CN117911011A - AC/DC series-parallel power line fault maintenance early warning method - Google Patents
AC/DC series-parallel power line fault maintenance early warning method Download PDFInfo
- Publication number
- CN117911011A CN117911011A CN202410310863.2A CN202410310863A CN117911011A CN 117911011 A CN117911011 A CN 117911011A CN 202410310863 A CN202410310863 A CN 202410310863A CN 117911011 A CN117911011 A CN 117911011A
- Authority
- CN
- China
- Prior art keywords
- membership
- data
- power line
- fault
- features
- 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.)
- Granted
Links
- 238000012423 maintenance Methods 0.000 title claims abstract description 59
- 238000000034 method Methods 0.000 title claims abstract description 37
- 238000012544 monitoring process Methods 0.000 claims abstract description 62
- 238000011156 evaluation Methods 0.000 claims abstract description 45
- 239000011159 matrix material Substances 0.000 claims abstract description 45
- 238000012545 processing Methods 0.000 claims abstract description 21
- 238000001514 detection method Methods 0.000 claims abstract description 16
- 238000000605 extraction Methods 0.000 claims description 11
- 230000002159 abnormal effect Effects 0.000 claims description 9
- 230000000737 periodic effect Effects 0.000 claims description 9
- 238000010606 normalization Methods 0.000 claims description 6
- 230000008569 process Effects 0.000 claims description 6
- 238000004364 calculation method Methods 0.000 claims description 5
- 238000007781 pre-processing Methods 0.000 claims description 5
- 230000003595 spectral effect Effects 0.000 description 6
- 230000005856 abnormality Effects 0.000 description 5
- 230000005540 biological transmission Effects 0.000 description 5
- 230000000630 rising effect Effects 0.000 description 4
- 238000001228 spectrum Methods 0.000 description 4
- 230000009466 transformation Effects 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 230000001052 transient effect Effects 0.000 description 2
- 102000002274 Matrix Metalloproteinases Human genes 0.000 description 1
- 108010000684 Matrix Metalloproteinases Proteins 0.000 description 1
- 238000004140 cleaning Methods 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000005684 electric field Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000009413 insulation Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000013450 outlier detection Methods 0.000 description 1
- 238000013021 overheating Methods 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Landscapes
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention discloses an AC/DC series-parallel power line fault maintenance early warning method, which relates to the technical field of power fault detection and comprises the following steps: collecting voltage, current, temperature and noise data of an alternating current-direct current series-parallel power line in a monitoring period; extracting features of the preprocessed data, wherein the features comprise statistical features, time-frequency features and frequency-domain features; establishing a fuzzy model, and setting the model output as fault probability; the data collected in each monitoring period are subjected to fuzzy processing to obtain a membership value matrix of each characteristic data, the membership value matrices of all the characteristic data are combined to form a first membership value matrix, and the first membership value matrix and the weight corresponding to each characteristic data are combined to obtain a second membership value matrix; and determining an evaluation result according to the specific membership value of each evaluation result in the second membership value matrix, and carrying out corresponding maintenance early warning according to the evaluation result. The invention realizes more accurate, reliable and timely fault maintenance and early warning of the AC/DC series-parallel power line.
Description
Technical Field
The invention relates to the technical field of power failure detection, in particular to an AC/DC hybrid power line failure maintenance early warning method.
Background
The ac/dc series-parallel power line refers to a power transmission mode in which an ac power transmission line and a dc power transmission line are simultaneously used to transmit electric power from one ac power system to another ac power system. The alternating current and direct current are combined, and the two types of electric energy are transmitted and distributed simultaneously, so that the electric power system has higher flexibility and efficiency, and has the advantages of large power transmission capacity, long distance, high efficiency and the like.
Along with the development of a power system and the improvement of the intelligent degree, the use of an alternating-current/direct-current hybrid power line is more and more common, and the alternating-current/direct-current hybrid power line comprises a plurality of fields of the power system, industry, commerce, traffic and the like. In the power system, the alternating-current/direct-current hybrid power line can realize long-distance and high-capacity power transmission, and the stability and reliability of the power system are improved; in the industrial and commercial fields, the alternating-current and direct-current series-parallel power line can meet various complicated power requirements; in the traffic field, the alternating current-direct current series-parallel power line can be used for electric automobile charging piles and the like. However, due to the complex operation environment of the power line, various faults, such as short circuit, open circuit, overload, etc., may occur in the ac/dc hybrid power line, and these faults may affect the stability and reliability of the power system, so the fault maintenance early warning of the ac/dc hybrid power line becomes an important task. The fault maintenance early warning can help power system operation staff to discover and process potential faults in time, and reliability and safety of the power system are improved.
The traditional fault maintenance early warning method is mostly dependent on manual experience and single index judgment, and the manual detection process may need a certain time and cost, lacks comprehensiveness and accuracy and is easily influenced by subjective factors; meanwhile, the traditional method often ignores the mutual influence and comprehensive consideration among the features, and can not fully utilize the implicit information in the data.
Disclosure of Invention
(One) solving the technical problems
Aiming at the technical problems in the background technology, the invention provides an AC/DC series-parallel power line fault maintenance early warning method, which comprises the following steps: collecting voltage, current, temperature and noise data of an alternating current-direct current series-parallel power line in a monitoring period; extracting features of the preprocessed data, wherein the features comprise statistical features, time-frequency features and frequency-domain features; establishing a fuzzy model, and setting the model output as fault probability; the data collected in each monitoring period are subjected to fuzzy processing to obtain a membership value matrix of each characteristic data, the membership value matrices of all the characteristic data are combined to form a first membership value matrix, and the first membership value matrix and the weight corresponding to each characteristic data are combined to obtain a second membership value matrix; determining an evaluation result according to the specific membership value of each evaluation result in the second membership value matrix, and carrying out corresponding maintenance early warning according to the evaluation result; the traditional AC/DC series-parallel power line fault maintenance and early warning method is optimized.
(II) technical scheme
In order to achieve the above purpose, the invention is realized by the following technical scheme:
A fault maintenance early warning method for an AC/DC series-parallel power line comprises the following steps:
S1, setting a monitoring period, collecting voltage, current, temperature and noise data in the working state of an alternating-direct current series-parallel power line in the monitoring period, and preprocessing the collected data;
s2, extracting features of the preprocessed data, including statistical features, time-frequency features and frequency domain features, and recording the features as actual features; the same characteristic extraction is carried out on the voltage, the current, the temperature and the noise data in the historical data, and the historical data is recorded as the historical characteristic;
S3, establishing a fuzzy model, setting the model output as fault probability, performing fuzzy logic processing on the normalized historical characteristic data to obtain a membership function, and formulating a corresponding fuzzy rule; combining the actual characteristic data with the corresponding membership functions to calculate membership degrees of the actual characteristic data to different fuzzy marks, and determining membership values of fault probability according to the membership degree of the fuzzy marks;
S4, carrying out fuzzy processing on each actual characteristic data in a period to obtain a membership value matrix of each actual characteristic for different fault probabilities, combining all membership degree matrixes in the monitoring period to form a first membership value matrix, and further combining weights corresponding to each actual characteristic to obtain a second membership value matrix of the monitoring period evaluation set result;
s5, determining an evaluation result according to the comprehensive membership value of each evaluation result in the second membership value matrix, and carrying out corresponding maintenance early warning according to the evaluation result.
Specifically, the feature extraction of the current and the voltage includes: calculating the average value and variance of the current and the voltage; converting the time domain signal into a frequency domain signal through Fourier transformation, and extracting frequency domain characteristics including a frequency spectrum peak value and a frequency spectrum bandwidth; detecting transient changes of current and voltage waveforms, such as peak value, rising time and falling time;
For feature extraction of temperature, comprising: calculating the mean and variance of the temperature data; monitoring the rising rate and the falling rate of the temperature;
Feature extraction for noise, comprising: calculating the mean value, variance, maximum value and minimum value statistical index of the noise data; the time domain signal is converted into a frequency domain signal through Fourier transformation, and spectral features such as spectral peaks and spectral bandwidths are extracted.
Specifically, step S3 establishes a fuzzy model according to the following manner:
s31, performing fuzzy processing on the characteristics, including establishing a factor set and an evaluation set of fault evaluation; wherein, Represents a factor set, an，Representing the total number of extracted features; represents an evaluation set, an ，Representing the number of evaluation results, the evaluation set of the methodThe probability of failure is small, in the probability of failure, the probability of failure is high;
S32, inquiring data detection records in a historical work log, wherein the data detection records comprise values of various parameters and judgment of whether faults occur when an alternating-current/direct-current series-parallel power line works; respectively taking out a monitoring value when an alternating current-direct current series-parallel power line fails and a detection value when the alternating current-direct current series-parallel power line works normally, and extracting historical characteristics by analyzing voltage, current, temperature and noise data in historical data;
s33, carrying out normalization processing on each item of data by using a linear normalization method, wherein each item of data is the extracted historical characteristic data;
S34, theory domain Is thatRespectively counting the monitoring values of various historical characteristic data when an AC/DC series-parallel power line fails and the size of the monitoring values of various historical characteristic data during normal operation, wherein the upper interval of the monitoring values when the AC/DC series-parallel power line fails isThe lower interval is; The interval of the monitoring value in normal operation is；
S35, carrying out fuzzy marking on each item of data, and marking the quantity type data as follows: small, medium, large; for rate data, marking as low, medium and high, and establishing a quadratic parabolic membership function;
S36, determining a fuzzy rule according to the fault record data;
s37, calculating the membership degree of each monitored actual characteristic data and the corresponding fuzzy mark, determining the fault probability rule of the actual characteristic data according to the section where the actual characteristic data is located, and assigning the calculated membership degree to the corresponding fault probability; according to And calculating the membership degree of each fuzzy mark in the section, and if the calculated membership degree of the fuzzy marks is not equal to 0, determining the membership value of the fault probability according to the membership degree of the fuzzy marks.
Further, the quadratic parabolic membership function established in step S35 is as follows:
The membership functions of the features labeled "small" and "low" are:
the membership functions of the features labeled "middle" are:
The membership functions for features labeled "large" and "high" are:
wherein, Represent the firstSpecific monitoring data of item data.
Specifically, step S36 includes:
S361, combining the lower limit of the monitoring value interval of each item of history feature data in the fault with the upper limit of the monitoring value interval of each item of history feature data in the normal operation, and combining the upper limit of the monitoring value interval of each item of history feature data in the fault with the lower limit of the monitoring value interval of each item of history feature data in the normal operation to obtain 5 intervals, wherein the 5 intervals are respectively: 、、、 And ；
S362, dividing the whole membership function of each class of features into 7 intervals according to the 5 intervals and the boundaries of the discourse domainThe device corresponds to full small (low), multiple middle and fewer, full middle, multiple middle and multiple large (high) and full large (high) respectively;
S363, determining that the fault probability is relatively high when the fuzzy rule is that the fuzzy mark is small (low) or large (high), and relatively low when the fuzzy mark is medium; the specific rules are as follows:
(1) When historical characteristic data Belonging toWhen the fault probability of the AC/DC series-parallel power line is high;
(2) When historical characteristic data Belonging toWhen the fault probability of the AC/DC series-parallel power line is high or medium;
(3) When historical characteristic data Belonging toWhen the fault probability of the AC/DC series-parallel power line is low or medium;
(4) When historical characteristic data Belonging toWhen the fault probability of the AC/DC series-parallel power line is low;
(5) When historical characteristic data Belonging toWhen the fault probability of the AC/DC series-parallel power line is low or medium;
(6) When historical characteristic data Belonging toWhen the fault probability of the AC/DC series-parallel power line is medium or high;
(7) When historical characteristic data Belonging toAnd when the fault probability of the alternating current-direct current series-parallel power line is high.
Specifically, step S4 includes:
S41, substituting actual characteristic data of a current monitoring period into the membership function to obtain membership degrees of each fuzzy characteristic corresponding to each actual characteristic; obtaining a membership value matrix corresponding to the fault probability according to the membership degree ；
S42, combining the membership value matrixes of all the actual features to form a first membership value matrix of the current monitoring periodThe expression is:
s43, determining a weight matrix of each feature through an AHP analytic hierarchy process ；
S44, combining the first membership value matrix with the weight of each actual feature to obtain a second membership value matrix of the evaluation set result of the monitoring periodThe expression is:
wherein, I.e. the comprehensive membership value representing each evaluation result.
Specifically, step S5 includes:
second membership value matrix In (a) and (b)、And (3) withRespectively representing membership values with low fault probability, medium fault probability and high fault probability, and normalizing the membership values to obtain numerical values、、I.e., the probability of three sets of failure probabilities occurring;
If it is The data detected in the current period does not generate abnormality, and the current time of last maintenance of the AC/DC series-parallel power line is inquiredIf (if)Greater than a threshold for periodic maintenanceThe central control system sends periodic maintenance early warning to maintenance personnel; if it isNot greater than a threshold for periodic maintenanceNo maintenance early warning is sent;
If it is And is also provided withThe data detected in the current period is abnormal, and the central control system sends fault maintenance early warning to maintenance personnel;
If a situation other than the two situations occurs, it indicates that the detected data in the current period may be abnormal, and further judgment is needed.
Further, the further judging step is as follows:
s51, collecting a first membership value matrix In (a)Is characterized by (1)，To meet the above constraint condition, and to take out the corresponding membership function、And；
S52, calculating the similarity of membership functions of the actual features in pairsIf the similarity of membership functionsGreater than a set thresholdIf the two characteristics have correlation, if the characteristics with correlation exceed a certain number, the two characteristics are likely to be caused by abnormality of certain equipment or circuit in the AC/DC series-parallel power circuit; let two membership functions to be calculated beAnd (3) withSimilarity of the membership functionsIs calculated as follows:
If the actual characteristics are calculated And featuresMembership function similarity of less than a thresholdThenAnd (3) withIs of the same type and is only calculated subsequentlyMembership function similarity with other actual features;
S53, counting the number of the actual features of the same type, if the number of the actual features of a certain type is larger than a set threshold value If the fault occurs, the central control system sends a fault maintenance early warning to a maintenance person;
If there is no actual feature number of a certain type greater than the set threshold The actual features of the same type are marked, and the similarity calculation is not needed in the detection of the actual features of the same type in the next monitoring period.
(III) beneficial effects
The invention provides a fault maintenance early warning method for an AC/DC hybrid power line, which has the following beneficial effects:
1. When the working state of the alternating current-direct current series-parallel power line is monitored, various data such as voltage, current, temperature and noise are collected, the working state of the power line can be reflected more comprehensively, and therefore faults can be early warned more accurately; through comprehensive collection and processing of the data, the running state of the power line can be better known, and the accuracy of fault early warning is improved;
2. The collected data is subjected to comprehensive feature extraction, including statistical features, time-frequency features, frequency domain features and the like, so that fault information in the data can be extracted more fully, and the accuracy of fault early warning is improved; through deep feature extraction, potential fault information in the power line can be better mined, and the accuracy and reliability of fault early warning are improved;
3. The fuzzy processing is carried out on the characteristics by establishing a factor set and an evaluation set of fault evaluation, so that the accuracy of characteristic extraction is improved; by establishing a secondary parabolic membership function, fuzzy marking and membership degree calculation are carried out on the features, so that the actual state of the features can be reflected more accurately, and the accuracy of fault early warning is improved;
4. obtaining the occurrence probability of three groups of fault probabilities by normalizing the fault probability membership values in the second membership value matrix; the fault probability of the line is estimated more accurately, so that more accurate maintenance early warning is realized; judging the correlation between the features by calculating the membership function similarity of the features; the accuracy and the reliability of judgment are improved.
Drawings
Fig. 1 is a flow chart of steps of a fault maintenance and early warning method for an ac/dc hybrid power line.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
Referring to fig. 1, the invention provides a fault maintenance and early warning method for an ac/dc series-parallel power line, comprising the following steps:
S1, setting a monitoring period, collecting voltage, current, temperature and noise data in the working state of an alternating-direct current series-parallel power line in the monitoring period, and preprocessing the collected data;
The voltage and current can directly reflect the running state of the power line; under the condition that an alternating current-direct current series-parallel power line works normally, the voltage, the current and the temperature should be kept in a stable range; if an abnormality occurs in the voltage or current, such as too high, too low, or too large a fluctuation, it may mean that there is some kind of fault in the power line; such faults may come from the line itself or from the connected equipment; too high or too low a temperature may mean that there is a fault in the line, e.g. overheating of the line may result in damage to the insulation; therefore, monitoring of voltage, current, and temperature is an important means of preventing and early discovery of power line faults;
During normal operation of the power line, certain noise is usually associated with the power line; these noises may be caused by a magnetic field, an electric field, or the like generated by a current flowing in the wire. If the noise is excessive or abnormal, it may mean that some kind of fault exists in the line, such as poor contact, short circuit, etc.
S11, designing a sensor network, connecting with a central control system, and comprising: a current transformer, a voltage sensor, a temperature sensor and a noise sensor; the current transformer is connected with the voltage sensor and the alternating-current/direct-current hybrid power line, and the temperature sensor and the noise sensor are arranged near the alternating-current/direct-current hybrid power line;
s12, setting a monitoring period, and monitoring current and voltage data in an alternating-current/direct-current series-parallel power line, the temperature around the power line and the noise condition generated in real time in each detection period;
S13, preprocessing operation is carried out on the collected data, wherein the preprocessing operation comprises data cleaning, outlier processing, noise reduction and other operations; specifically, filters are used to remove noise, errors, or outlier data points that may be present, outlier detection algorithms are used to detect and process outliers in the acquired data.
The central control system is used for receiving the data of the sensor, processing and analyzing the data and sending the early warning.
S2, extracting features of the preprocessed data, including statistical features, time-frequency features and frequency domain features, and recording the features as actual features;
For the feature extraction of current and voltage, comprising: calculating the average value and variance of the current and the voltage, and reflecting the stability and fluctuation condition; converting the time domain signal into a frequency domain signal through Fourier transformation, and extracting frequency domain characteristics including a frequency spectrum peak value and a frequency spectrum bandwidth; detecting transient changes of current and voltage waveforms, such as peak value, rising time, falling time and the like;
For feature extraction of temperature, comprising: calculating the average value and variance of the temperature data, and reflecting the change and stability of the temperature; monitoring the rising rate and the falling rate of the temperature, and reflecting whether the temperature of the power line has abnormal change or not;
Feature extraction for noise, comprising: calculating the mean value, variance, maximum value and minimum value statistical index of the noise data, and reflecting the intensity and distribution condition of the noise; the time domain signal is converted into a frequency domain signal through Fourier transformation, and spectral features such as spectral peaks and spectral bandwidths are extracted.
The calculated statistical, time-frequency and frequency-domain features of each monitoring period are taken as actual features at the end of the period for subsequent blurring processing.
S3, extracting characteristics similar to S2 from voltage, current, temperature and noise data in the historical data, and recording the characteristics as historical characteristics; establishing a fuzzy model, setting model output as fault probability, performing fuzzy logic processing on normalized historical characteristic data to obtain a membership function, and formulating a corresponding fuzzy rule; combining the actual characteristic data with the corresponding membership functions to calculate membership degrees of the actual characteristic data to different fuzzy marks, and determining membership values of fault probability according to the membership degree of the fuzzy marks;
A fuzzy model is a model based on fuzzy logic and fuzzy reasoning that is capable of processing, reasoning and deciding on uncertain, inaccurate or fuzzy information. In fault pre-warning, the fuzzy model can be used for analyzing and processing data acquired by the sensor so as to identify possible faults or abnormal conditions. Performing fault evaluation on the extracted features based on the fuzzy model, specifically including:
s31, performing fuzzy processing on the characteristics, including establishing a factor set and an evaluation set of fault evaluation;
The factor set is a common set composed of elements of various factors affecting the evaluation object, namely the extracted features; using Representing the factor set, then，Representing the total number of extracted features;
The evaluation set is a set formed by various results possibly made on an evaluation object, namely judging whether the fault exists or not; using Representing the evaluation set, then，Representing the number of evaluation results, the evaluation set of the methodThe probability of failure is small, in the probability of failure, the probability of failure is high;
S32, inquiring data detection records in a historical work log, wherein the data detection records comprise values of various parameters and judgment of whether faults occur when an alternating-current/direct-current series-parallel power line works; respectively taking out a monitoring value when an alternating current-direct current series-parallel power line fails and a detection value when the alternating current-direct current series-parallel power line works normally, and extracting historical characteristics by analyzing voltage, current, temperature and noise data in historical data;
s33, carrying out normalization processing on each item of data by using a linear normalization method, wherein each item of data is the extracted historical characteristic data;
S34, theory domain Is thatNamely, the definition domain of the normalized characteristic is used for respectively counting the monitoring value of each item of history characteristic data when an AC/DC series-parallel power line fails and the size of the monitoring value of each item of history characteristic data during normal operation, wherein the upper interval of the monitoring value when the AC/DC series-parallel power line fails isThe lower interval is; The interval of the monitoring value in normal operation isWherein, the method comprises the steps of, wherein,Represent the firstSpecific monitoring data of item data;
S35, carrying out fuzzy marking on each item of data, and marking the quantity type data as follows: small, medium, large; for rate type data, labeled low, medium, high, a quadratic parabolic membership function is established, wherein membership functions for features labeled small and low are:
the membership functions of the features labeled as:
The membership functions for features marked large and high are:
S36, determining a fuzzy rule according to the fault record data, wherein the method comprises the following steps:
S361, combining the lower limit of the monitoring value interval of each item of history feature data in the fault with the upper limit of the monitoring value interval of each item of history feature data in the normal operation, and combining the upper limit of the monitoring value interval of each item of history feature data in the fault with the lower limit of the monitoring value interval of each item of history feature data in the normal operation to obtain 5 intervals, wherein the 5 intervals are respectively: 、、、 And ；
S362, dividing the whole membership function of each class of features into 7 intervals according to the 5 intervals and the boundaries of the discourse domainThe device corresponds to full small (low), multiple middle and fewer, full middle, multiple middle and multiple large (high) and full large (high) respectively;
S363, determining that the fault probability is relatively high when the fuzzy rule is that the fuzzy mark is small (low) or large (high), and relatively low when the fuzzy mark is medium; the specific rules are as follows:
(1) When historical characteristic data Belonging toWhen the fault probability of the AC/DC series-parallel power line is high;
(2) When historical characteristic data Belonging toWhen the fault probability of the AC/DC series-parallel power line is high or medium;
(3) When historical characteristic data Belonging toWhen the fault probability of the AC/DC series-parallel power line is low or medium;
(4) When historical characteristic data Belonging toWhen the fault probability of the AC/DC series-parallel power line is low;
(5) When historical characteristic data Belonging toWhen the fault probability of the AC/DC series-parallel power line is low or medium;
(6) When historical characteristic data Belonging toWhen the fault probability of the AC/DC series-parallel power line is medium or high;
(7) When historical characteristic data Belonging toWhen the fault probability of the AC/DC series-parallel power line is high;
s37, calculating the membership degree of each monitored actual characteristic data and the corresponding fuzzy mark, determining the fault probability rule of the actual characteristic data according to the section where the actual characteristic data is located, and assigning the calculated membership degree to the corresponding fault probability; according to Calculating the membership degree of each fuzzy mark in the section, and if the calculated membership degree of the fuzzy marks is not equal to 0, determining the membership value of the fault probability according to the magnitude relation of the membership degrees of the fuzzy marks; such as whenWhen it is calculated that its membership to small (low) isThe membership degree in the pair isWhileThe membership value of the data with high probability of power line fault isThe membership value in the fault probability is; When (when)When it is calculated that its membership to small (low) isThe membership degree in the pair isWhileThe membership value of the data in the probability of the power line fault isMembership value with low failure probability is。
S4, carrying out fuzzy processing on each actual characteristic data in a period to obtain a membership value matrix of each actual characteristic for different fault probabilities, combining all membership degree matrixes in the monitoring period to form a first membership value matrix, and further combining weights corresponding to each actual characteristic to obtain a second membership value matrix of the monitoring period evaluation set result;
S41, substituting actual characteristic data of a current monitoring period into the membership function to obtain membership degrees of each fuzzy characteristic corresponding to each actual characteristic; obtaining a membership value matrix corresponding to the fault probability according to the membership degree ；
S42, combining the membership value matrixes of all the actual features to form a first membership value matrix of the current monitoring periodThe expression is:
s43, determining a weight matrix of each actual feature through an AHP analytic hierarchy process ; The AHP weight analysis method is a qualitative and quantitative calculation weight research method, adopts a pairwise comparison method, establishes a matrix, utilizes the relativity of the number size, and finally calculates the importance of each factor by utilizing the principle that the larger the number is, the higher the important weight is; the method comprises the steps of carrying out weight analysis on selected features according to related data in a history log and whether faults exist or not to obtain the weight of each feature; this weight will change correspondingly as the data increases;
S44, combining the first membership value matrix with the weight of each actual feature to obtain a second membership value matrix of the evaluation set result of the monitoring period The expression is:
wherein, I.e. the comprehensive membership value representing each evaluation result.
S5, according to the second membership value matrixThe specific membership value of each evaluation result in the system is used for determining the evaluation result, and corresponding maintenance early warning is carried out according to the evaluation result.
Second membership value matrixIn (a) and (b)、And (3) withRespectively representing membership values with low fault probability, medium fault probability and high fault probability, and normalizing the membership values to obtain numerical values、、I.e., the probability of three sets of failure probabilities occurring;
If it is The data detected in the current period does not generate abnormality, and the current time of last maintenance of the AC/DC series-parallel power line is inquiredIf (if)Greater than a threshold for periodic maintenanceThe central control system sends periodic maintenance early warning to maintenance personnel; if it isNot greater than a threshold for periodic maintenanceNo maintenance early warning is sent;
If it is And is also provided withThe data detected in the current period is abnormal, and the central control system sends fault maintenance early warning to maintenance personnel;
if the conditions other than the two conditions are generated, the condition that the detected data in the current period is possibly abnormal is indicated, and further judgment is needed, wherein the specific judgment steps are as follows:
s51, collecting a first membership value matrix In (a)Is characterized by (1)，To meet the above constraint condition, and to take out the corresponding membership function、And；
S52, calculating the similarity of membership functions of the actual features in pairsIf the similarity of membership functionsGreater than a set thresholdIf the two characteristics have correlation, if the characteristics with correlation exceed a certain number, the two characteristics are likely to be caused by abnormality of certain equipment or circuit in the AC/DC series-parallel power circuit; let two membership functions to be calculated beAnd (3) withSimilarity of the membership functionsIs calculated as follows:
If the actual characteristics are calculated And featuresMembership function similarity of less than a thresholdThenAnd (3) withIs of the same type and is only calculated subsequentlyMembership function similarity with other actual features;
S53, counting the number of the actual features of the same type, if the number of the actual features of a certain type is larger than a set threshold value If the fault occurs, the central control system sends a fault maintenance early warning to a maintenance person;
If there is no actual feature number of a certain type greater than the set threshold The actual features of the same type are marked, and the similarity calculation is not needed in the detection of the actual features of the same type in the next monitoring period.
The foregoing is merely illustrative of the present application, and the present application is not limited thereto, and any person skilled in the art will readily recognize that variations or substitutions are within the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
The foregoing description of the preferred embodiments of the invention is not intended to limit the invention, but to enable any modification, equivalent or improvement to be made without departing from the spirit and principles of the invention.
Claims (10)
1. A fault maintenance and early warning method for an AC/DC series-parallel power line is characterized in that: comprising the following steps:
collecting voltage, current, temperature and noise data of an alternating-current/direct-current hybrid power line in a monitoring period, and preprocessing the collected data;
extracting features of the preprocessed data, including statistical features, time-frequency features and frequency-domain features, and recording the features as actual features; the same characteristic extraction is carried out on the voltage, the current, the temperature and the noise data in the historical data, and the historical data is recorded as the historical characteristic;
Establishing a fuzzy model, setting model output as fault probability, performing fuzzy logic processing on normalized historical characteristic data to obtain a membership function, and formulating a corresponding fuzzy rule; combining the actual characteristic data with the corresponding membership functions to calculate membership degrees of the actual characteristic data to different fuzzy marks, and determining membership values of fault probability according to the membership degree of the fuzzy marks;
Carrying out fuzzy processing on each actual characteristic data in a period to obtain a membership value matrix of each actual characteristic for different fault probabilities, combining all membership degree matrixes in the monitoring period to form a first membership value matrix, and further combining weights corresponding to each actual characteristic to obtain a second membership value matrix of the monitoring period evaluation set result;
And determining an evaluation result according to the comprehensive membership value of each evaluation result in the second membership value matrix, and carrying out corresponding maintenance early warning according to the evaluation result.
2. The method for early warning fault maintenance of the alternating current-direct current series-parallel power line is characterized in that:
The fuzzy model is established according to the following mode: performing fuzzy processing on the extracted various features, including establishing a factor set and an evaluation set of fault evaluation; wherein, Represents a set of factors, and/>，/>Representing the total number of extracted features; /(I)Represents an evaluation set, and/>，/>The number of evaluation results is represented, and the evaluation set/>The probability of failure is small, in the probability of failure, the probability of failure is high;
Inquiring a data detection record in a historical work log, wherein the data detection record comprises values of various parameters and judgment of whether faults occur when an alternating-current/direct-current hybrid power line works; and respectively taking out a monitoring value when the AC/DC series-parallel power line fails and a detection value when the AC/DC series-parallel power line works normally, recording the size of each monitored item of data, and carrying out linear normalization processing on each item of data, wherein each item of data is the extracted historical characteristic data.
3. The method for early warning fault maintenance of the alternating current-direct current series-parallel power line is characterized in that the method comprises the following steps of:
Theory domain For/>Respectively counting the monitoring values of various historical characteristic data when an AC/DC series-parallel power line fails and the size of the monitoring values of various historical characteristic data during normal operation, wherein the upper interval of the monitoring values when the AC/DC series-parallel power line fails is/>The lower interval is/>; The interval of monitoring values during normal operation is/>；
Each item of data is subjected to fuzzy marking, and for the quantity type data, the marking is as follows: small, medium, large; for rate type data, labeled low, medium, high, a quadratic parabolic membership function is established.
4. The method for early warning fault maintenance of an ac/dc series-parallel power line according to claim 3, wherein:
Combining the lower limit of the monitoring value interval of each item of history feature data in the fault with the upper limit of the monitoring value interval of each item of history feature data in the normal operation, and combining the upper limit of the monitoring value interval of each item of history feature data in the fault with the lower limit of the monitoring value interval of each item of history feature data in the normal operation to obtain 5 intervals, wherein the 5 intervals are respectively: 、/>、/>、/> /> ; And determining the fuzzy rule of 7 sections consisting of the boundaries of the 5 sections and the discourse domain.
5. The method for early warning fault maintenance of the alternating current-direct current series-parallel power line according to claim 4 is characterized in that:
Calculating the membership degree of each item of actual characteristic data and the corresponding fuzzy mark, determining the fault probability rule of the actual characteristic data according to the section where the actual characteristic data is located, and assigning the calculated membership degree to the corresponding fault probability; according to And calculating the membership degree of each fuzzy mark in the section, and if the calculated membership degree of the fuzzy marks is not equal to 0, determining the membership value of the fault probability according to the membership degree of the fuzzy marks.
6. The method for early warning fault maintenance of the alternating current-direct current series-parallel power line according to claim 5 is characterized in that:
Substituting the actual characteristic data of the current monitoring period into the secondary parabolic membership function to obtain membership degrees of each fuzzy characteristic corresponding to each actual characteristic; obtaining a membership value matrix corresponding to the fault probability according to the membership degree ；
Combining the membership value matrixes of all the actual features to form a first membership value matrix of the current monitoring periodThe expression is:
；
Determining a weight matrix of each actual feature by AHP analytic hierarchy process 。
7. The method for early warning fault maintenance of the alternating current-direct current series-parallel power line according to claim 6 is characterized in that:
matrix the first membership value Weight matrix/>, with actual featuresCombining to obtain a second membership value matrix/>, of the evaluation set result of the monitoring periodThe expression is: /(I)Wherein/>、/>、/>I.e. the comprehensive membership value representing each evaluation result.
8. The method for early warning fault maintenance of the alternating current-direct current series-parallel power line according to claim 7 is characterized in that:
The membership value matrix In/>、/>And/>Respectively representing membership values with low fault probability, medium fault probability and high fault probability, and carrying out normalization treatment on the membership values to obtain a numerical value/>、/>、/>I.e., the probability of three sets of failure probabilities occurring;
If it is Inquiring the time/>, from the last overhaul of an alternating current/direct current series-parallel power lineIf/>Greater than the threshold of periodic maintenance/>Sending periodic maintenance early warning to maintenance personnel; if/>Not greater than the threshold value of periodic maintenance/>No maintenance early warning is sent;
If it is And/>The data detected in the current period is abnormal, and a fault maintenance early warning is sent to maintenance personnel;
if a situation other than the two situations occurs, a judgment instruction is sent.
9. The method for early warning fault maintenance of the alternating current-direct current series-parallel power line is characterized in that:
After receiving the judging instruction, collecting a first membership value matrix Middle/>Actual characteristics/>，/>To meet the number of the characteristics of the constraint conditions, and to take out the corresponding membership function/>、/>/>；
Calculating the similarity of membership functions of the actual features in pairsIf the similarity of membership functions/>Greater than a set thresholdLet two membership functions to be calculated be/>And/>Similarity/>, of the membership functionsIs calculated as follows:
；
If the actual characteristics are calculated And/>Membership function similarity of less than threshold/>Only calculate/>Membership function similarity to other actual features.
10. The method for early warning fault maintenance of the alternating current-direct current series-parallel power line according to claim 9 is characterized in that:
counting the number of the actual features of the same type, if the number of the actual features of the same type is larger than a set threshold value The method comprises the steps of indicating that a certain device or circuit in an alternating-current/direct-current series-parallel power circuit is about to be abnormal, and sending a fault maintenance early warning to a maintenance person;
If there is no actual feature number of a certain type greater than the set threshold These same types of actual features are labeled, and no similarity calculation is required in the detection of these same types of actual features in the next monitoring period. /(I)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410310863.2A CN117911011B (en) | 2024-03-19 | 2024-03-19 | AC/DC series-parallel power line fault maintenance early warning method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410310863.2A CN117911011B (en) | 2024-03-19 | 2024-03-19 | AC/DC series-parallel power line fault maintenance early warning method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117911011A true CN117911011A (en) | 2024-04-19 |
| CN117911011B CN117911011B (en) | 2024-05-28 |
Family
ID=90689275
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410310863.2A Active CN117911011B (en) | 2024-03-19 | 2024-03-19 | AC/DC series-parallel power line fault maintenance early warning method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117911011B (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101051738A (en) * | 2007-02-06 | 2007-10-10 | 天津大学 | Negative sequence direction longitudinal protective method for very high voltage transmission line |
| US20100085669A1 (en) * | 2008-10-03 | 2010-04-08 | General Electric Company | Arc fault detection using fuzzy logic |
| CN102621434A (en) * | 2012-04-19 | 2012-08-01 | 山东大学 | Nonlinear fuzzy detection method for operating safety of power cable tunnel |
| CN102967798A (en) * | 2012-11-15 | 2013-03-13 | 深圳大学 | Failure warning method and system of power device |
| CN111178725A (en) * | 2019-12-24 | 2020-05-19 | 贵州电网有限责任公司 | Protective equipment state early warning method based on analytic hierarchy process |
| CN113325308A (en) * | 2021-04-13 | 2021-08-31 | 北京中大科慧科技发展有限公司 | Power supply fault detection method for data center |
| CN114202304A (en) * | 2021-12-06 | 2022-03-18 | 中国南方电网有限责任公司超高压输电公司南宁监控中心 | Intelligent monitoring processing method and system for power grid faults |
| CN116961215A (en) * | 2023-06-19 | 2023-10-27 | 蒋洁 | Rapid fault response processing method for power system |
| CN117131946A (en) * | 2023-08-30 | 2023-11-28 | 广东电网有限责任公司 | Energy storage battery fault early warning method and device, electronic equipment and medium |
| CN117371207A (en) * | 2023-10-11 | 2024-01-09 | 国网宁夏电力有限公司 | Extra-high voltage converter valve state evaluation method, medium and system |
-
2024
- 2024-03-19 CN CN202410310863.2A patent/CN117911011B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101051738A (en) * | 2007-02-06 | 2007-10-10 | 天津大学 | Negative sequence direction longitudinal protective method for very high voltage transmission line |
| US20100085669A1 (en) * | 2008-10-03 | 2010-04-08 | General Electric Company | Arc fault detection using fuzzy logic |
| CN102621434A (en) * | 2012-04-19 | 2012-08-01 | 山东大学 | Nonlinear fuzzy detection method for operating safety of power cable tunnel |
| CN102967798A (en) * | 2012-11-15 | 2013-03-13 | 深圳大学 | Failure warning method and system of power device |
| CN111178725A (en) * | 2019-12-24 | 2020-05-19 | 贵州电网有限责任公司 | Protective equipment state early warning method based on analytic hierarchy process |
| CN113325308A (en) * | 2021-04-13 | 2021-08-31 | 北京中大科慧科技发展有限公司 | Power supply fault detection method for data center |
| CN114202304A (en) * | 2021-12-06 | 2022-03-18 | 中国南方电网有限责任公司超高压输电公司南宁监控中心 | Intelligent monitoring processing method and system for power grid faults |
| CN116961215A (en) * | 2023-06-19 | 2023-10-27 | 蒋洁 | Rapid fault response processing method for power system |
| CN117131946A (en) * | 2023-08-30 | 2023-11-28 | 广东电网有限责任公司 | Energy storage battery fault early warning method and device, electronic equipment and medium |
| CN117371207A (en) * | 2023-10-11 | 2024-01-09 | 国网宁夏电力有限公司 | Extra-high voltage converter valve state evaluation method, medium and system |
Non-Patent Citations (2)
| Title |
|---|
| 李岩;滕云;冷欧阳;苑舜;蔡志远;: "数据驱动的输电线路在线监测装置可靠性评估", 中国电机工程学报, no. 15, 5 August 2018 (2018-08-05) * |
| 梁敏;李兴源;: "基于模糊理论的电缆故障风险预估模型", 四川电力技术, no. 06, 20 December 2015 (2015-12-20) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117911011B (en) | 2024-05-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US6192317B1 (en) | Statistical pattern analysis methods of partial discharge measurements in high voltage insulation | |
| CN109933905B (en) | Mechanical equipment health state assessment method based on multi-dimensional early warning analysis | |
| CN103617568B (en) | Setting method for abnormal data determination threshold in steady-state power quality early-warning mechanism | |
| CN111209934B (en) | Fan fault pre-alarm method and system | |
| CN115018139A (en) | Current transformer error state online identification method and system based on interphase characteristics | |
| CN106972626A (en) | The running status inspection method of power equipment, apparatus and system | |
| CN112327100B (en) | Power failure detection method and system based on Internet of things | |
| CN103103570B (en) | Based on the aluminium cell condition diagnostic method of pivot similarity measure | |
| CN105548744A (en) | Substation equipment fault identification method based on operation-detection large data and system thereof | |
| CN115170000A (en) | Remote monitoring method and system based on electric energy meter communication module | |
| CN103699668A (en) | Power distribution network electric equipment combination state evaluation method based on data section consistency | |
| CN116993329B (en) | Communication equipment operation maintenance decision management system based on data analysis | |
| CN116800199A (en) | Distributed photovoltaic operation electric energy quality monitoring and analyzing method | |
| CN111598165A (en) | Density clustering outlier detection method based on extreme learning machine | |
| CN116315173A (en) | Battery temperature sampling system based on new energy automobile | |
| CN116771610A (en) | Method for adjusting fault evaluation value of variable pitch system of wind turbine | |
| CN115051310A (en) | Intelligent power-off switch system and method for circuit fault analysis | |
| CN117560300B (en) | Intelligent internet of things flow prediction and optimization system | |
| CN117911011B (en) | AC/DC series-parallel power line fault maintenance early warning method | |
| Dong et al. | Fault diagnosis and classification in photovoltaic systems using scada data | |
| CN109784777B (en) | Power grid equipment state evaluation method based on time sequence information fragment cloud similarity measurement | |
| CN114167837B (en) | Intelligent fault diagnosis method and system for railway signal system | |
| CN117031201A (en) | Multi-scene topology anomaly identification method and system for power distribution network | |
| CN109507468B (en) | Header box branch current detection method and system based on correlation characteristics | |
| CN113848347A (en) | Health state detection method for wind meter of wind driven generator |
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 |