable partial discharge fault real-time monitoring method based on data driving Download PDFInfo
- Publication number
- CN118130984B CN118130984B CN202410577172.9A CN202410577172A CN118130984B CN 118130984 B CN118130984 B CN 118130984B CN 202410577172 A CN202410577172 A CN 202410577172A CN 118130984 B CN118130984 B CN 118130984B
- Authority
- CN
- China
- Prior art keywords
- partial discharge
- fault
- signal
- data
- prediction model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000012544 monitoring process Methods 0.000 title claims abstract description 32
- 238000000034 method Methods 0.000 title claims abstract description 30
- 238000003745 diagnosis Methods 0.000 claims abstract description 21
- 238000007781 pre-processing Methods 0.000 claims abstract description 15
- 238000013480 data collection Methods 0.000 claims abstract description 10
- 238000010276 construction Methods 0.000 claims abstract description 9
- 230000000694 effects Effects 0.000 claims abstract description 7
- 238000013527 convolutional neural network Methods 0.000 claims description 32
- 238000004364 calculation method Methods 0.000 claims description 29
- 238000012549 training Methods 0.000 claims description 28
- 238000011156 evaluation Methods 0.000 claims description 19
- 238000004458 analytical method Methods 0.000 claims description 18
- 230000008859 change Effects 0.000 claims description 12
- 238000005516 engineering process Methods 0.000 claims description 12
- 238000005259 measurement Methods 0.000 claims description 12
- 238000012360 testing method Methods 0.000 claims description 11
- 230000007246 mechanism Effects 0.000 claims description 9
- 238000000605 extraction Methods 0.000 claims description 7
- 238000012795 verification Methods 0.000 claims description 7
- 230000007613 environmental effect Effects 0.000 claims description 6
- 238000012545 processing Methods 0.000 claims description 5
- 230000015556 catabolic process Effects 0.000 claims description 3
- 238000012937 correction Methods 0.000 claims description 3
- 238000006731 degradation reaction Methods 0.000 claims description 3
- 230000004927 fusion Effects 0.000 claims description 3
- 238000005457 optimization Methods 0.000 claims description 3
- 238000002360 preparation method Methods 0.000 claims description 3
- 230000001360 synchronised effect Effects 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 238000007599 discharging Methods 0.000 claims 1
- 238000001514 detection method Methods 0.000 abstract description 5
- 238000012423 maintenance Methods 0.000 description 9
- 230000008901 benefit Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000002265 prevention Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000032683 aging Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001066 destructive effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000012854 evaluation process Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 239000011810 insulating material Substances 0.000 description 1
- 230000004807 localization Effects 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 230000003449 preventive effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/12—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing
- G01R31/1227—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing of components, parts or materials
- G01R31/1263—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing of components, parts or materials of solid or fluid materials, e.g. insulation films, bulk material; of semiconductors or LV electronic components or parts; of cable, line or wire insulation
- G01R31/1272—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing of components, parts or materials of solid or fluid materials, e.g. insulation films, bulk material; of semiconductors or LV electronic components or parts; of cable, line or wire insulation of cable, line or wire insulation, e.g. using partial discharge measurements
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/12—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing
- G01R31/1209—Testing dielectric strength or breakdown voltage ; Testing or monitoring effectiveness or level of insulation, e.g. of a cable or of an apparatus, for example using partial discharge measurements; Electrostatic testing using acoustic measurements
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y04—INFORMATION OR COMMUNICATION TECHNOLOGIES HAVING AN IMPACT ON OTHER TECHNOLOGY AREAS
- Y04S—SYSTEMS INTEGRATING TECHNOLOGIES RELATED TO POWER NETWORK OPERATION, COMMUNICATION OR INFORMATION TECHNOLOGIES FOR IMPROVING THE ELECTRICAL POWER GENERATION, TRANSMISSION, DISTRIBUTION, MANAGEMENT OR USAGE, i.e. SMART GRIDS
- Y04S10/00—Systems supporting electrical power generation, transmission or distribution
- Y04S10/50—Systems or methods supporting the power network operation or management, involving a certain degree of interaction with the load-side end user applications
- Y04S10/52—Outage or fault management, e.g. fault detection or location
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Acoustics & Sound (AREA)
- Testing Relating To Insulation (AREA)
- Locating Faults (AREA)
Abstract
The invention relates to the technical field of cable fault monitoring, in particular to a cable partial discharge fault real-time monitoring method based on data driving, which comprises the following steps: s1, data collection and pretreatment: collecting data of partial discharge signals of the cable, and preprocessing the collected data; s2, extracting features: extracting partial discharge signal characteristics; s3, model construction: a partial discharge fault prediction model is constructed, and the partial discharge fault prediction model learns the relation between the partial discharge signal characteristics and faults, so that the prediction and diagnosis of the faults are realized; s4, real-time monitoring and early warning: monitoring partial discharge activity of the cable in real time by using a partial discharge fault prediction model; s5, fault location: after the early warning signal is sent out, the partial discharge signal characteristics are analyzed, and the position of the fault is estimated. The invention provides high-quality input data for the fault prediction model, thereby greatly improving the accuracy and efficiency of cable partial discharge fault detection.
Description
Technical Field
The invention relates to the technical field of cable fault monitoring, in particular to a cable partial discharge fault real-time monitoring method based on data driving.
Background
In the operation and maintenance of the existing power system, the cable partial discharge is a common and potentially destructive phenomenon, which usually predicts the aging or damage of the insulating material of the power equipment, possibly leading to equipment faults and even system accidents, so that the real-time monitoring and accurate diagnosis of the cable partial discharge activity become particularly important, and the traditional partial discharge monitoring method mainly depends on periodic physical inspection and simple signal analysis technology, and is long in time consumption and high in cost, continuous real-time monitoring of the cable partial discharge behavior cannot be realized, and meanwhile, the accuracy and timeliness of fault diagnosis are also insufficient.
With the rapid development of information technology and artificial intelligence, the intelligent monitoring method based on data driving provides a new solution for the detection and diagnosis of the cable partial discharge, and the method can collect cable partial discharge signals in real time by utilizing various sensors, automatically analyze the characteristics of the partial discharge signals through advanced signal processing and machine learning technologies, so as to realize the instant early warning and accurate positioning of the cable partial discharge faults, however, how to more effectively extract key characteristics in the partial discharge signals, construct an accurate and reliable fault prediction model and realize the rapid positioning of the cable partial discharge faults is still a key problem to be solved in current researches and applications.
Disclosure of Invention
Based on the above purpose, the invention provides a cable partial discharge fault real-time monitoring method based on data driving.
The cable partial discharge fault real-time monitoring method based on data driving comprises the following steps:
S1, data collection and pretreatment: collecting data of partial discharge signals of the cable, wherein the data comprise electromagnetic waveforms, acoustic waveforms and current waveforms, and preprocessing the collected data;
S2, extracting features: extracting partial discharge signal characteristics from the preprocessed data, wherein the partial discharge signal characteristics comprise discharge capacity, discharge frequency and discharge mode;
S3, model construction: based on the extracted partial discharge signal characteristics, an improved convolutional neural network CNN is adopted to construct a partial discharge fault prediction model, the partial discharge fault prediction model learns the relation between the partial discharge signal characteristics and faults, so that the prediction and diagnosis of the faults are realized, and the faults are dynamically adjusted and optimized according to newly collected data;
s4, real-time monitoring and early warning: monitoring the partial discharge activity of the cable in real time by using a partial discharge fault prediction model, and automatically generating and sending an early warning signal when the partial discharge fault prediction model predicts a fault risk;
s5, fault location: after the early warning signal is sent out, the partial discharge signal characteristics are analyzed, and the position of the fault is estimated.
Further, the data collecting and preprocessing in S1 includes:
s11, data collection: collecting data of partial discharge signals by a plurality of sensors disposed around the cable, the plurality of sensors including an electromagnetic waveform sensor, an acoustic waveform sensor, and a current waveform sensor;
S12, data preprocessing: and preprocessing the collected partial discharge signal data, including denoising, normalization and synchronous correction.
Further, the feature extraction in S2 includes:
And (3) extracting discharge quantity characteristics: the discharge capacity is expressed as a charge quantity, is a measure of the released charge in the partial discharge event, and the preprocessed current waveform data is calculated by integrating the partial discharge current waveform;
Extracting discharge frequency characteristics: the discharge frequency is the number of times of occurrence of partial discharge events in unit time, and is calculated by the reciprocal of the interval time of the partial discharge events;
And (3) extracting discharge mode characteristics: the discharge mode involves waveform analysis of the partial discharge signal, which converts the time domain signal into a frequency domain signal by fourier transformation, thereby analyzing different frequency components.
Further, the model construction in S3 includes:
Feature fusion: fusing the extracted partial discharge signal features to form a comprehensive feature vector;
model selection and training: constructing a partial discharge fault prediction model through an improved convolutional neural network CNN, using data of partial discharge signals and corresponding fault diagnosis results as a training set, and training the partial discharge fault prediction model to learn the relationship between the partial discharge signal characteristics and fault types;
model verification and tuning: evaluating the performance of the partial discharge fault prediction model on an independent test set, wherein the performance comprises accuracy, recall rate and F1 score, and optimizing parameters of the partial discharge fault prediction model based on an evaluation result;
dynamic adjustment mechanism: introducing a dynamic adjustment mechanism, and allowing the partial discharge fault prediction model to learn and adjust according to the newly collected data;
Real-time prediction and diagnosis: and inputting the data of the partial discharge signals collected in real time into a trained partial discharge fault prediction model, and carrying out fault prediction and diagnosis on the basis of the learned partial discharge signal characteristics and fault relation by the partial discharge fault prediction model.
Further, the improved convolutional neural network CNN introduces partial discharge signal strength in the convolutional layerThe modulation factor makes the application of the convolution kernel more sensitive to the intensity change of the partial discharge signal, and the calculation formula is as follows:
；
Wherein, Representing position on output feature mapIs used as a reference to the value of (a),Representing locations on an input feature mapIs used as a reference to the value of (a),Is the relative position index covered by the convolution kernel,Is the size of the convolution kernel and,Is indicated in the positionThe intensity information of the partial discharge signal is located,Is the adjustment parameter of the device, which is used for adjusting the parameters,Is the position of the convolution kernelIs a weight value of (a).
Further, the partial discharge signal strengthThe charge amount reflecting the partial discharge event is calculated by analyzing the peak intensity in the electromagnetic waveform, acoustic waveform or current waveform of the partial discharge signal, and the calculation method comprises:
Signal intensity extraction: for each partial discharge signal sample, the peak intensities in different channels (such as electromagnetic, acoustic, current) are calculated as follows:
PeakIntensity；
Wherein, Representing the absolute value of the partial discharge signal in the channel;
and (3) standardization treatment: the extracted signal intensity is standardized, and the calculation formula is as follows:
；
Each channel after processing The value will lie inWithin the interval, wherein 0 represents the lowest signal strength and 1 represents the highest signal strength;
and (3) calculating a modulation factor: integrating the channels The final value is obtainedThe modulation factor, the calculation formula is:
Wherein, the method comprises the steps of, wherein, Is the number of channels.
Further, the model verification and tuning includes:
preparation of independent test sets: dividing a part of data from the collected cable partial discharge data to be used as a test set;
Calculating a performance evaluation index:
Accuracy rate of The calculation formula is as follows:；
Wherein, (True examples) and(True negative example) is the number of fault and non-fault events correctly identified by the partial discharge fault prediction model,(False positive example) and(False negative example) is the number of fault and non-fault events that the partial discharge fault prediction model incorrectly identifies, respectively;
Recall rate of recall The calculation formula of (2) is as follows:；
the calculation formula of the score is as follows: ；
Wherein the accuracy is The calculation formula of (2) is as follows:；
Tuning model parameters based on the evaluation results: and according to the result of the evaluation index, performing fine adjustment on parameters of the partial discharge fault prediction model to optimize performance, including adjusting the learning rate, modifying the number or the size of the network layers and adjusting regularization items.
Further, the dynamic adjustment mechanism includes:
model performance was continuously monitored: periodically evaluating performance evaluation indexes of the partial discharge fault prediction model, and automatically triggering a dynamic adjustment flow when the performance evaluation indexes are lower than a preset threshold value;
data-driven adjustment decision: analyzing the data of the newly collected partial discharge signals, analyzing the reason for the performance degradation of the partial discharge fault prediction model, and determining to execute an adjustment strategy based on the analysis result, wherein the adjustment strategy comprises fine adjustment of parameters of the partial discharge fault prediction model, updating of a training set and retraining;
Fine tuning of model parameters: fine tuning the partial discharge fault prediction model, wherein the fine tuning comprises the steps of adjusting the weight of a convolution layer and a bias item;
Updating the training set: the data of the newly collected partial discharge signals and the corresponding fault diagnosis results are added into a training set regularly, the coverage range of the training set is enlarged, and the partial discharge fault prediction model learns the change of the cable operation condition and the information of the new fault mode;
retraining: and when the training set changes or the performance of the partial discharge fault prediction model is reduced, retraining the partial discharge fault prediction model, wherein retraining comprises using the updated training set or adjusting a model structure.
Further, the fault location in S5 includes:
Signal characteristic analysis: after the early warning signal is sent out, analyzing the partial discharge signal characteristics related to the early warning, wherein the analysis comprises the relation between the partial discharge signal characteristics and the correlation between the partial discharge signal characteristics and the known fault positions;
using spatial information: combining the spatial information of the partial discharge signals, and estimating the fault position by applying a time difference positioning technology;
the time difference positioning technology specifically comprises the following steps:
signal time recording: recording the exact time of arrival of the partial discharge signal on a plurality of sensors of known locations;
Calculating a time difference: for each pair of sensors, the time difference between the arrival of the signal at each pair of sensors is calculated, i.e ；
Multilateral measurement positioning: according to the propagation speed of the signalSum and time differenceSolving the fault location using multilateral measurements, the fault-to-sensor distance difference being equal to the signal propagation velocity multiplied by the time difference for each pair of sensors, i.e；
Calculating fault positions: and (3) calculating the position of the fault by combining time difference data of all the sensor pairs.
Further, the fault location further includes a signal strength difference technique, where the signal strength difference technique specifically includes:
signal strength measurement: measuring the intensity of the partial discharge signal received by each sensor;
and (3) establishing an attenuation model: the relationship between signal strength and distance is represented using a path loss model, which is expressed as:
；
Wherein, Is at a distance ofWhere the measured signal strength is measured and,Is at a reference distanceThe signal strength at which the signal is to be received,Is an environmental factor that is used to determine the environmental factor,Is the distance from the fault point to the sensor;
Distance estimation: based on the measurements Estimating the distance from the fault point to each sensor by using the value and the path loss model;
position estimation: and (3) calculating the position of the fault by combining the distance estimation of all the sensors.
The invention has the beneficial effects that:
According to the invention, key signal data from partial discharge of the cable can be accurately captured and processed through efficient data collection and preprocessing steps, including electromagnetic waveforms, acoustic waveforms and current waveforms, the preprocessing operation improves the quality of the data and the accuracy of subsequent analysis, the advantage ensures that the characteristics extracted from the signals can accurately reflect the actual condition of a partial discharge event, and high-quality input data is provided for a fault prediction model, so that the accuracy and the efficiency of cable partial discharge fault detection are greatly improved.
According to the invention, the partial discharge fault prediction model constructed by the improved convolutional neural network CNN can deeply learn the complex relation between the partial discharge signal characteristics and faults, so that accurate prediction and diagnosis of the faults are realized, particularly, the model enhances the recognition capability of the cable partial discharge fault characteristics by introducing the partial discharge signal intensity modulation factor, particularly, the model is more sensitive when analyzing the change of the partial discharge signal intensity, which is critical to the recognition and prediction of the cable partial discharge faults, and in addition, the model also has the capability of dynamically adjusting and optimizing according to newly collected data, so that the change of the state of a cable system along with the time is ensured, and the model can continuously adapt to the new fault mode and condition change.
The invention can estimate the position of fault occurrence through time difference positioning technology and signal intensity difference technology, not only improves the accuracy of fault positioning, but also greatly quickens the speed of fault diagnosis, provides powerful support for rapidly solving the cable partial discharge fault, greatly reduces the maintenance cost of fault positioning information, improves the reliability and safety of the power system, and can provide accurate guidance for maintenance work and rapidly solving the fault by combining the space information of the partial discharge signal, thereby greatly improving the efficiency of cable maintenance and management.
Drawings
In order to more clearly illustrate the invention or the technical solutions of the prior art, the drawings which are used in the description of the embodiments or the prior art will be briefly described, it being obvious that the drawings in the description below are only of the invention and that other drawings can be obtained from them without inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow chart of a monitoring method according to an embodiment of the invention;
Fig. 2 is a schematic diagram of a model construction flow according to an embodiment of the present invention.
Detailed Description
The present invention will be further described in detail with reference to specific embodiments in order to make the objects, technical solutions and advantages of the present invention more apparent.
It is to be noted that unless otherwise defined, technical or scientific terms used herein should be taken in a general sense as understood by one of ordinary skill in the art to which the present invention belongs. The terms "first," "second," and the like, as used herein, do not denote any order, quantity, or importance, but rather are used to distinguish one element from another. The word "comprising" or "comprises", and the like, means that elements or items preceding the word are included in the element or item listed after the word and equivalents thereof, but does not exclude other elements or items. The terms "connected" or "connected," and the like, are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "upper", "lower", "left", "right", etc. are used merely to indicate relative positional relationships, which may also be changed when the absolute position of the object to be described is changed.
As shown in fig. 1-2, the real-time monitoring method for the partial discharge faults of the cable based on data driving comprises the following steps:
s1, data collection and pretreatment: collecting data of partial discharge signals of the cable, wherein the data comprise electromagnetic waveforms, acoustic waveforms and current waveforms, and preprocessing the collected data to improve the quality of the data and the accuracy of subsequent analysis;
s2, extracting features: extracting partial discharge signal characteristics from the preprocessed data, wherein the partial discharge signal characteristics comprise discharge capacity, discharge frequency and discharge mode and are used for describing and identifying partial discharge faults;
s3, model construction: based on the extracted partial discharge signal characteristics, an improved convolutional neural network CNN is adopted to construct a partial discharge fault prediction model, the relation between the partial discharge signal characteristics and faults is learned by the partial discharge fault prediction model, the prediction and diagnosis of the faults are realized, and the dynamic adjustment and optimization are carried out according to the newly collected data so as to improve the accuracy of the prediction;
S4, real-time monitoring and early warning: monitoring the partial discharge activity of the cable in real time by using a partial discharge fault prediction model, and automatically generating and sending an early warning signal when the partial discharge fault prediction model predicts a fault risk so as to take corresponding maintenance measures or preventive actions in time;
S5, fault location: after an early warning signal is sent out, analyzing the characteristics of the partial discharge signal, estimating the position of the fault, and ensuring that the fault positioning is critical for guiding maintenance work and rapidly solving the fault;
The invention realizes real-time monitoring and early warning of the cable partial discharge faults through the steps, not only improves the accuracy and efficiency of fault detection, but also can realize the timely positioning of faults, greatly reduces the maintenance cost and improves the reliability of the power system.
The data collection and preprocessing in S1 includes:
S11, data collection: collecting data of partial discharge signals by a plurality of sensors disposed around the cable, the plurality of sensors including an electromagnetic waveform sensor (for capturing electromagnetic fluctuations generated during the partial discharge of the cable), an acoustic waveform sensor (for recording acoustic characteristics induced by the partial discharge event, which can provide important clues about the nature and location of the partial discharge), and a current waveform sensor (for monitoring changes in current in the cable), the plurality of sensors being precisely arranged on the cable to ensure that all necessary data relating to the partial discharge of the cable can be captured comprehensively;
S12, data preprocessing: and preprocessing the collected partial discharge signal data to improve the accuracy and efficiency of data analysis, including denoising, standardization and synchronous correction.
The feature extraction in S2 includes:
And (3) extracting discharge quantity characteristics: the discharge capacity is expressed as a charge quantity, is a measure of the released charge in the partial discharge event, the preprocessed current waveform data is calculated by integrating the partial discharge current waveform, the discharge capacity is a key index reflecting the strength and severity of the partial discharge, and the severity of the partial discharge fault can be estimated by comparing the magnitudes of the discharge capacity;
the calculation formula of the integral partial discharge current waveform is as follows: Wherein, the method comprises the steps of, wherein, Is the amount of discharge to be conducted,Is the current waveform during the partial discharge event,Time is;
Extracting discharge frequency characteristics: the discharge frequency is the number of times of occurrence of the partial discharge event in unit time, the discharge frequency is calculated by the reciprocal of the interval time of the partial discharge event, and the discharge frequency is an important basis for judging the type of discharge activity (such as continuous discharge or intermittent discharge);
the discharge frequency is calculated by the following formula: Wherein, the method comprises the steps of, wherein, Is the discharge frequency, is the time interval between two consecutive partial discharge events;
And (3) extracting discharge mode characteristics: the discharge mode relates to waveform analysis of partial discharge signals, the time domain signals are converted into frequency domain signals through Fourier transformation, so that different frequency components are analyzed, and different discharge modes (such as internal discharge, surface discharge and the like) can be accurately divided and identified through analyzing specific modes and waveform characteristics in data, which is important for fault analysis and prevention;
The calculation formula of the frequency domain signal is as follows: Wherein, the method comprises the steps of, wherein, Is a frequency domain representation of the signal,Is the angular frequency of the wave form,Is a partial discharge signal in the time domain;
Through feature extraction, key partial discharge features including discharge capacity, discharge frequency and discharge mode can be accurately identified from complex partial discharge signal data, key information is provided for subsequent model construction, real-time monitoring and early warning and fault positioning, and the accuracy and efficiency of the cable partial discharge fault monitoring method are greatly improved.
The model construction in S3 includes:
Feature fusion: fusing the extracted partial discharge signal features to form a comprehensive feature vector, wherein the comprehensive feature vector comprehensively reflects the multidimensional characteristics of the partial discharge event and provides rich input data for the model;
model selection and training: constructing a partial discharge fault prediction model through an improved convolutional neural network CNN, using data of partial discharge signals and corresponding fault diagnosis results as a training set, and training the partial discharge fault prediction model to learn the relationship between the partial discharge signal characteristics and fault types;
Model verification and tuning: evaluating the performance of the partial discharge fault prediction model on an independent test set, wherein the performance comprises accuracy, recall rate and F1 score, and optimizing parameters of the partial discharge fault prediction model based on an evaluation result to ensure that the model has optimal prediction performance;
dynamic adjustment mechanism: introducing a dynamic adjustment mechanism, and allowing the partial discharge fault prediction model to learn and adjust according to the newly collected data;
real-time prediction and diagnosis: inputting the data of the partial discharge signals collected in real time into a trained partial discharge fault prediction model, wherein the partial discharge fault prediction model predicts and diagnoses faults based on the learned partial discharge signal characteristics and the fault relation, and the prediction result can be used for immediate fault early warning and subsequent maintenance decision support;
Through the process, not only can the high-accuracy partial discharge fault prediction and diagnosis be realized, but also the change of the cable partial discharge characteristic can be adapted through continuous learning and optimization, so that the high performance can be maintained in a dynamically-changed power grid environment.
Improved convolutional neural network CNN introduces partial discharge signal strength in convolutional layerThe modulation factor makes the application of the convolution kernel more sensitive to the intensity change of the partial discharge signal, and the calculation formula is as follows:
；
Wherein, Representing position on output feature mapIs used as a reference to the value of (a),Representing locations on an input feature mapIs used as a reference to the value of (a),Is the relative position index covered by the convolution kernel,Is the size of the convolution kernel, i.e. the number of elements on the input feature map covered by the convolution operation,Is indicated in the positionThe intensity information of the partial discharge signal is located,Is an adjusting parameter for controlling the modulating function of the intensity of the partial discharge signal on the convolution output,Is the position of the convolution kernelWeight value of (2);
Improved convolutional neural network CNN by introducing partial discharge signal strength in convolutional operations The modulating factor enhances the identification and analysis capacity of the partial discharge fault prediction model to the cable partial discharge fault characteristics, and further improves the effect and accuracy of the cable partial discharge fault real-time monitoring method.
Partial discharge signal strengthThe charge amount reflecting the partial discharge event is calculated by analyzing the peak intensity in the electromagnetic waveform, acoustic waveform or current waveform of the partial discharge signal, and the calculation method comprises:
Signal intensity extraction: for each partial discharge signal sample, the peak intensities in different channels (such as electromagnetic, acoustic, current) are calculated as follows:
PeakIntensity；
Wherein, Representing the absolute value of the partial discharge signal in the channel;
and (3) standardization treatment: in order to make The modulation factor is suitable for data of different scales and partial discharge events of different intensities, the extracted signal intensity is subjected to standardized processing, and the calculation formula is as follows:
；
Each channel after processing The value will lie inWithin the interval, wherein 0 represents the lowest signal strength and 1 represents the highest signal strength;
and (3) calculating a modulation factor: integrating the channels The final value is obtainedThe modulation factor, the calculation formula is:
Wherein, the method comprises the steps of, wherein, Is the number of channels;
by introducing the strength of the partial discharge signal The modulation factor can make the convolutional neural network CNN pay more attention to the strength characteristics of the signals when analyzing the partial discharge signals, is very important for identifying and predicting the cable partial discharge faults, and the signal strength is often closely related to the severity of the faults, so that the convolutional neural network CNN is more effective and accurate in identifying potential defects of the cables and predicting possible faults.
Model verification and tuning includes:
Preparation of independent test sets: a part of data is segmented from the collected cable partial discharge data to serve as a test set, and the test set is not used in the training stage of the partial discharge fault prediction model, so that independent verification can be provided in the model evaluation process;
Calculating a performance evaluation index:
Accuracy rate of The calculation formula (of the proportion of the model to be correctly predicted) is:；
Wherein, (True examples) and(True negative example) is the number of fault and non-fault events correctly identified by the partial discharge fault prediction model,(False positive example) and(False negative example) is the number of fault and non-fault events that the partial discharge fault prediction model incorrectly identifies, respectively;
Recall rate of recall The calculation formula (the proportion of the fault event correctly identified by the model to all actual fault events) is as follows:；
The calculation formula of the score (the harmonic mean of the accuracy and the recall) is: ；
Wherein the accuracy is The calculation formula of (2) is as follows:；
tuning model parameters based on the evaluation results: according to the result of the evaluation index, the parameters of the partial discharge fault prediction model are finely adjusted to optimize performance, including adjusting the learning rate, modifying the number or the size of the network layers, and adjusting regularization items so as to achieve better accuracy, recall rate and F1 score;
By evaluating the accuracy, recall and F1 score of the improved convolutional neural network CNN on an independent test set, the ability of the model to predict cable partial discharge faults can be fully understood, with recall being particularly important for fault prediction, as it is directly related to the integrity of fault detection, i.e., whether the model can capture all fault events.
Based on the evaluation result, the identification accuracy and generalization capability of the model to the partial discharge faults can be further improved by adjusting the parameters of the model (such as the size of a convolution kernel, the number of layers, the learning rate and the like), and the adjustment process ensures that the model can effectively predict the cable partial discharge faults in practical application, provides scientific basis for timely maintenance and prevention, and greatly improves the reliability and safety of a cable system.
The dynamic adjustment mechanism includes:
model performance was continuously monitored: periodically evaluating performance evaluation indexes of the partial discharge fault prediction model, and automatically triggering a dynamic adjustment flow when the performance evaluation indexes are lower than a preset threshold value;
Data-driven adjustment decision: analyzing the data of the newly collected partial discharge signals, analyzing the reasons for the performance degradation of the partial discharge fault prediction model, such as the newly occurring fault mode or cable state change, and deciding to execute an adjustment strategy based on the analysis result, wherein the adjustment strategy comprises fine adjustment of the parameters of the partial discharge fault prediction model, update of a training set (such as the introduction of the latest partial discharge event and the analysis result thereof) and retraining;
Fine tuning of model parameters: fine tuning is carried out on the partial discharge fault prediction model so as to reflect information in the latest data, wherein the fine tuning comprises adjustment of convolution layer weight and bias items;
The fine adjustment of the weight of the convolution layer is realized by a gradient descent method, and the calculation formula is as follows:
；
Wherein, Is the weight after the update of the weight,Is the current weight of the current weight,Is the learning rate, usually a smaller value is selected when fine tuning,Is a loss functionRegarding weightsIs a gradient of (2);
The calculation formula of bias term fine tuning is:
；
Wherein, Is an updated bias term that is used to determine,Is the current bias term and,Is a loss functionWith respect to bias termsIs a gradient of (2);
Updating the training set: the data of the newly collected partial discharge signals and the corresponding fault diagnosis results are added into a training set regularly, the coverage range of the training set is enlarged, and the partial discharge fault prediction model learns the change of the cable operation condition and the information of the new fault mode;
Retraining: when the training set is significantly changed or the performance of the partial discharge fault prediction model is reduced, retraining the partial discharge fault prediction model is carried out, wherein retraining comprises using the updated training set or adjusting a model structure (such as adding a convolution layer, adjusting the size of a convolution kernel and the like);
The improved convolutional neural network CNN can learn and adjust according to new partial discharge data, and continuously optimize the performance of the convolutional neural network CNN, so that not only is the prediction accuracy of the model in initial deployment improved, but also the model is ensured to adapt to new fault modes and condition changes along with the time lapse and the cable system state change, and the long-term effectiveness and adaptability of the model are maintained.
Fault location in S5 includes:
Signal characteristic analysis: after the early warning signal is sent out, analyzing the partial discharge signal characteristics related to the early warning, wherein the analysis comprises the relation between the partial discharge signal characteristics and the correlation between the partial discharge signal characteristics and the known fault positions;
using spatial information: in combination with spatial information of the partial discharge signals, such as signal strength differences and arrival time differences captured by sensors at different locations, a time difference localization technique is applied to estimate the location of the fault;
The time difference positioning technology specifically comprises the following steps:
signal time recording: recording the exact time of arrival of the partial discharge signal on a plurality of sensors of known locations;
Calculating a time difference: for each pair of sensors, the time difference between the arrival of the signal at each pair of sensors is calculated, i.e ；
Multilateral measurement positioning: according to the propagation speed of the signalSum and time differenceSolving the fault location using multilateral measurements, the fault-to-sensor distance difference being equal to the signal propagation velocity multiplied by the time difference for each pair of sensors, i.e；
Calculating fault positions: and (3) calculating the position of the fault by combining time difference data of all the sensor pairs.
The fault location further includes a signal strength difference technique, which specifically includes:
signal strength measurement: measuring the intensity of the partial discharge signal received by each sensor;
and (3) establishing an attenuation model: the relationship between signal strength and distance is represented using a path loss model, which is expressed as:
；
Wherein, Is at a distance ofWhere the measured signal strength is measured and,Is at a reference distanceThe signal strength at this point, typically taken to be 1 meter,Is an environmental factor, depending on the environmental characteristics of the signal propagation,Is the distance from the fault point to the sensor;
Distance estimation: based on the measurements Estimating the distance from the fault point to each sensor by using the value and the path loss model;
position estimation: and (3) calculating the position of the fault by combining the distance estimation of all the sensors.
Those of ordinary skill in the art will appreciate that: the discussion of any of the embodiments above is merely exemplary and is not intended to suggest that the scope of the invention is limited to these examples; the technical features of the above embodiments or in the different embodiments may also be combined within the idea of the invention, the steps may be implemented in any order and there are many other variations of the different aspects of the invention as described above, which are not provided in detail for the sake of brevity.
The present invention is intended to embrace all such alternatives, modifications and variances which fall within the broad scope of the appended claims. Therefore, any omission, modification, equivalent replacement, improvement, etc. of the present invention should be included in the scope of the present invention.
Claims (5)
1. The cable partial discharge fault real-time monitoring method based on data driving is characterized by comprising the following steps of:
S1, data collection and pretreatment: collecting data of partial discharge signals of the cable, wherein the data comprise electromagnetic waveforms, acoustic waveforms and current waveforms, and preprocessing the collected data;
S2, extracting features: extracting partial discharge signal characteristics from the preprocessed data, wherein the partial discharge signal characteristics comprise discharge capacity, discharge frequency and discharge mode;
S3, model construction: based on the extracted partial discharge signal characteristics, an improved convolutional neural network CNN is adopted to construct a partial discharge fault prediction model, the partial discharge fault prediction model learns the relation between the partial discharge signal characteristics and faults, so that the prediction and diagnosis of the faults are realized, and the faults are dynamically adjusted and optimized according to newly collected data;
s4, real-time monitoring and early warning: monitoring the partial discharge activity of the cable in real time by using a partial discharge fault prediction model, and automatically generating and sending an early warning signal when the partial discharge fault prediction model predicts a fault risk;
S5, fault location: after an early warning signal is sent out, analyzing the characteristics of the partial discharge signal, and estimating the position of fault occurrence;
the data collection and preprocessing in S1 includes:
s11, data collection: collecting data of partial discharge signals by a plurality of sensors disposed around the cable, the plurality of sensors including an electromagnetic waveform sensor, an acoustic waveform sensor, and a current waveform sensor;
s12, data preprocessing: preprocessing the collected partial discharge signal data, including denoising, standardization and synchronous correction;
The feature extraction in S2 includes:
And (3) extracting discharge quantity characteristics: the discharge capacity is expressed as a charge quantity, is a measure of the released charge in the partial discharge event, and the preprocessed current waveform data is calculated by integrating the partial discharge current waveform;
Extracting discharge frequency characteristics: the discharge frequency is the number of times of occurrence of partial discharge events in unit time, and is calculated by the reciprocal of the interval time of the partial discharge events;
And (3) extracting discharge mode characteristics: the discharging mode relates to waveform analysis of partial discharge signals, and a time domain signal is converted into a frequency domain signal through Fourier transformation, so that different frequency components are analyzed;
the model construction in S3 includes:
Feature fusion: fusing the extracted partial discharge signal features to form a comprehensive feature vector;
model selection and training: constructing a partial discharge fault prediction model through an improved convolutional neural network CNN, using data of partial discharge signals and corresponding fault diagnosis results as a training set, and training the partial discharge fault prediction model to learn the relationship between the partial discharge signal characteristics and fault types;
model verification and tuning: evaluating the performance of the partial discharge fault prediction model on an independent test set, wherein the performance comprises accuracy, recall rate and F1 score, and optimizing parameters of the partial discharge fault prediction model based on an evaluation result;
dynamic adjustment mechanism: introducing a dynamic adjustment mechanism, and allowing the partial discharge fault prediction model to learn and adjust according to the newly collected data;
Real-time prediction and diagnosis: inputting the data of the partial discharge signals collected in real time into a trained partial discharge fault prediction model, and carrying out fault prediction and diagnosis on the basis of the learned partial discharge signal characteristics and the fault relation by the partial discharge fault prediction model;
The improved convolutional neural network CNN introduces partial discharge signal strength in the convolutional layer The modulation factor makes the application of the convolution kernel more sensitive to the intensity change of the partial discharge signal, and the calculation formula is as follows:
；
Wherein, Representing position on output feature mapIs used as a reference to the value of (a),Representing locations on an input feature mapIs used as a reference to the value of (a),Is the relative position index covered by the convolution kernel,Is the size of the convolution kernel and,Is indicated in the positionThe intensity information of the partial discharge signal is located,Is the adjustment parameter of the device, which is used for adjusting the parameters,Is the position of the convolution kernelWeight value of (2);
The partial discharge signal strength The charge amount reflecting the partial discharge event is calculated by analyzing the peak intensity in the electromagnetic waveform, acoustic waveform or current waveform of the partial discharge signal, and the calculation method comprises:
Signal intensity extraction: for each partial discharge signal sample, peak intensities in different channels are calculated according to the following calculation formula:
PeakIntensity；
Wherein, Representing the absolute value of the partial discharge signal in the channel;
and (3) standardization treatment: the extracted signal intensity is standardized, and the calculation formula is as follows:
；
Each channel after processing The value will lie inWithin the interval, wherein 0 represents the lowest signal strength and 1 represents the highest signal strength;
and (3) calculating a modulation factor: integrating the channels The final value is obtainedThe modulation factor, the calculation formula is:
Wherein, the method comprises the steps of, wherein, Is the number of channels.
2. The method for monitoring the partial discharge fault of the cable based on data driving according to claim 1, wherein the model verification and optimization comprises:
preparation of independent test sets: dividing a part of data from the collected cable partial discharge data to be used as a test set;
Calculating a performance evaluation index:
Accuracy rate of The calculation formula is as follows:；
Wherein, AndThe number of fault and non-fault events correctly identified by the partial discharge fault prediction model,AndThe number of fault and non-fault events which are wrongly identified by the partial discharge fault prediction model are respectively;
Recall rate of recall The calculation formula of (2) is as follows:；
the calculation formula of the score is as follows: ；
Wherein the accuracy is The calculation formula of (2) is as follows:；
Tuning model parameters based on the evaluation results: and according to the result of the evaluation index, performing fine adjustment on parameters of the partial discharge fault prediction model to optimize performance, including adjusting the learning rate, modifying the number or the size of the network layers and adjusting regularization items.
3. The method for monitoring the partial discharge fault of the cable based on data driving according to claim 2, wherein the dynamic adjustment mechanism comprises:
model performance was continuously monitored: periodically evaluating performance evaluation indexes of the partial discharge fault prediction model, and automatically triggering a dynamic adjustment flow when the performance evaluation indexes are lower than a preset threshold value;
data-driven adjustment decision: analyzing the data of the newly collected partial discharge signals, analyzing the reason for the performance degradation of the partial discharge fault prediction model, and determining to execute an adjustment strategy based on the analysis result, wherein the adjustment strategy comprises fine adjustment of parameters of the partial discharge fault prediction model, updating of a training set and retraining;
Fine tuning of model parameters: fine tuning the partial discharge fault prediction model, wherein the fine tuning comprises the steps of adjusting the weight of a convolution layer and a bias item;
Updating the training set: the data of the newly collected partial discharge signals and the corresponding fault diagnosis results are added into a training set regularly, the coverage range of the training set is enlarged, and the partial discharge fault prediction model learns the change of the cable operation condition and the information of the new fault mode;
retraining: and when the training set changes or the performance of the partial discharge fault prediction model is reduced, retraining the partial discharge fault prediction model, wherein retraining comprises using the updated training set or adjusting a model structure.
4. The method for monitoring a partial discharge fault of a cable based on data driving according to claim 3, wherein the locating of the fault in S5 comprises:
Signal characteristic analysis: after the early warning signal is sent out, analyzing the partial discharge signal characteristics related to the early warning, wherein the analysis comprises the relation between the partial discharge signal characteristics and the correlation between the partial discharge signal characteristics and the known fault positions;
using spatial information: combining the spatial information of the partial discharge signals, and estimating the fault position by applying a time difference positioning technology;
the time difference positioning technology specifically comprises the following steps:
signal time recording: recording the exact time of arrival of the partial discharge signal on a plurality of sensors of known locations;
Calculating a time difference: for each pair of sensors, the time difference between the arrival of the signal at each pair of sensors is calculated, i.e ；
Multilateral measurement positioning: according to the propagation speed of the signalSum and time differenceSolving the fault location using multilateral measurements, the fault-to-sensor distance difference being equal to the signal propagation velocity multiplied by the time difference for each pair of sensors, i.e；
Calculating fault positions: and (3) calculating the position of the fault by combining time difference data of all the sensor pairs.
5. The method for monitoring the partial discharge fault of the cable based on data driving according to claim 4, wherein the fault location further comprises a signal strength difference technology, and the signal strength difference technology specifically comprises:
signal strength measurement: measuring the intensity of the partial discharge signal received by each sensor;
and (3) establishing an attenuation model: the relationship between signal strength and distance is represented using a path loss model, which is expressed as:
；
Wherein, Is at a distance ofWhere the measured signal strength is measured and,Is at a reference distanceThe signal strength at which the signal is to be received,Is an environmental factor that is used to determine the environmental factor,Is the distance from the fault point to the sensor;
Distance estimation: based on the measurements Estimating the distance from the fault point to each sensor by using the value and the path loss model;
position estimation: and (3) calculating the position of the fault by combining the distance estimation of all the sensors.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410577172.9A CN118130984B (en) | 2024-05-10 | 2024-05-10 | Cable partial discharge fault real-time monitoring method based on data driving |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410577172.9A CN118130984B (en) | 2024-05-10 | 2024-05-10 | Cable partial discharge fault real-time monitoring method based on data driving |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN118130984A CN118130984A (en) | 2024-06-04 |
| CN118130984B true CN118130984B (en) | 2024-07-12 |
Family
ID=91248418
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410577172.9A Active CN118130984B (en) | 2024-05-10 | 2024-05-10 | Cable partial discharge fault real-time monitoring method based on data driving |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118130984B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118362847B (en) * | 2024-06-20 | 2024-09-24 | 珠海康晋电气股份有限公司 | Automatic data acquisition method |
| CN118584399B (en) * | 2024-08-02 | 2024-11-05 | 红光电气集团有限公司 | A method for locating partial discharge position of complete set of equipment in 10KV transformer area |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116736051A (en) * | 2023-05-22 | 2023-09-12 | 内蒙古工业大学 | GIS insulation defect intelligent diagnosis method, device, equipment and storage medium |
| CN117590173A (en) * | 2023-11-29 | 2024-02-23 | 伊犁新天煤化工有限责任公司 | Cable partial discharge pattern recognition method based on convolutional neural network |
Family Cites Families (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105548832B (en) * | 2015-12-10 | 2018-09-21 | 国网四川省电力公司电力科学研究院 | High voltage power cable fault recognition method |
| CN106353648B (en) * | 2016-08-30 | 2024-04-09 | 浙江新图维电子科技有限公司 | Comprehensive partial discharge monitoring method |
| CN109856517B (en) * | 2019-03-29 | 2021-02-23 | 国家电网有限公司 | A Discrimination Method for Partial Discharge Detection Data of UHV Equipment |
| KR102284958B1 (en) * | 2019-11-15 | 2021-08-04 | 한전케이디엔주식회사 | Partial discharge position estimation appratus and method |
| CN113985318A (en) * | 2021-10-26 | 2022-01-28 | 国网山东省电力公司检修公司 | Multi-type partial discharge signal real-time extraction method for transformer operation state |
| JP2024003552A (en) * | 2022-06-27 | 2024-01-15 | 東芝インフラシステムズ株式会社 | Partial discharge inspection method and partial discharge inspection device |
| CN116008751A (en) * | 2023-01-06 | 2023-04-25 | 重庆邮电大学 | Sample enhancement-based cable partial discharge fault characteristic identification method |
| CN116027158A (en) * | 2023-02-07 | 2023-04-28 | 国网辽宁省电力有限公司电力科学研究院 | High-voltage cable partial discharge fault prediction method and system |
| CN116796187A (en) * | 2023-05-23 | 2023-09-22 | 国网上海市电力公司 | Power transmission line partial discharge detection method |
| CN117828333A (en) * | 2024-01-04 | 2024-04-05 | 重庆邮电大学 | A cable partial discharge feature extraction method based on signal mixing enhancement and CNN |
-
2024
- 2024-05-10 CN CN202410577172.9A patent/CN118130984B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116736051A (en) * | 2023-05-22 | 2023-09-12 | 内蒙古工业大学 | GIS insulation defect intelligent diagnosis method, device, equipment and storage medium |
| CN117590173A (en) * | 2023-11-29 | 2024-02-23 | 伊犁新天煤化工有限责任公司 | Cable partial discharge pattern recognition method based on convolutional neural network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN118130984A (en) | 2024-06-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN118130984B (en) | Cable partial discharge fault real-time monitoring method based on data driving | |
| CN110647539B (en) | Prediction method and system for vehicle faults | |
| CN109186813B (en) | Temperature sensor self-checking device and method | |
| CN113657221B (en) | Power plant equipment state monitoring method based on intelligent sensing technology | |
| CN112187528B (en) | On-line monitoring method of communication flow in industrial control system based on SARIMA | |
| CN112173636B (en) | Method for detecting faults of belt conveyor carrier roller by inspection robot | |
| CN110209999A (en) | A kind of mobile unit failure trend prediction method | |
| CN107844067B (en) | A kind of gate of hydropower station on-line condition monitoring control method and monitoring system | |
| CN109655298A (en) | A kind of the failure real time early warning method and device of large span metal Roof | |
| CN118050591A (en) | Distribution network 10KV overhead line fault early warning and positioning system based on AI | |
| CN115238785A (en) | Rotary machine fault diagnosis method and system based on image fusion and integrated network | |
| CN115618205A (en) | Portable voiceprint fault detection system and method | |
| CN114486259A (en) | Signal Processing Method for Distributed Optical Fiber Acoustic Sensing System with Optimal Variational Mode Decomposition | |
| US20230024104A1 (en) | Identification of false transformer humming using machine learning | |
| CN115165362B (en) | A bearing fault intelligent detection system and method based on inspection and vibration signals | |
| CN118840818B (en) | A fire warning method for intelligent cables | |
| CN103487250A (en) | Coal mining equipment predictive maintenance method based on two-dimensional projection | |
| CN118244110B (en) | Industrial robot motor fault early warning method and system | |
| CN117630797A (en) | Electric meter health status detection method, system and storage medium based on working current | |
| CN118038295A (en) | Real-time online identification system for defects of power transmission line | |
| CN117951581A (en) | Pile foundation integrity assessment method based on deep learning and ultrasonic reflection | |
| CN117744029A (en) | Computer host running environment adaptability test system based on data analysis | |
| CN116910879A (en) | A method and device for diagnosing abnormal cable force of bridge stay cables under random vehicle loads | |
| CN114778115A (en) | Bearing fault detection system and method based on acoustic emission signals | |
| CN117113259B (en) | Coal mine state data processing method and system for predicting potential safety hazards |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |