CN113934982B - Method for predicting mechanical life of breaker operating mechanism based on vibration-electric signal fusion - Google Patents
Method for predicting mechanical life of breaker operating mechanism based on vibration-electric signal fusion Download PDFInfo
- Publication number
- CN113934982B CN113934982B CN202111207687.2A CN202111207687A CN113934982B CN 113934982 B CN113934982 B CN 113934982B CN 202111207687 A CN202111207687 A CN 202111207687A CN 113934982 B CN113934982 B CN 113934982B
- Authority
- CN
- China
- Prior art keywords
- degradation
- circuit breaker
- key
- operating mechanism
- energy storage
- 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
- 230000007246 mechanism Effects 0.000 title claims abstract description 85
- 238000000034 method Methods 0.000 title claims abstract description 70
- 230000004927 fusion Effects 0.000 title claims abstract description 27
- 238000006731 degradation reaction Methods 0.000 claims abstract description 149
- 230000015556 catabolic process Effects 0.000 claims abstract description 135
- 238000004146 energy storage Methods 0.000 claims abstract description 86
- 230000036541 health Effects 0.000 claims abstract description 47
- 230000006870 function Effects 0.000 claims abstract description 40
- 230000009471 action Effects 0.000 claims abstract description 38
- 230000008569 process Effects 0.000 claims abstract description 29
- 238000012544 monitoring process Methods 0.000 claims abstract description 22
- 238000007619 statistical method Methods 0.000 claims abstract description 7
- 230000009467 reduction Effects 0.000 claims abstract description 5
- 238000012216 screening Methods 0.000 claims abstract description 4
- 230000003595 spectral effect Effects 0.000 claims description 25
- 239000011159 matrix material Substances 0.000 claims description 18
- 238000003646 Spearman's rank correlation coefficient Methods 0.000 claims description 12
- 239000013598 vector Substances 0.000 claims description 9
- 238000012886 linear function Methods 0.000 claims description 8
- 238000012545 processing Methods 0.000 claims description 7
- 238000013507 mapping Methods 0.000 claims description 4
- 230000001419 dependent effect Effects 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 239000000284 extract Substances 0.000 abstract description 2
- 238000012360 testing method Methods 0.000 description 19
- 238000004458 analytical method Methods 0.000 description 10
- 238000012423 maintenance Methods 0.000 description 9
- 101000812677 Homo sapiens Nucleotide pyrophosphatase Proteins 0.000 description 7
- 102100039306 Nucleotide pyrophosphatase Human genes 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 230000000694 effects Effects 0.000 description 6
- 238000000513 principal component analysis Methods 0.000 description 6
- 238000011160 research Methods 0.000 description 6
- 101100353526 Neurospora crassa (strain ATCC 24698 / 74-OR23-1A / CBS 708.71 / DSM 1257 / FGSC 987) pca-2 gene Proteins 0.000 description 5
- 238000007689 inspection Methods 0.000 description 5
- 230000008859 change Effects 0.000 description 4
- 230000007850 degeneration Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000005070 sampling Methods 0.000 description 4
- 238000003745 diagnosis Methods 0.000 description 3
- 238000012417 linear regression Methods 0.000 description 3
- 238000013459 approach Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000012512 characterization method Methods 0.000 description 2
- 238000011156 evaluation Methods 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000000611 regression analysis Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 description 1
- 108091033411 PCA3 Proteins 0.000 description 1
- 235000010627 Phaseolus vulgaris Nutrition 0.000 description 1
- 244000046052 Phaseolus vulgaris Species 0.000 description 1
- 238000005299 abrasion Methods 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 230000001925 catabolic effect Effects 0.000 description 1
- 238000013499 data model Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011835 investigation Methods 0.000 description 1
- 238000005184 irreversible process Methods 0.000 description 1
- 238000004445 quantitative analysis Methods 0.000 description 1
- 238000011158 quantitative evaluation Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/02—Reliability analysis or reliability optimisation; Failure analysis, e.g. worst case scenario performance, failure mode and effects analysis [FMEA]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/04—Ageing analysis or optimisation against ageing
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Mathematical Optimization (AREA)
- Computational Mathematics (AREA)
- Pure & Applied Mathematics (AREA)
- Mathematical Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Operations Research (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Probability & Statistics with Applications (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computer Hardware Design (AREA)
- Computing Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Geometry (AREA)
- Testing Electric Properties And Detecting Electric Faults (AREA)
Abstract
The invention relates to a mechanical life prediction method of a breaker operating mechanism based on vibration-electric signal fusion, which comprises the following steps of collecting a current signal of a motor and a vibration signal of an energy storage operating mechanism in an energy storage process of a universal breaker to obtain state monitoring data; extracting time domain features of the current signals and time domain and frequency domain features of the vibration signals, and screening out key degradation features; secondly, carrying out statistical analysis on all key degradation characteristics to obtain a degradation starting point; thirdly, performing dimension reduction and fusion on key degradation features in a degradation period by adopting a PCA method, selecting a main component with the largest contribution rate as a fusion feature, and taking the normalized fusion feature as a comprehensive health index; fitting the degradation process of the comprehensive health index along with the increase of the action times by utilizing multiple unary regression function models; and selecting the function model with the smallest fitting error as a prediction model. The method extracts various signal characteristics for fusion, and comprehensively reflects the degradation state from multiple angles.
Description
Technical Field
The invention belongs to the technical field of online prediction of residual life of a universal circuit breaker, and particularly relates to a mechanical life prediction method of a circuit breaker operating mechanism based on vibration-electric signal fusion.
Background
The universal circuit breaker is the circuit breaker with the largest capacity in a low-voltage power distribution system, and is widely applied to the total incoming line position or important equipment position of the power distribution system. Under the action of internal and external factors, the health state of the universal circuit breaker shows a declining trend in the operation process, and when declining reaches a certain degree, normal functions cannot be completed, so that the real-time health state of the circuit breaker needs to be known and the residual life is predicted, and the operation reliability of the circuit breaker is improved. The investigation result of the reliability of the circuit breaker shows that the duty ratio of mechanical faults in various fault causes is the largest, and the reliable operation of the energy storage operating mechanism is the premise of normal opening and closing of the circuit breaker in consideration of the operation principle of the universal circuit breaker, so that the mechanical faults are most likely to occur in the energy storage operating mechanism, the energy storage operating mechanism is also the key of the normal operation of the universal circuit breaker, and the residual life prediction of the universal circuit breaker can be equivalently used as the residual life prediction of the energy storage operating mechanism. If the residual service life of the energy storage operating mechanism can be timely and accurately predicted according to the monitoring information in the performance degradation process of the energy storage operating mechanism, the running reliability of the circuit breaker can be greatly improved, faults are avoided, meanwhile, the downtime can be effectively reduced, the maintenance period is shortened, the maintenance steps are simplified, and the maintenance cost is reduced.
At present, the research on the prediction of the residual life of the circuit breaker is mainly focused on faults of the high-voltage circuit breaker, the research objects are mainly contact systems and switching-on and switching-off operation accessories of the circuit breaker, and the single signals are analyzed and researched, such as Zhao Lihua and the like (Zhao Lihua, jin Haowen, huang Xiaolong, and the like; bean Longjiang (Doulongjiang. Breaker spring operating mechanism failure mechanism analysis and diagnosis method research [ D ]. Beijing: north China university of electric power, 2019.) utilizes vibration signals to conduct breaker spring operating mechanism failure mechanism analysis and diagnosis method research; zhao Shutao et al (Zhao Shutao, xu Wenjie, li Yunpeng, etc. vacuum circuit breaker spring mechanism stored energy state identification method based on preferential general characteristics [ J/OL ]. High voltage technology.); the above-mentioned researches are mainly focused on dominant fault diagnosis, but the research on the energy storage process is not yet deep enough.
The reliable operation of the energy storage operating mechanism is the premise of normal opening and closing of the circuit breaker, namely, any link of energy storage has an influence on opening and closing operation of the circuit breaker, and the mechanical performance of the energy storage operating mechanism of the universal circuit breaker is degraded due to abrasion, jamming and the like during the actual operation, no faults are generated at the moment, so that the performance degradation trend of the energy storage operating mechanism is predicted in real time, and further, the residual service life is predicted to be very necessary, and the faults can be avoided.
Most of the existing residual life prediction methods are artificial intelligence methods, a large number of training samples are usually needed in the methods, typical degradation data is used as a training sample set to establish a prediction model, and further the residual life of a sample to be detected is predicted.
The statistical data model method mainly comprises wiener theory, regression analysis and the like, and is characterized in that (Sun S G,Wang Q F,Du T H,et al.Quantitative Evaluation of electrical life ofAC contactor based on initial characteristic parameters[J].IEEE Transactions on Instrumentation and Measurement,2021,70:1-10.) such as Sun Shuguang and the like utilizes Spearman rank correlation coefficient to analyze the correlation between the electrical parameter characteristics and the service life of the contactor, the characteristics are fused to obtain a service life comprehensive evaluation index, and finally a service life prediction model based on unitary regression analysis is established. The method does not need prior data accumulation, establishes a prediction model based on historical degradation data, and performs off-line estimation and on-line update on model parameters by using state monitoring data.
Disclosure of Invention
Aiming at the defects of the prior art, the invention aims to provide a mechanical life prediction method of a breaker operating mechanism based on vibration-electric signal fusion.
The technical scheme adopted for solving the technical problems is as follows:
a method for predicting mechanical life of a breaker operating mechanism based on vibration-electric signal fusion is characterized by comprising the following steps:
the method comprises the steps of firstly, collecting a current signal of a motor and a vibration signal of an energy storage operating mechanism in an energy storage process of a universal circuit breaker to obtain state monitoring data; extracting time domain features of the current signals and time domain and frequency domain features of the vibration signals, and screening out key degradation features with high service life correlation degree;
secondly, carrying out statistical analysis on all key degradation characteristics to obtain a degradation starting point of the universal circuit breaker;
Thirdly, performing dimension reduction and fusion on key degradation features in a degradation period by adopting a PCA method, selecting a main component with the largest contribution rate as a fusion feature, and taking the normalized fusion feature as a comprehensive health index of the universal circuit breaker;
fitting the degradation process of the comprehensive health index along with the increase of the action times by utilizing multiple unary regression function models; and selecting a function model with the minimum fitting error as a prediction model for predicting the residual life of the universal circuit breaker.
The specific process of the third step is as follows:
3-1, setting the number of samples as n, wherein each sample contains m key degradation characteristics, and the total characteristic data matrix X n×m of the state monitoring data in the formula (9) is provided; column vector x j of the total feature data matrix corresponds to a certain key degradation feature column of all samples, x j＝(xj(1),xj(2),…,xj (n)), and j is more than 0 and less than or equal to m;
normalizing the key degradation features using equation (10);
In the method, in the process of the invention, The value after the key degradation characteristic x j (i) is normalized is more than 0 and less than or equal to n; /(I)The mean value of the j-th key degradation feature; s (x j) is the standard deviation of the j-th column key degradation feature;
3-2, setting the total characteristic data matrix after the standardized processing as Then/>The covariance matrix P of (2) is:
Calculating eigenvalues lambda j and eigenvectors e j of the covariance matrix P according to equation (12);
P＝EDET (12)
Wherein D is a diagonal matrix arranged in descending order of eigenvalue, d=diag (λ 1,λ2,…,λj,…,λm); e is the set of feature vectors E j corresponding to the feature value lambda j ,E＝(e1,e2,…em),e1＝(e11,e12,…,e1m),e2＝(e21,e22,…,e2m),em＝(em1,em2,…,emm);
Fusing all key degradation characteristics through a formula (13) to obtain all main components y 1,y2,…,ym;
The contribution rate α j of the j-th principal component is:
And taking the main component with the largest contribution rate as a fusion characteristic and carrying out normalization processing.
The fourth step is specifically as follows:
Taking the action times z as independent variables, taking the comprehensive health index y as dependent variables, fitting by using linear function, logarithmic function, exponential function and power function models respectively, and selecting an exponential function model with the minimum fitting error as a prediction model along with the degradation process of the comprehensive health index along with the increase of the action times;
The mapping relation between the comprehensive health index and the action times is as follows:
When the comprehensive health index exceeds the failure threshold value, indicating that the universal circuit breaker fails; the action number when the comprehensive health index exceeds the failure threshold value is recorded as z, namely the total service life of the universal circuit breaker is z;
the remaining life of the universal circuit breaker satisfies the formula (22):
z Residual life time ＝z-z the current number of actions (22)
In the prediction process, parameters of a prediction model are updated through the acquired state monitoring data, and the action times when the comprehensive health index exceeds the failure threshold value are obtained based on the degradation model, so that the residual life times of the universal circuit breaker are obtained according to a formula (22).
In the first step, the time domain features include average value, standard deviation, skewness, kurtosis, peak-to-peak value, root mean square, amplitude factor, form factor, impact factor, margin factor, energy and signal duration, and the frequency domain features of the vibration signal include spectral kurtosis average value, spectral kurtosis standard deviation, spectral kurtosis and spectral kurtosis skewness;
acquiring key degradation characteristics with high performance degradation correlation degree by utilizing a Spearman rank correlation coefficient;
Assuming that (X k,Yk) is a sample taken from the state monitoring data (X, Y), the elements in X, Y are respectively ranked in order from small to large, R k represents the rank order of X k in X, Q k represents the rank order of Y k in Y, and Spearman rank correlation coefficient R of the sample (X k,Yk) is calculated by formula (2);
in formula (2), k=1, 2, …, n, n is the total number of samples of the state monitoring data;
When |r| is less than or equal to 0.3, the correlation between X and Y does not exist; x and Y have low correlation when r is less than or equal to 0.3 and less than or equal to 0.5; when r is less than or equal to 0.5 and less than or equal to 0.8, obvious correlation exists between X and Y; when r is more than or equal to 0.8, the X and Y have high correlation; the characteristic that X and Y have obvious correlation and high correlation is taken as a key degradation characteristic with high performance degradation correlation of the universal circuit breaker.
Compared with the prior art, the invention has the beneficial effects that:
The method has the outstanding substantial characteristics that: in order to realize the online prediction of the residual life of the universal circuit breaker, the operation reliability of the universal circuit breaker is improved, the multi-signal characteristics are extracted to reflect the mechanical state of the universal circuit breaker, namely, not only the motor current signal but also the vibration signal of the energy storage operating mechanism are extracted, and the degradation state of the universal circuit breaker is reflected in a multi-angle all-around manner; acquiring key degradation characteristics with high service life correlation by using a Spearman rank correlation coefficient, and then fusing the key degradation characteristics by using Principal Component Analysis (PCA) to obtain a comprehensive health index; the method can realize quantitative prediction of the residual mechanical life, has engineering practicability, realizes online quantitative prediction evaluation of the residual life of the universal circuit breaker, and can effectively improve the operation and maintenance efficiency of the universal circuit breaker. The method extracts and fuses the current signal and the vibration signal in the energy storage process, quantifies the degradation process of the energy storage operating mechanism of the universal circuit breaker, and the fused comprehensive health index can reflect the microscopic degradation process of the mechanism better than the single characteristic.
The method of the invention has the remarkable advantages that:
(1) According to the invention, the current signal of the universal circuit breaker motor and the vibration signal of the energy storage operating mechanism are selected as analysis signals, and the multi-dimensional characteristic parameters of the time domain and the frequency domain of the current signal and the vibration signal are calculated, so that the test result error caused by single characteristic measurement deviation can be effectively reduced. The motor current signal is easy to detect, contains abundant mechanical state information of the energy storage operating mechanism and the interlocking mechanism thereof, comprehensively reflects the mechanical state of the mechanism by combining the mechanical vibration signal of the energy storage process, and can well explain the action change mechanism of the energy storage operating mechanism.
(2) On the basis of extracting multidimensional characteristic parameters, all the characteristics are not selected to serve as degradation characteristics of the universal energy storage operating mechanism, key degradation characteristics with high service life correlation degree are obtained by utilizing spearman rank correlation coefficients, and the key degradation characteristics are selected for subsequent analysis, so that the mechanical state of the energy storage operating mechanism is accurately reflected.
(3) Compared with the conventional normal distribution inspection, which only analyzes from a qualitative angle, the method can reduce the influence of subjective factors and improve the accuracy and effectiveness of inspection by changing qualitative to quantitative.
(4) According to the invention, the service life of the energy storage operating mechanism is not predicted directly, but whether the energy storage operating mechanism is mechanically degenerated is judged, and the degeneration phenomenon of the mechanism is judged by the fact that the parameter value of the key degeneration characteristic exceeds the parameter range of the normal period, so that the residual service life of the mechanism is predicted.
(5) According to the invention, the residual life of the energy storage operating mechanism of the circuit breaker is predicted by using a unitary linear regression method, and the performance prediction model can be built by using the state monitoring information and the index prediction model of a single energy storage operation, so that the energy storage operating mechanism is not required to be physically modeled, a large number of test samples are not required, and a large amount of time and cost are saved.
(6) The method solves the key problems of the conventional circuit breaker prediction and health management (Prognostics ANDHEALTH MANAGEMENT, PHM) technology, provides technical guidance for the circuit breaker maintenance according to conditions, promotes the transition of the circuit breaker maintenance technology from timing inspection to the condition maintenance technology, improves the reliability of the circuit breaker, and greatly reduces the daily maintenance cost due to the safety and the utilization rate.
Drawings
FIG. 1 is a general flow chart of the present invention;
FIG. 2 is a schematic diagram of a universal circuit breaker life test system;
FIG. 3 (a) is a graph showing the amplitude variation of the current signal;
FIG. 3 (b) is a graph of vibration signal amplitude variation;
FIG. 4 (a) is a graph of the standard deviation variation of the current signal;
FIG. 4 (b) is a graph of the crest factor variation of the current signal;
FIG. 4 (c) is a graph of the energy variation of the current signal;
FIG. 4 (d) is a graph of the duration of the current signal;
FIG. 5 (a) is a graph of the duration of a vibration signal;
FIG. 5 (b) is a graph of the standard deviation of spectral kurtosis of a vibration signal;
FIG. 5 (c) is a graph of spectral kurtosis variation of a vibration signal;
FIG. 5 (d) is a graph of spectral kurtosis variation of a vibration signal;
FIG. 6 is a Spearman rank correlation coefficient plot for various features and lifetimes;
FIG. 7 is a heat map of the various features and Spearman rank correlations;
FIG. 8 (a) is a probability density distribution diagram of the standard deviation of the current signal;
FIG. 8 (b) is a probability density distribution plot of the energy of the current signal;
FIG. 8 (c) is a probability density distribution plot of the duration of the current signal;
FIG. 8 (d) is a probability density distribution plot of the standard deviation of spectral kurtosis of a vibration signal;
fig. 9 (a) is a normal distribution verification result diagram of the standard deviation of the current signal;
FIG. 9 (b) is a graph of the normal distribution test result of the energy of the current signal;
FIG. 9 (c) is a graph of the normal distribution test result for the duration of the current signal;
fig. 9 (d) is a normal distribution test result diagram of the spectral kurtosis standard deviation of the vibration signal;
FIG. 10 is a graph showing the comparison of principal components PCA1 and PCA 2;
FIG. 11 is a graph of degradation resulting from fitting different functional models;
FIG. 12 (a) is a graph showing the trend of the remaining life of the energy storage operating mechanism;
FIG. 12 (b) is a graph of probability density for remaining life of the energy storage operating mechanism;
FIG. 13 is a graph comparing the predicted remaining life using a single signal feature and the integrated health indicator of the present invention.
Detailed Description
The technical scheme of the present application is further described below with reference to specific embodiments and drawings, but the scope of the present application is not limited thereto.
The invention provides a method for predicting the mechanical life of a breaker operating mechanism based on vibration-electric signal fusion (a method for short), which comprises the following steps:
firstly, collecting a current signal of a motor and a vibration signal of an energy storage operating mechanism in an energy storage process of the universal circuit breaker to obtain state monitoring (condition monitoring, abbreviated as CM) data, wherein the state monitoring data are recorded as (X, Y) X, Y are vectors, X represents characteristics, and Y is the service life of the universal circuit breaker; extracting time domain features of the current signals and time domain and frequency domain features of the vibration signals, and screening out key degradation features with high service life correlation by using Spearman rank correlation coefficients;
1-1, extracting time domain characteristics of a current signal and a vibration signal; the time domain features comprise 12 of average value, standard deviation, skewness, kurtosis, peak-to-peak value, root mean square, amplitude factor, form factor, impact factor, margin factor, energy and signal duration, and the calculation formulas of the time domain features are shown in table 1; in table 1, s i is the amplitude corresponding to the sampling time i, s max、smin is the maximum amplitude and the minimum amplitude respectively, N is the total number of sampling times, t start、tend is the start and stop time of signal sampling respectively, X1-X12 represent 12 time domain features of the current signal, and X13-X24 represent 12 time domain features of the vibration signal;
TABLE 1 time domain characterization
1-2, Extracting frequency domain characteristics of vibration signals; the frequency domain features are described by statistical indexes derived from spectral kurtosis (Spectral Kurtosis, abbreviated as SK), wherein the statistical indexes comprise a spectral kurtosis average value, a spectral kurtosis standard deviation, a spectral kurtosis and a spectral kurtosis skewness, and the calculation formula is shown in table 2; wherein the spectral kurtosis SK is calculated by formula (1);
In the formula (1), f is frequency, t is time, SK is vector, S (t, f) is amplitude of vibration signal after short time Fourier transform under specific moment and frequency;
TABLE 2 frequency domain characterization
In Table 2, N f is the dimension of the vector SK, and X25-X28 represent the four frequency domain characteristics of the vibration signal;
1-3, acquiring key degradation features with high performance degradation correlation degree by utilizing a Spearman rank correlation coefficient based on the time domain features and the frequency domain features extracted in the steps 1-1 and 1-2;
Assuming that (X k,Yk) is a sample taken from the state monitoring data (X, Y), the elements in X, Y are respectively ranked in order from small to large, R k represents the rank order of X k in X, Q k represents the rank order of Y k in Y, and Spearman rank correlation coefficient R of the sample (X k,Yk) is calculated by formula (2);
in formula (2), k=1, 2, …, n, n is the total number of samples of the state monitoring data;
when |r| is less than or equal to 0.3, the correlation between X and Y does not exist; x and Y have low correlation when r is less than or equal to 0.3 and less than or equal to 0.5; when r is less than or equal to 0.5 and less than or equal to 0.8, obvious correlation exists between X and Y; when r is more than or equal to 0.8, the X and Y have high correlation; taking the characteristic that X and Y have obvious correlation and high correlation as a key degradation characteristic with high performance degradation correlation of the universal circuit breaker;
Secondly, respectively carrying out statistical analysis on all key degradation characteristics, calculating a normal period interval of each key degradation characteristic, and if the key degradation characteristic value is located outside the normal period interval, considering that the universal circuit breaker enters a degradation period; the action times of the universal circuit breaker entering the degradation period are reflected by all key degradation characteristics to serve as degradation starting points;
2-1, respectively carrying out statistical analysis on each key degradation characteristic in the normal operation process of the energy storage operating mechanism, and solving the mean value mu and the standard deviation sigma of each key degradation characteristic to obtain a normal period interval [ mu-3 sigma, mu+3 sigma ] corresponding to each key degradation characteristic; checking whether the key degradation features conform to normal distribution from a qualitative angle by utilizing Quantile-Quantile (Q-Q) diagram, if the key degradation features conform to normal distribution, the normal distribution of the key degradation features is consistent with standard normal quantile distribution, and the Q-Q diagram is approximate to a straight line;
2-2, after qualitative normal distribution inspection is completed, normal distribution inspection is carried out on key degradation characteristics from a quantitative angle by using a skewness coefficient and a kurtosis coefficient; the skewness coefficient is used for representing an index of the degree of asymmetry of the probability distribution density curve relative to the average value, and the skewness coefficient b S of the key degradation characteristic X is calculated according to the formula (3):
wherein E (X) represents the mean, var (X) represents the variance, The 3 rd order central moment of the key degradation characteristic X is represented, and sigma is the standard deviation; for a normal distribution probability density curve, its bias coefficient b S =0;
Standard error S S of the skewness coefficient b S is calculated according to equation (4):
The test formula of the bias coefficient b S obtained by the formula (3) and the formula (4) is as follows:
Wherein Z S represents the test value of b S, and when the significance level α=0.05, if Z S is less than 1.96, the skewness coefficient b S is considered to be significantly equal to zero, i.e., the sample data obeys normal distribution;
Similar to the skewness factor, the kurtosis factor is used to characterize the index of the probability density curve peak at the average, and the kurtosis factor b K for the key degradation feature X is calculated according to equation (6):
where E represents the mean, var represents the variance, A central moment of order 4 representing a key degradation feature X; for a probability density curve of a normal distribution, the kurtosis coefficient b K =0;
The standard error S K of the kurtosis coefficient b K is calculated according to equation (7):
the test formula for the kurtosis coefficient b K obtainable by the formula (6) and the formula (7) is:
Wherein Z K represents the test value of b K, and when the significance level α=0.05, if Z K is less than 1.96, the kurtosis coefficient b K is considered to be significantly equal to zero, i.e., the sample data obeys a normal distribution;
Therefore, when both the test value Z S of the skewness coefficient and the test value Z K of the kurtosis coefficient are smaller than 1.96, the key degradation feature X is considered to be subject to normal distribution; when the key degradation feature obeys the normal distribution N (mu, sigma 2), the probability that the value of the key degradation feature falls to the interval [ mu-3 sigma, mu+3 sigma ] is 99.73%; if the value of the key degradation characteristic falls outside the interval, the value is regarded as a small probability event, so that the universal circuit breaker is judged to enter the degradation period; the action times of the universal circuit breaker entering the degradation period are reflected by all key degradation characteristics to serve as degradation starting points;
Thirdly, performing dimension reduction and fusion on key degradation characteristics in a degradation period by adopting a PCA method, and extracting the main component with the most abundant degradation information to obtain degradation performance indexes, namely comprehensive health indexes, of the universal circuit breaker;
3-1, assuming that each sample contains m key degradation features, a total feature data matrix X n×m of the state monitoring data of the formula (9); column vector x j of the total feature data matrix corresponds to a certain key degradation feature column of all samples, x j＝(xj(1),xj(2),…,xj (n)), and j is more than 0 and less than or equal to m;
in order to eliminate data differences caused by different dimensions, the key degradation characteristics are standardized by using a formula (10);
In the method, in the process of the invention, The value after the key degradation characteristic x j (i) is normalized is more than 0 and less than or equal to n; /(I)The mean value of the j-th key degradation feature; s (x j) is the standard deviation of the j-th column key degradation feature;
3-2, setting the total characteristic data matrix after the standardized processing as Then/>The covariance matrix P of (2) is:
Calculating eigenvalues lambda j and eigenvectors e j of the covariance matrix P according to equation (12);
P＝EDET (12)
Wherein D is a diagonal matrix arranged in descending order of eigenvalue, d=diag (λ 1,λ2,…,λj,…,λm); e is the set of feature vectors E j corresponding to the feature value lambda j ,E＝(e1,e2,…em),e1＝(e11,e12,…,e1m),e2＝(e21,e22,…,e2m),em＝(em1,em2,…,emm);
Fusing all key degradation characteristics through a formula (13) to obtain all main components y 1,y2,…,ym;
The contribution rate α j of the j-th principal component is:
Taking the main component with the largest contribution rate as a fusion characteristic, carrying out normalization processing, taking the normalized fusion characteristic as a comprehensive health index for judging the residual life of the universal circuit breaker, and marking the main component with the largest contribution rate as y, namely the comprehensive health index y;
Fitting a degradation process of comprehensive health indexes along with the increase of the action times by utilizing a plurality of function models; selecting a function model with the minimum fitting error as a prediction model for predicting the residual life of the universal circuit breaker;
4-1, taking the action times z as independent variables, taking the comprehensive health index y as dependent variables, and respectively utilizing linear functions, logarithmic functions, exponential functions and power function model fitting to carry out degradation processes of the comprehensive health index along with the increase of the action times; in the process, a logarithmic function, an exponential function and a power function model are required to be transformed into a linear function according to a transformation mode in a table 3, and unitary linear regression analysis is carried out on the action times and the comprehensive health indexes through the linear function, so that a mapping relation between the comprehensive health indexes and the action times meets a formula (15), namely a degradation curve of the follow-up action times of the universal circuit breaker is obtained;
TABLE 3 linearization equation for nonlinear function model
Wherein, beta 0、β1 is the slope and intercept, epsilon is the random error, E (epsilon) is the mean value of the random error, D (epsilon) is the variance of the random error, and sigma is the standard deviation of the random error;
Assuming that the observed value of the kth action and the integrated health indicator is (z k,yk), k=1, 2, …, n; obtaining an estimated value of the parameter β 0、β1 by using the least square method of equation (16) Let observation y k and regression/>The sum of the squares of the errors is minimal;
Wherein the method comprises the steps of Satisfy formula (17):
In the method, in the process of the invention, Average value of z k、yk;
in summary, the mapping relationship between the comprehensive health index and the action times satisfies the formula (18):
When the comprehensive health index exceeds the failure threshold value, indicating that the universal circuit breaker fails; the action number when the comprehensive health index exceeds the failure threshold value is recorded as z, namely the total service life of the universal circuit breaker is z;
4-2, calculating fitting errors of the function models, wherein the indexes of the fitting errors comprise Root Mean Square Errors (RMSE), average absolute errors (MAE) and average errors (MRE), and the formulas (19) - (21) are respectively satisfied; when the three errors are minimum, the fitting error of the function model is minimum, and the function model is used as a prediction model for predicting the residual life of the universal circuit breaker;
the remaining life of the universal circuit breaker satisfies the formula (22):
z Residual life time ＝z-z the current number of actions (22)
In the actual prediction process, the slope of the prediction model is updated through the acquired state monitoring data Intercept/>And obtaining the action times when the comprehensive health index exceeds the failure threshold value based on the prediction model, thereby obtaining the residual life times of the universal circuit breaker according to the formula (22).
Example 1
As shown in FIG. 2, the universal circuit breaker life test system comprises a Hall current sensor, a vibration sensor, a USB-7648A acquisition card, an industrial personal computer LabVIEW platform, a PCL-720+ board card and a solid state relay set; the industrial personal computer LabVIEW platform controls the PCL-720+ board card to operate the solid-state relay group through the ISA bus to control the action process of the universal circuit breaker, wherein the switching-on, switching-off, energy storage and undervoltage of the universal circuit breaker are controlled by a switching-on relay, a switching-off relay, an energy storage relay and an undervoltage relay in the solid-state relay group respectively; the current signal of the motor is collected by a CHB-50SF Hall current sensor, the vibration signal of the energy storage operating mechanism is collected by a vibration sensor, then the analog signal collected by the Hall current sensor and the analog signal collected by the vibration sensor are converted into digital signals through a USB-7648A collecting card and are uploaded to an industrial personal computer LabVIEW platform through a USB bus, and the industrial personal computer LabVIEW platform adopts a program developed by MATLAB software to conduct data processing and residual life prediction.
In the embodiment, an energy storage operating mechanism of the DW15-1600 type universal circuit breaker is used as a test object, the energy storage operating mechanism bears the function of controlling the energy storage of the universal circuit breaker, and the premise is that the energy storage operating mechanism is required to normally store energy in opening and closing of the universal circuit breaker; in the energy storage process, the motor drives the eccentric wheel to rotate, so that the connecting rod and the pawl do up and down reciprocating motion, the ratchet wheel, the pin and the cam are pushed to do rotary motion, the cam is connected with the square shaft, and the square shaft rotates to drive the energy storage operating mechanism to move, so that the spring stores energy; the energy storage motor adopts an alternating current power supply mode, and combines the industry standard of the universal circuit breaker and the performance index of the energy storage process, and the embodiment sets the energy storage threshold value of the energy storage operating mechanism as a parameter when the energy storage time reaches 1350 ms.
The mechanical life prediction method of the breaker operating mechanism based on vibration-electric signal fusion of the embodiment comprises the following steps:
firstly, acquiring a current signal of an energy storage motor and a vibration signal of an energy storage operating mechanism in an energy storage process by using a service life test system of the universal circuit breaker, and setting a test operating frequency to 20 times/h according to a related standard of the universal circuit breaker, wherein the universal circuit breaker runs in an idle state, the time length of each signal acquisition is 2.5s, and the sampling frequency is 20kHz, so as to obtain the current signal and the vibration signal shown in the figures 3 (a) and 3 (b); in fig. 3, the current signal mainly shows that the peak-to-valley amplitude gradually changes, and the vibration signal also changes along with the mechanical action process; time t 0: the motor is started just before, and the starting current is larger so as to obtain larger torque; stage t1-t 2: the motor stably runs to drive the energy storage operating mechanism to stretch the spring, and the current is basically kept constant; in the stage t2-t3, the load torque is further increased and the current is obviously increased due to the larger spring stretching amount; in the stage t3-t4, the spring stretches to approach to the limit position, the current is obviously increased, then the energy storage process is finished, and the current returns to zero; whereas for vibration signals, the t0-t1 phase: when the motor is started, the spring starts to be stretched, and the relative mechanical stress and the motion quantity among all parts of the energy storage operating mechanism are smaller, so that the amplitude of the vibration signal is smaller; stage t1-t 2: along with the gradual stretching of the energy storage spring by the energy storage motor, the amplitude of the vibration signal is gradually increased, and the phase t2-t4 is as follows: the spring is gradually stretched to the limit position, the relative mechanical stress and the motion degree between the parts are increased, the vibration signal is further increased, and the peak value is reached at the moment t 4; therefore, the load of the energy storage operating mechanism and the mechanical characteristic change of the energy storage operating mechanism in the energy storage process can be reflected to a certain extent through time domain analysis of the current signal and the vibration signal.
In order to comprehensively analyze the operation state of an energy storage operation mechanism of the universal circuit breaker, 12 time domain features of a current signal are extracted through a table 1; extracting 16 features in total from the time domain and the frequency domain of the vibration signal by table 1 and table 2, wherein the characteristic variation curves of the standard deviation, the crest factor, the energy and the duration of the current signal are shown in fig. 4 (a) -4 (d); characteristic variation curves of the duration of the vibration signal, the spectral kurtosis standard deviation, the spectral kurtosis and the spectral kurtosis skewness are shown in fig. 5 (a) -5 (d); as can be seen from the figure, the degradation capability of the energy storage operating mechanism represented by each feature is different, and three features in fig. 4 (b) and fig. 5 (c) and 5 (d) are basically stable in the life span; however, the five features in fig. 4 (a), 4 (c), 4 (d) and fig. 5 (a), 5 (b) show a significant trend of increasing in the lifetime, so that the feature indicating strong degradation capability of the energy storage operating mechanism needs to be screened out; the correlation between 28 features and the service life of the energy storage operating mechanism is analyzed by using the Spearman rank correlation coefficient, the analysis results are shown in fig. 6 and 7, the feature with higher correlation degree is selected as the key degradation feature with high correlation degree with the service life of the energy storage operating mechanism, and the standard deviation X2, the energy X11, the duration X12 and the spectral kurtosis standard deviation X26 of the vibration signal are finally selected as the key degradation features for subsequent analysis.
Secondly, carrying out statistical analysis on all key degradation characteristics to distinguish a normal period and a degradation period of the energy storage operating mechanism, wherein the change trend of the 4 key degradation characteristics of fig. 4 (a), 4 (c), 4 (d) and 5 (b) can be obviously seen, the values of all the characteristics in the early stage fluctuate in a small range in the life test process of the energy storage operating mechanism, the values have relatively stable distribution, the performance degradation of the energy storage operating mechanism is not obvious, and the energy storage operating mechanism is in the normal period; the value fluctuation of each characteristic in the later period is obviously increased, the performance degradation of the energy storage operating mechanism is obvious, the characteristic distribution is correspondingly changed, and the degradation period is entered; therefore, the normal period and the degradation period of the energy storage operating mechanism need to be distinguished, and the starting point of degradation is monitored to start the residual life prediction, so that the effect of improving the residual life prediction precision can be achieved;
carrying out probability statistical analysis on 4 key degradation features in the previous 2000 action processes to obtain probability density distribution curves shown in fig. 8 (a) -8 (d), wherein the probability density of each key degradation feature can be found to accord with normal distribution in fig. 8;
Carrying out normal distribution test on each key degradation characteristic from a qualitative angle, and checking whether each key degradation characteristic accords with normal distribution or not through Quantile-Quantile graphs, if so, approximating a Q-Q graph to a straight line, and obtaining results shown in fig. 9 (a) -9 (d); as is evident from the figure, each key degradation feature substantially conforms to a normal distribution;
The analysis visually describes the distribution of each key degradation characteristic from a qualitative aspect, and quantitative analysis is carried out on whether each key degradation characteristic obeys normal distribution or not by adopting a skewness coefficient and a kurtosis coefficient; when the significance level is 0.05, if the absolute values of the skewness coefficient and the kurtosis coefficient are smaller than 1.96, the key degradation features can be considered to be subjected to normal distribution, the results shown in the table 4 are obtained, and the skewness coefficient and the kurtosis coefficient of each key degradation feature are smaller than 1.96, namely each key degradation feature is subjected to normal distribution.
TABLE 4 skewness coefficient and kurtosis coefficient of key degradation features
And (5) obtaining a normal period interval: for the normal period, when each key degeneration feature obeys the normal distribution N (mu, sigma 2), the probability that the value of the key degeneration feature falls into the interval [ mu-3 sigma, mu+3 sigma ] is 99.73%; if the value of the key degradation characteristic falls outside the interval, the value can be regarded as a small probability event, so that the degradation of the running state of the energy storage operating mechanism is judged, and the degradation period is entered; the normal period interval [ mu-3σ, mu+3σ ] of each key degradation feature shown in table 5 is calculated, and this interval is taken as a defining threshold of normal period and degradation period, and in order to eliminate the influence of random noise, if each key degradation feature has a period exceeding the normal period interval [ mu-3σ, mu+3σ ] 2 times in succession, it is a degradation start point, and fig. 4 (a), (c), (d) and fig. 5 (b) show that the degradation start point is when the number of actions is 2017.
TABLE 5 Normal phase intervals for key degradation features
Thirdly, performing dimension reduction and fusion on 4 key degradation features by adopting a PCA method, wherein the degradation process of an energy storage operating mechanism is difficult to comprehensively reflect by a single feature, and the selected different features are different in numerical value, and have larger reflection difference on the running state of the energy storage operating mechanism, so that each key degradation feature is subjected to standardized treatment firstly, and then feature fusion is performed by adopting the PCA method to obtain the contribution rate of 4 main components, wherein the contribution rate of a first main component PCA1 is 92.8%, the contribution rate of a second main component PCA2 is 7.1%, and as the contribution rates of the main components PCA3 and PCA4 are close to zero, the difference can be ignored, and only the extracted PCA1 and PCA2 information is analyzed to obtain a result shown in a figure 10; comparing PCA1 and PCA2, it was found that the value of PCA2 decreased and increased with increasing number of lives after entering the degradation period, while the value of PCA1 increased continuously, i.e. monotonicity with changing number of actions was stronger. In the whole service life process, maintenance is not carried out on the universal circuit breaker, so that the degradation process of an energy storage operating mechanism of the universal circuit breaker is regarded as an irreversible process, therefore, the PCA1 can better represent the characteristic that the component value is increased along with the approach of failure of the energy storage operating mechanism, and the PCA1 is selected as a fused characteristic value and is also used as a comprehensive health index;
And fourthly, utilizing a unitary linear regression model theory, respectively utilizing linear function, logarithmic function, exponential function and power function model fitting to synthesize the degradation process of the health index along with the increase of the action times, and selecting a function model with the minimum fitting error for predicting the residual life of the energy storage operating mechanism of the universal circuit breaker. According to the transformation relation in table 3, each function model is converted into a linear function, and the comprehensive health index degradation curve of the energy storage operating mechanism shown in fig. 11 is obtained. As can be seen from fig. 11, compared with the comprehensive health index degradation curve, the slope of the linear function is unchanged, and the fitting effect is poor in the early stage and the later stage of degradation; the logarithmic function has poor fitting effect in the later period of degradation, the power function has good fitting effect in the earlier period of degradation, but the fitting effect in the middle period and the later period of degradation is poor, and the fitting effect of the exponential function in the whole life span is best through comprehensive analysis.
Table 6 shows the fitting error of each function model, and as can be seen from table 6, the fitting error of the exponential function model is the smallest, that is, in the life test process, the degradation process of the energy storage operating mechanism of the universal circuit breaker most accords with the change trend of the exponential function model, so that the exponential function model is selected as a prediction model for predicting the residual life of the energy storage operating mechanism.
TABLE 6 fitting errors for models
After the degradation starting point, updating the slope and intercept parameters of the prediction model through the acquired state monitoring data, determining the critical action times of the comprehensive health index value exceeding the failure threshold value based on the parameters, and obtaining the residual life times of the energy storage operating mechanism according to a formula (22).
FIG. 12 is a graph showing the result of predicting the single remaining life, taking the operation to 2570 times as an example, after the parameters of the prediction model are updated by using the monitored data in FIG. 12 (a), the degradation trend of the energy storage operating mechanism is predicted, and the number of operations when the comprehensive health index value exceeds the failure threshold value is 2740 times and the number of operations when the comprehensive health index value is actually failed is 2727 times by using the prediction model; fig. 12 (b) is a probability density graph of the remaining life of the energy storage operating mechanism, the upper and lower limits of the 95% confidence interval are reasonable intervals of the remaining life of the energy storage operating mechanism predicted by the model, and as can be seen from fig. 12 (b), the predicted value of the remaining life is 170 times, the actual value is 157 times, and the prediction error is 13 times.
FIG. 13 is a comparison of the results of predicting the overall remaining life of the catabolic operating mechanism using the integrated health indicator, the single current signal characteristic and the single vibration signal characteristic, respectively; selecting the current duration and the standard deviation of the spectral kurtosis of the vibration signal as a single current signal characteristic and a single vibration signal respectively; after the action times are 2300, comparing the difference between the predicted residual life of the comprehensive health index and the predicted residual life of the two single features with the actual residual life, it can be found that the predicted residual life of the comprehensive health index is closest to the actual residual life, and the difference between the predicted value and the actual value is reduced along with the increase of the monitored data quantity, and at the moment, the predicted value of the residual life is within a 95% confidence interval, which indicates that the error is within a reasonable range.
TABLE 7 prediction error of residual Life at each stage
Table 7 shows the prediction errors of the residual life of each stage under different characteristics, and as can be seen from Table 7, the prediction errors of the method provided by the invention are minimum, so that the residual life of the energy storage operating mechanism of the universal circuit breaker can be effectively predicted, and the prediction precision is higher.
The steps are implemented by using LabVIEW and MATLAB software, and the LabVIEW and MATLAB software are well known to those skilled in the art. The percentages in the above examples are all numerical percentages.
The invention is applicable to the prior art where it is not described.
Claims (2)
1. A method for predicting mechanical life of a breaker operating mechanism based on vibration-electric signal fusion is characterized by comprising the following steps:
the method comprises the steps of firstly, collecting a current signal of a motor and a vibration signal of an energy storage operating mechanism in an energy storage process of a universal circuit breaker to obtain state monitoring data; extracting time domain features of the current signals and time domain and frequency domain features of the vibration signals, and screening out key degradation features with high service life correlation degree;
secondly, carrying out statistical analysis on all key degradation characteristics to obtain a degradation starting point of the universal circuit breaker;
Thirdly, performing dimension reduction and fusion on key degradation features in a degradation period by adopting a PCA method, selecting a main component with the largest contribution rate as a fusion feature, and taking the normalized fusion feature as a comprehensive health index of the universal circuit breaker;
3-1, setting the number of samples as n, wherein each sample contains m key degradation characteristics, and the total characteristic data matrix X n×m of the state monitoring data in the formula (9) is provided; column vector x j of the total feature data matrix corresponds to a certain key degradation feature column of all samples, x j＝(xj(1),xj(2),…,xj (n)), and j is more than 0 and less than or equal to m;
normalizing the key degradation features using equation (10);
In the method, in the process of the invention, The value after the key degradation characteristic x j (i) is normalized is more than 0 and less than or equal to n; /(I)The mean value of the j-th key degradation feature; s (x j) is the standard deviation of the j-th column key degradation feature;
3-2, setting the total characteristic data matrix after the standardized processing as Then/>The covariance matrix P of (2) is:
Calculating eigenvalues lambda j and eigenvectors e j of the covariance matrix P according to equation (12);
P＝EDET (12)
Wherein D is a diagonal matrix arranged in descending order of eigenvalue, d=diag (λ 1,λ2,…,λj,…,λm); e is the set of feature vectors E j corresponding to the feature value lambda j ,E＝(e1,e2,…em),e1＝(e11,e12,…,e1m),e2＝(e21,e22,…,e2m),em＝(em1,em2,…,emm);
Fusing all key degradation characteristics through a formula (13) to obtain all main components y 1,y2,…,ym;
The contribution rate α j of the j-th principal component is:
taking the main component with the largest contribution rate as a fusion characteristic and carrying out normalization processing;
Fitting the degradation process of the comprehensive health index along with the increase of the action times by utilizing multiple unary regression function models; selecting a function model with the minimum fitting error as a prediction model for predicting the residual life of the universal circuit breaker;
Taking the action times z as independent variables, taking the comprehensive health index y as dependent variables, fitting by using linear function, logarithmic function, exponential function and power function models respectively, and selecting an exponential function model with the minimum fitting error as a prediction model along with the degradation process of the comprehensive health index along with the increase of the action times;
The mapping relation between the comprehensive health index and the action times is as follows:
When the comprehensive health index exceeds the failure threshold value, indicating that the universal circuit breaker fails; the action number when the comprehensive health index exceeds the failure threshold value is recorded as z, namely the total service life of the universal circuit breaker is z;
the remaining life of the universal circuit breaker satisfies the formula (22):
z Residual life time ＝z-z the current number of actions (22)
In the prediction process, parameters of a prediction model are updated through the acquired state monitoring data, and the action times when the comprehensive health index exceeds the failure threshold value are obtained based on the prediction model, so that the residual life times of the universal circuit breaker are obtained according to a formula (22).
2. The method for predicting mechanical life of a circuit breaker operating mechanism based on vibration-electric signal fusion according to claim 1, wherein in the first step, the time domain features include average value, standard deviation, skewness, kurtosis, peak-to-peak value, root mean square, amplitude factor, form factor, impact factor, margin factor, energy and signal duration, and the frequency domain features of the vibration signal include spectral kurtosis average value, spectral kurtosis standard deviation, spectral kurtosis and spectral kurtosis skewness;
acquiring key degradation characteristics with high performance degradation correlation degree by utilizing a Spearman rank correlation coefficient;
Assuming that (X k,Yk) is a sample taken from the state monitoring data (X, Y), the elements in X, Y are respectively ranked in order from small to large, R k represents the rank order of X k in X, Q k represents the rank order of Y k in Y, and Spearman rank correlation coefficient R of the sample (X k,Yk) is calculated by formula (2);
in formula (2), k=1, 2, …, n, n is the total number of samples of the state monitoring data;
When |r| is less than or equal to 0.3, the correlation between X and Y does not exist; x and Y have low correlation when r is less than or equal to 0.3 and less than or equal to 0.5; when r is less than or equal to 0.5 and less than or equal to 0.8, obvious correlation exists between X and Y; when r is more than or equal to 0.8, the X and Y have high correlation; the characteristic that X and Y have obvious correlation and high correlation is taken as a key degradation characteristic with high performance degradation correlation of the universal circuit breaker.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111207687.2A CN113934982B (en) | 2021-10-18 | 2021-10-18 | Method for predicting mechanical life of breaker operating mechanism based on vibration-electric signal fusion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111207687.2A CN113934982B (en) | 2021-10-18 | 2021-10-18 | Method for predicting mechanical life of breaker operating mechanism based on vibration-electric signal fusion |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113934982A CN113934982A (en) | 2022-01-14 |
| CN113934982B true CN113934982B (en) | 2024-04-19 |
Family
ID=79279988
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111207687.2A Active CN113934982B (en) | 2021-10-18 | 2021-10-18 | Method for predicting mechanical life of breaker operating mechanism based on vibration-electric signal fusion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113934982B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115758094B (en) * | 2023-01-09 | 2023-05-09 | 河北工业大学 | Method for extracting degradation index and predicting service life of speed reducer |
| CN116070527B (en) * | 2023-03-07 | 2023-06-16 | 南京航空航天大学 | Milling cutter residual life prediction method based on degradation model |
| CN116595449B (en) * | 2023-06-06 | 2023-12-12 | 西安科技大学 | Asynchronous motor fault diagnosis method based on improved SSA optimization support vector machine |
| CN116822383A (en) * | 2023-08-31 | 2023-09-29 | 成都态坦测试科技有限公司 | Equipment life prediction model construction method and device, readable storage medium and equipment |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101701881A (en) * | 2009-11-10 | 2010-05-05 | 广州锐翔电力工程设备有限公司 | Mechanical property state monitor of circuit breaker |
| CN108760266A (en) * | 2018-05-31 | 2018-11-06 | 西安交通大学 | The virtual degeneration index building method of mechanical key component based on learning distance metric |
| AU2020102377A4 (en) * | 2020-09-23 | 2020-11-12 | Chordiya, S. B. Prof.(Dr.) | CNI-Water Management Technology: WATER COLLECTION AND NOTIFICATION INTELLIGENT MANAGEMENT TECHNOLOGY |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106323373A (en) * | 2016-08-12 | 2017-01-11 | 工业和信息化部电子第五研究所 | Electronic equipment health monitoring and early warning system and electronic equipment health monitoring and early warning method |
-
2021
- 2021-10-18 CN CN202111207687.2A patent/CN113934982B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101701881A (en) * | 2009-11-10 | 2010-05-05 | 广州锐翔电力工程设备有限公司 | Mechanical property state monitor of circuit breaker |
| CN108760266A (en) * | 2018-05-31 | 2018-11-06 | 西安交通大学 | The virtual degeneration index building method of mechanical key component based on learning distance metric |
| AU2020102377A4 (en) * | 2020-09-23 | 2020-11-12 | Chordiya, S. B. Prof.(Dr.) | CNI-Water Management Technology: WATER COLLECTION AND NOTIFICATION INTELLIGENT MANAGEMENT TECHNOLOGY |
Non-Patent Citations (1)
| Title |
|---|
| 燕晨耀.基于多特征量的滚动轴承退化状态评估和剩余寿命预测方法研究.全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113934982A (en) | 2022-01-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113934982B (en) | Method for predicting mechanical life of breaker operating mechanism based on vibration-electric signal fusion | |
| CN109444791B (en) | Error state evaluation method and system for capacitor voltage transformer | |
| CN112200032B (en) | Online monitoring method for mechanical state of high-voltage circuit breaker based on attention mechanism | |
| CN109034191B (en) | ELM-based one-dimensional telemetry data abnormal interpretation method | |
| CN113125095B (en) | Universal circuit breaker contact system residual mechanical life prediction method based on deep learning | |
| CN111382809B (en) | Isolating switch mechanical fault diagnosis method based on motor output power | |
| CN116756595B (en) | Conductive slip ring fault data acquisition and monitoring method | |
| CN111275295B (en) | Distributed photovoltaic fault diagnosis method based on inverse distance weight interpolation | |
| CN112131797B (en) | Main shaft bearing life prediction and reliability assessment method based on stress analysis | |
| CN115877205A (en) | Intelligent fault diagnosis system and method for servo motor | |
| CN113918642A (en) | Data filtering, monitoring and early warning method based on power Internet of things equipment | |
| CN114722895A (en) | Photovoltaic array fault diagnosis method based on non-invasive state detection | |
| CN115600928A (en) | OAO-RVM-based GIS basin-type insulator state evaluation method | |
| CN115728628A (en) | On-line monitoring and early warning method and device for circuit breaker control loop fault | |
| CN115659768A (en) | Method and device for predicting residual service life of IGBT | |
| US11867763B2 (en) | Method for monitoring circuit breaker and apparatus and internet of things using the same | |
| Zhang et al. | Abnormal identification of dissolved gas in oil monitoring device based on multivariate statistical process monitoring | |
| CN111537889A (en) | Data-driven echelon battery RUL prediction and classification method | |
| CN112710972A (en) | GIS contact poor contact fault diagnosis method, device, system and storage medium | |
| CN112231624A (en) | Real-time evaluation system for short-circuit resistance of multi-transformer winding based on Internet of things | |
| CN113721142B (en) | Method and device for evaluating failure risk of circuit breaker | |
| CN117951633B (en) | Photovoltaic power generation equipment fault diagnosis method and system | |
| Pei et al. | Mechanical remaining Life Prediction of Circuit Breaker based on long-term and short-term memory Neural Network | |
| Ma et al. | An Online Detection Method of Short Circuit for Battery Packs | |
| Wei et al. | Research on transmission line trip prediction based on logistic regression algorithm under icing condition |
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 |