CN112578213A - Fault prediction method and device for rail power supply screen - Google Patents
Fault prediction method and device for rail power supply screen Download PDFInfo
- Publication number
- CN112578213A CN112578213A CN202011535641.9A CN202011535641A CN112578213A CN 112578213 A CN112578213 A CN 112578213A CN 202011535641 A CN202011535641 A CN 202011535641A CN 112578213 A CN112578213 A CN 112578213A
- Authority
- CN
- China
- Prior art keywords
- power supply
- fault
- sample
- target
- circuit parameter
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 63
- 230000002159 abnormal effect Effects 0.000 claims abstract description 85
- 238000001514 detection method Methods 0.000 claims abstract description 49
- 238000012360 testing method Methods 0.000 claims abstract description 30
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 27
- 239000013598 vector Substances 0.000 claims description 50
- 238000012549 training Methods 0.000 claims description 31
- 230000001932 seasonal effect Effects 0.000 claims description 13
- 238000004590 computer program Methods 0.000 claims description 11
- 238000002372 labelling Methods 0.000 claims description 9
- 239000011159 matrix material Substances 0.000 claims description 7
- 238000000354 decomposition reaction Methods 0.000 claims description 6
- 238000012423 maintenance Methods 0.000 description 15
- 230000007774 longterm Effects 0.000 description 9
- 238000010586 diagram Methods 0.000 description 5
- 238000012544 monitoring process Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 230000005540 biological transmission Effects 0.000 description 3
- 238000004891 communication Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 2
- 239000004576 sand Substances 0.000 description 2
- 230000032683 aging Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 230000008878 coupling Effects 0.000 description 1
- 238000010168 coupling process Methods 0.000 description 1
- 238000005859 coupling reaction Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 238000007670 refining Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012706 support-vector machine Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/005—Testing of electric installations on transport means
- G01R31/008—Testing of electric installations on transport means on air- or spacecraft, railway rolling stock or sea-going vessels
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/28—Testing of electronic circuits, e.g. by signal tracer
- G01R31/282—Testing of electronic circuits specially adapted for particular applications not provided for elsewhere
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Supply And Distribution Of Alternating Current (AREA)
Abstract
The invention provides a fault prediction method and a fault prediction device for a rail power supply screen, wherein the method comprises the following steps: performing anomaly detection on the circuit parameter time sequence samples based on an anomaly detection algorithm to obtain abnormal samples, and performing fault type marking on the abnormal samples; acquiring a target circuit parameter sample according to the abnormal sample and the fault type marking data; acquiring a fault prediction model according to the target circuit parameter sample; and inputting the circuit parameters and the prediction period of the target power supply screen in the test set into the fault prediction model to obtain the probability of the target power supply screen failing. The device is used for executing the method. According to the fault prediction method and device for the track power supply screen, provided by the invention, the circuit parameters of the track power supply screen are subjected to abnormity detection, the fault type of abnormal data is labeled, and a time sequence characteristic method is combined to construct an accurate power supply screen fault prediction model, so that the automatic prediction of the power supply screen fault is realized, and the fault prediction accuracy of the track power supply screen is improved.
Description
Technical Field
The invention relates to the technical field of rail transit, in particular to a fault prediction method and device of a rail power supply screen.
Background
The railway power supply screen is an important device for supplying power to railway signal equipment, whether the safe and stable operation of the power supply screen can directly influence the driving safety, and the power supply screen is influenced by factors such as self component aging and environment in work, so that frequent faults can be caused, and the driving safety is further seriously influenced. Therefore, the running state of the power supply screen is monitored, the fault prediction is carried out, the fault point is found in advance, and the maintenance of the power supply screen can play an extremely important role.
At present, most of power supply screen monitoring is completed by a signal centralized monitoring system, monitored parameters of the power supply screen monitoring system can not meet the requirements for predicting the time and the fault type of the power supply screen fault, and most of the current fault prediction methods stay in setting related threshold values and are carried out according to a comparison method. There are significant limitations and limited data utilization.
For the operation and maintenance of urban rail transit power supply screen equipment, the operation and maintenance data that operation and maintenance personnel relied on are mainly concentrated on, the input and output monitoring of each power distribution module of power supply screen, and data is comparatively single, sets up the threshold value interval when current power supply screen failure prediction technology used mostly moreover, compares the data of gathering with the threshold value, and then the time of prediction trouble emergence. The method is seriously dependent on the threshold range set by the prior experience of operation and maintenance personnel, needs to be independently set for power screens of different manufacturers and power screens in different states, and has low efficiency, low universality and insufficient accuracy of power screen fault prediction.
Disclosure of Invention
The fault prediction method of the track power supply screen is used for overcoming the defects of low efficiency, low universality and insufficient precision of power supply screen fault prediction in the prior art, can realize automatic prediction of power supply screen faults, improves the prediction efficiency, and meanwhile, constructs a precise fault prediction model based on the circuit parameters of the power supply screen, and improves the applicability and precision of power supply screen fault prediction.
The invention provides a fault prediction method of a rail power supply screen, which comprises the following steps:
performing anomaly detection on the circuit parameter time sequence sample based on an anomaly detection algorithm to obtain an anomaly sample, and performing fault type marking on the anomaly sample;
acquiring a target circuit parameter sample according to the abnormal sample and the data after the abnormal sample is subjected to fault type marking;
acquiring a fault prediction model according to the target circuit parameter sample;
inputting circuit parameters and a prediction period of a target power supply screen in a test set into a fault prediction model to obtain the probability of the target power supply screen failing;
wherein the circuit parameter timing sequence sample is determined according to circuit parameters of a target power supply screen in a training set;
and the acquisition of the circuit parameters of the target power supply screen in the training set is prior to the acquisition of the circuit parameters of the target power supply screen in the testing set.
According to the fault prediction method of the rail power supply screen, provided by the invention, the circuit parameter time sequence sample is obtained in the following way:
and according to the type of the circuit parameters of the target power supply screen in the training set, carrying out time sequence division on the circuit parameters of the target power supply screen in the training set at intervals of preset time so as to obtain the circuit parameter time sequence sample.
According to the fault prediction method of the track power supply screen provided by the invention, the abnormal detection is carried out on the circuit parameter time sequence sample based on the abnormal detection algorithm so as to obtain the abnormal sample, and the method comprises the following steps:
obtaining the statistical characteristics of the circuit parameter time sequence sample, and determining a characteristic matrix of the circuit parameter time sequence sample according to the statistical characteristics;
anomaly detection is carried out on the feature matrix based on an isolated forest anomaly detection algorithm or a local outlier coefficient anomaly detection algorithm so as to obtain an anomaly score of the circuit parameter time sequence sample;
determining the abnormal sample according to the abnormal score;
wherein the statistical features include at least: mean, standard deviation, and entropy.
According to the fault prediction method of the rail power supply screen provided by the invention, the step of obtaining the fault prediction model according to the target circuit parameter sample comprises the following steps:
analyzing the target circuit parameter samples to determine the transaction time of each fault type;
acquiring a maximum value of the transaction time, and taking the maximum value as the prediction period;
scaling the target circuit parameter sample in an equal ratio according to the prediction period to obtain a prediction vector;
determining parameters of the fault prediction model according to the prediction vector;
determining the fault prediction model according to the parameters of the fault model;
and the prediction vector is used for representing the probability of the target power supply screen failing.
According to the fault prediction method for the track power supply screen, the step of acquiring the fault probability of the target power supply screen comprises the following steps:
acquiring a slope vector in a prediction period from the starting time to the current time according to the circuit parameters of the target power supply screen in the test set and the prediction period;
matching the slope vector with the prediction vector output by the fault prediction model, and determining the probability of the fault of the target power supply screen according to the matching result;
wherein the slope vector is determined from adjacent data points in the circuit parameter;
the prediction vector is an average value of slope vectors of target circuit parameter samples after equal ratio scaling.
According to the fault prediction method of the rail power supply screen provided by the invention, the fault prediction model is obtained according to the target circuit parameter sample, and the fault prediction method further comprises the following steps:
dividing the target circuit parameter samples at intervals of the preset time according to the target circuit parameter samples to obtain target circuit parameter time sequence samples;
performing frequency domain decomposition on the target circuit parameter time sequence sample to obtain a season time sequence item, a trend time sequence item and an error item corresponding to the target circuit parameter time sequence sample;
determining parameters of the fault prediction model according to the seasonal time sequence item, the trend time sequence item and the error item;
and determining the fault prediction model according to the parameters of the fault prediction model.
According to the fault prediction method for the rail power supply screen, provided by the invention, the circuit parameters comprise:
voltage, current, frequency, phase and harmonics of the rail power supply;
voltage, current, frequency, phase and harmonics of the local power supply;
the phase angle of the local power supply and the phase angle of the rail power supply;
the suction release time and the return coefficient of the relay in the track power supply screen are determined;
track relay occupancy and idle state data;
voltage, current, phase, harmonics and frequency of the supply;
power on-off data between the transformer box and the signal end of the choke transformer;
voltage, current, phase, harmonic and frequency provided by a power transformer to the target power screen; and the temperature and humidity in the transformer box.
The invention also provides a fault prediction device of the rail power supply screen, which comprises the following components: the system comprises an abnormal sample labeling module, a target sample obtaining module, a fault prediction model determining module and a fault prediction module;
the abnormal sample labeling module is used for carrying out abnormal detection on the circuit parameter time sequence sample based on an abnormal detection algorithm so as to obtain an abnormal sample, and carrying out fault type labeling on the abnormal sample;
the target sample acquisition module is used for acquiring a target circuit parameter sample according to the abnormal sample and the data after the fault type marking is carried out on the abnormal sample;
the fault prediction model determining module is used for acquiring a fault prediction model according to the target circuit parameter sample;
the fault prediction module is used for inputting the circuit parameters and the prediction period of the target power supply screen in the test set into the fault prediction model to obtain the probability of the target power supply screen failing;
wherein the circuit parameter timing sequence sample is determined according to circuit parameters of a target power supply screen in a training set;
and the acquisition of the circuit parameters of the target power supply screen in the training set is prior to the acquisition of the circuit parameters of the target power supply screen in the testing set.
The invention also provides an electronic device, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor executes the program to realize the steps of the fault prediction method of the rail power supply screen.
The present invention also provides a non-transitory computer readable storage medium having stored thereon a computer program which, when executed by a processor, carries out the steps of the method of fault prediction of a rail power supply panel as described in any one of the above.
The fault prediction method of the track power supply screen provided by the invention obtains an abnormal sample by carrying out abnormal detection on the circuit parameters of the track power supply screen, and an accurate power supply screen fault prediction model is constructed by marking the fault type of the abnormal sample and combining a time sequence characteristic method, so that the automatic prediction of the power supply screen fault can be realized based on the constructed accurate fault prediction model, meanwhile, the accuracy of power supply panel fault prediction is improved, compared with the prior art that operation and maintenance personnel predict the power supply panel fault according to prior experience, the efficiency of power supply panel fault prediction is improved, and operation and maintenance personnel need to carry out fault prediction on different power supply panels independently when carrying out fault prediction on different power supply panels, the universality is lower, the fault prediction of different power supply screens can be realized by constructing a fault prediction model, and the applicability of the fault prediction of the power supply screens is improved.
Drawings
In order to more clearly illustrate the technical solutions of the present invention or the prior art, the drawings needed for the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.
FIG. 1 is a schematic flow chart of a fault prediction method for a rail power supply panel provided by the invention;
FIG. 2 is a schematic diagram of a process for circuit parameter data acquisition of a rail power supply panel provided by the present invention;
FIG. 3 is a schematic structural diagram of a fault prediction device of a rail power supply panel provided by the invention;
fig. 4 is a schematic structural diagram of an electronic device provided in the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings, and it is obvious that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Fig. 1 is a schematic flow chart of a fault prediction method for a rail power supply panel provided by the present invention, as shown in fig. 1, the method includes:
s1, performing anomaly detection on the circuit parameter time sequence samples based on an anomaly detection algorithm to obtain abnormal samples, and performing fault type marking on the abnormal samples;
s2, acquiring a target circuit parameter sample according to the abnormal sample and the data after the fault type marking is carried out on the abnormal sample;
s3, acquiring a fault prediction model according to the target circuit parameter sample;
s4, inputting the circuit parameters and the prediction period of the target power supply screen in the test set into a fault prediction model, and acquiring the probability of the target power supply screen failing;
the circuit parameter time sequence sample is determined according to circuit parameters of a target power supply screen in a training set;
the acquisition of the circuit parameters of the target power supply panel in the training set is prior to the acquisition of the circuit parameters of the target power supply panel in the testing set.
It should be noted that the execution subject of the method may be a computer device.
Optionally, the data acquisition module acquires circuit parameter related data of the track power supply screen, and divides the acquired circuit parameter related data into a training set and a test set according to the acquisition time sequence.
Optionally, the first collected circuit parameter related data is used as a training set, and the second collected circuit parameter related data is used as a test set. The circuit parameters of the power supply screen in the training set are used for obtaining a target circuit parameter sample required by building a fault prediction model, the circuit parameters of the power supply screen in the testing set are used as input data and input into the fault prediction model, and the probability of the power supply screen failing is obtained according to the trained fault prediction model.
Optionally, the circuit parameters of the power panel in the collected training set are subjected to time sequence division to obtain circuit parameter time sequence samples of the power panel, the circuit parameter time sequence samples are subjected to anomaly detection based on an anomaly detection algorithm to obtain anomaly data in the circuit parameter time sequence samples, and the anomaly samples are formed according to the anomaly data.
And marking the fault type of the power supply screen by operation and maintenance personnel according to the circuit parameter time sequence sample corresponding to the fault type in the expert database and in combination with the screened abnormal sample. The expert database may be pre-established according to historical power panel fault information and fault data corresponding to the historical power panel fault information, for example, the expert database may be further established by refining the fault data type corresponding to the power panel fault information.
Optionally, the fault information refers to description of a fault type of the power panel, for example, the fault type is a "start circuit fault" of the power panel, fault information corresponding to the fault type is an "indoor device fault" and an "outdoor device fault" are visible, the fault information actually includes a fault cause, only the cause is not specified to a more specific device, and the fault information is further refined through an expert database such as a pre-established relational database, that is, the cause of the power panel fault is determined step by step until certain power panel fault data corresponding to fault information that cannot be refined is traced back.
Optionally, when the fault information is "indoor device fault", the fault information may be further refined as: "track power failure", "local power failure" and "track circuit relay failure", when the fault information is "outdoor equipment trouble", can further refine to: "failure of power transmission side equipment", "failure of transformer tank and choke transformer equipment", and "failure of power supply transformer".
It should be noted that "rail power failure", "local power failure", "rail circuit relay failure", "power transmission end equipment failure", "transformer box and choke transformer equipment failure", and "power supply voltage equipment failure" in the present invention may still be further refined according to operation and maintenance experience until a corresponding failure equipment source is found, which is not specifically limited by the present invention.
And finally establishing an expert database for describing the fault type of the power supply screen according to the relation of the fault type, the fault information and the fault data.
Optionally, the abnormal sample and the fault type marking data corresponding to the abnormal sample are integrated to form a data set, and the obtained data set is used as a target circuit parameter sample to be used for subsequently constructing a fault prediction model based on the target circuit parameter sample. Optionally, the target circuit parameter samples are input into a pre-established fault prediction model for training, parameters of the pre-established fault prediction model are determined, and then the fault prediction model for power supply screen fault prediction is constructed according to the parameters of the pre-established fault prediction model.
The pre-established fault prediction model may be a corresponding decision tree machine learning model or a corresponding support vector machine model, which is not specifically limited in the present invention.
And inputting the collected circuit parameters and the prediction period of the power supply screen in the test set into the trained fault prediction model according to the constructed fault prediction model, and determining the fault probability of the power supply screen according to the output result of the fault prediction model, such as a prediction vector.
The fault prediction method of the track power supply screen provided by the invention obtains an abnormal sample by carrying out abnormal detection on the circuit parameters of the track power supply screen, and an accurate power supply screen fault prediction model is constructed by marking the fault type of the abnormal sample and combining a time sequence characteristic method, so that the automatic prediction of the power supply screen fault can be realized based on the constructed accurate fault prediction model, meanwhile, the accuracy of power supply panel fault prediction is improved, compared with the prior art that operation and maintenance personnel predict the power supply panel fault according to prior experience, the efficiency of power supply panel fault prediction is improved, and operation and maintenance personnel need to carry out fault prediction on different power supply panels independently when carrying out fault prediction on different power supply panels, the universality is lower, the fault prediction of different power supply screens can be realized by constructing a fault prediction model, and the applicability of the fault prediction of the power supply screens is improved.
Further, in one embodiment, the circuit parameter timing samples are obtained by:
a1, performing time sequence division on the circuit parameters of the target power supply screen in the training set at preset time intervals according to the type of the circuit parameters of the target power supply screen in the training set to obtain circuit parameter time sequence samples;
optionally, the circuit parameter related data of the power screen is acquired according to a data acquisition module, wherein the data acquisition module can be an industrial personal computer, and the circuit parameters of the power screen in continuous time can be acquired based on the industrial personal computer.
In the actual acquisition process, various different types of circuit parameters of the power supply screen are acquired as much as possible, and the acquired related data of various continuous circuit parameters are shunted according to the types of the circuit parameters of the power supply screen.
Optionally, the same type of circuit parameters of the power supply panel are taken as the same circuit sequence, such as voltage, current, phase and other circuit parameters; the different types of circuit parameters of the power supply screen are divided into different path sequences, for example, the related circuit parameters of the temperature and the humidity of the outdoor transformer box corresponding to the power supply screen are used as one path of sequence, and the power supply on-off between the corresponding transformer box of the power supply screen and the signal end of the choke transformer is used as the other path of sequence.
Optionally, the circuit parameters of the power panel corresponding to each path of sequence are divided into time sequences at preset time intervals, for example, the sequences corresponding to various continuous circuit parameter related data are divided into time sequence samples at T second intervals, so as to obtain time sequence samples corresponding to various circuit parameter related data, and the obtained time sequence samples corresponding to different types of circuit parameter related data are integrated to obtain circuit parameter time sequence samples of the power panel.
According to the fault prediction method of the track power supply screen, the circuit parameters of the power supply screen are refined by carrying out time sequence division on the collected different types of circuit parameters of the power supply screen, and the circuit parameter time sequence sample required by constructing the fault prediction model is obtained based on the refined circuit parameters.
Further, in an embodiment, the step S1 may specifically include:
s11, obtaining the statistical characteristics of the circuit parameter time sequence sample, and determining the characteristic matrix of the circuit parameter time sequence sample according to the statistical characteristics;
s12, carrying out anomaly detection on the feature matrix based on an isolated forest anomaly detection algorithm or a local outlier coefficient anomaly detection algorithm to obtain an anomaly score of a circuit parameter time sequence sample;
s13, determining an abnormal sample according to the abnormal score;
wherein the statistical features include at least: mean, standard deviation, and entropy.
Optionally, the circuit parameter timing samples obtained from the above are selected as
Circuit parameter timing samples t in (1)iIn other words, an unsupervised time sequence anomaly detection algorithm based on statistical characteristics is used for anomaly detection, the statistical characteristics, such as mean, standard deviation and entropy characteristics, of the circuit parameter time sequence samples corresponding to each path i are extracted, and a characteristic matrix F of the circuit parameter time sequence samples is formed.
And (3) calculating the abnormal score of each circuit parameter time sequence sample by adopting a statistical learning abnormal detection algorithm of isolated forest and/or local outlier coefficients, determining the abnormal sample in the circuit parameter time sequence sample based on the angle of statistical significance according to the abnormal score of the circuit parameter time sequence sample, repeating the process, and screening abnormal sample data from mass circuit parameter data to finish the subsequent fault type marking of the power screen according to the abnormal sample.
The process of calculating the abnormal score by the isolated forest abnormal detection algorithm specifically comprises the following steps:
b1, eighty percent of circuit parameter time sequence samples are used as training data, psi points are randomly selected from the training data to be used as sub samples, and the sub samples are placed into a root node of an isolated tree;
b2, randomly appointing a dimension, and randomly generating a cutting point p in the data range of the current node, wherein the cutting point is generated between the maximum value and the minimum value of the appointed dimension in the data of the current node;
b3, selecting the cutting point to generate a hyperplane, and dividing the data space of the current node into 2 subspaces: placing points smaller than p in the currently selected dimension on the left branch of the current node, and placing points larger than or equal to p on the right branch of the current node;
b4, recursion steps B2 and B3 at the left and right branch nodes of the node, and new leaf nodes are continuously constructed until only one data (cutting can not be continued) or the tree grows to the set height on the leaf node;
b5, repeating the steps B1-B4 until a preset number of isolated trees are obtained, integrating all the isolated trees, evaluating the remaining twenty percent of circuit parameter time sequence samples by using the generated isolated trees, and calculating the abnormal score of the obtained circuit parameter time sequence samples according to the formula (1):
where x represents the circuit parameter time series samples, h (x) represents the height of x per tree, and c (ψ) represents the average of the path lengths for a given number of samples ψ, for normalizing the path height of the sample x.
If the anomaly score is close to 1, then it must be an anomaly point;
if the anomaly score is much less than 0.5, then it must not be an anomaly point;
if all points are scored around 0.5 for outliers, then there is likely no outlier in the sample.
Similarly, an anomaly score for obtaining a time series sample of a circuit parameter may be calculated by a local outlier coefficient anomaly detection algorithm, such as the LOF algorithm.
According to the fault prediction method for the track power supply screen, the statistical characteristics such as the mean value, the standard deviation or the entropy of the circuit parameter time sequence samples of the power supply screen are obtained, the characterization of the characteristics of the circuit parameter time sequence samples is more accurate, the circuit parameter time sequence samples are subjected to abnormity detection based on an isolated forest abnormity detection algorithm or a local outlier coefficient abnormity detection algorithm, and compared with the traditional method of utilizing Kmeans and DBScan algorithms, due to the fact that distance related indexes do not need to be calculated, the abnormity detection of the circuit parameter time sequence samples can be rapidly completed, the abnormity detection speed is greatly increased, and further, the speed of obtaining the abnormity samples is increased.
Further, in an embodiment, the step S3 may specifically include:
s31, analyzing the target circuit parameter samples to determine the transaction time of each fault type;
s32, acquiring a maximum value of the transaction time, and taking the maximum value as a prediction period;
s33, scaling the target circuit parameter sample in an equal ratio according to the prediction period to obtain a prediction vector;
s34, determining parameters of the fault prediction model according to the prediction vector;
s35, determining a fault prediction model according to the parameters of the fault model;
the prediction vector is used for representing the probability of the target power supply screen failing.
Optionally, starting from the service logic, the fault prediction of the power supply panel is mainly divided into 2 parts, namely short-time prediction and long-time prediction, wherein the short-time fault mainly predicts the occurrence probability of a certain type of fault in a unit time in the future, and the long-time fault prediction mainly predicts the occurrence probability of a certain type of fault in a week or even longer in the future.
Optionally, the short-term prediction is implemented as follows:
analyzing the obtained target circuit parameter sample to obtain the time from the normal state to the fault state of each different type of circuit parameter depended by each fault type of the power supply screen, defining the time from the normal state to the fault state of the circuit parameter as the abnormal time, and recording the abnormal time as T1～TnumAnd num represents the total number of fault samples of each type in the target circuit parameter samples.
The prediction period T is used as the prediction of the short-time prediction by the maximum value of the abnormal action time zonesAnd T iss＝max[T1,T2,...,Tnum]According to the prediction period T, num fault sample datasScaling the data into the same dimension in an equal ratio, and calculating the slope k between adjacent data points to obtain the slope vector of num fault dataIs denoted by k1～num＝(k1,k2,...,knum) And determining a prediction vector of a fault prediction model for short-time prediction according to the obtained slope vector.
And taking the prediction vector of the obtained short-time prediction fault prediction model as a parameter of the fault prediction model, and constructing the fault prediction model according to the parameter of the fault prediction model for performing short-time fault prediction on the power supply screen.
According to the fault prediction method of the track power supply screen, parameters of a fault prediction model for short-time prediction are obtained by analyzing the target circuit parameter samples, the fault prediction model is determined according to the obtained parameters of the fault model, and accurate prediction of short-time faults of the power supply screen can be achieved.
Further, in an embodiment, the step S4 may specifically include:
s41, obtaining a slope vector in a prediction period from the starting time to the current time according to the circuit parameters and the prediction period of the target power supply screen in the test set;
s42, matching the slope vector with a prediction vector output by the fault prediction model, and determining the probability of the fault of the target power supply screen according to the matching result;
wherein the slope vector is determined from adjacent data points in the circuit parameter;
the prediction vector is the average of the slope vectors of the scaled target circuit parameter samples. Optionally, the circuit parameters and the predicted period T of the target power supply panel in the test setsAnd inputting the data into a fault prediction model to obtain a prediction vector. Wherein the prediction vector is obtained by calculating the prediction period TsAn average of the slope vectors of the scaled samples of the target circuit parameter.
Optionally, based on target power in test setCircuit parameters and predicted period T of screensCalculating a prediction period T from the start time to the current timesThe slope vector of the inner part is calculated to obtain a prediction period T from the starting time to the current timesThe slope vector in the prediction period is matched with the obtained prediction vector, and if the obtained prediction period T is calculated from the starting time to the current timesIf the difference value between the inner slope vector and the obtained prediction vector is within the preset threshold range, determining the next time point and having higher probability of the power supply screen failure, wherein the difference value between the next time point and the current time point is a prediction period Ts。
Alternatively, different threshold ranges may be set according to different prediction accuracies, for example, if a prediction period T from the start time to the current time is to be setsIf the difference value between the inner slope vector and the obtained prediction vector is within 0-0.05, the probability of the power supply screen failure at the next time point is determined to be more than 95%; if a prediction period T from the start time to the current time is to be usedsIf the difference value between the slope vector and the obtained prediction vector is within 0-0.5, the probability of the failure of the power supply screen at the next time point is determined to be more than 50%, and a prediction period T is from the starting time to the current timesThe relationship between the difference between the inner slope vector and the obtained prediction vector and the occurrence probability of the power supply screen fault can be set according to the prior experience of professional operation and maintenance personnel, and the method is not particularly limited in this respect.
According to the fault prediction method of the track power supply screen, the prediction vector output by the fault prediction model is matched with the slope vector in a prediction period from the starting time to the current time of the circuit parameter of the test centralized power supply screen obtained through calculation, so that the probability of the power supply screen failing in the prediction period from the current time to the next time is determined, meanwhile, the probability of the power supply screen failing is quantized according to the result of matching of the prediction vector and the slope vector, and the precision of short-time fault prediction of the power supply screen is improved.
Further, in one embodiment, step S3 may further include:
s36, dividing the target circuit parameter samples at preset time intervals according to the target circuit parameter samples to obtain target circuit parameter time sequence samples;
s37, performing frequency domain decomposition on the target circuit parameter time sequence sample to obtain a season time sequence item, a trend time sequence item and an error item corresponding to the target circuit parameter time sequence sample;
s38, determining parameters of a fault prediction model according to the seasonal time sequence item, the trend time sequence item and the error item;
and S39, determining the fault prediction model according to the parameters of the fault prediction model.
Optionally, when the long-term fault prediction is performed on the power supply screen, the long-term prediction of each type of fault of the power supply screen can be converted into prediction of each type of basic data trend of the power supply screen, and the fault occurrence time is reversely deduced according to the predicted trend so as to achieve the purpose of long-term fault prediction.
Therefore, when long-term fault prediction is performed on the power supply screen, firstly, the obtained target circuit parameter samples are divided at preset time intervals to obtain target circuit parameter time sequence samples yt；
Secondly, since a time series of basic data can be formed by overlapping or coupling several time series in the frequency domain, based on the decomposition model of the time series evolved, the target parameter time sequence sample y is sequenced according to the formula (2) or the formula (3)tPerforming frequency domain decomposition, and acquiring a season time sequence item, a trend time sequence item and an error item corresponding to the target circuit parameter time sequence sample:
yt＝St+Tt+Rt (2)
yt＝St*Tt*Rt (3)
in the formula, StRepresenting seasonal time series terms, TtRepresents a trend time series term, RtRepresents an error term;
wherein the seasonal time series term StThe trend time sequence term T changes along with seasonal changes of days, weeks, months, years and the liketRepresenting the trend of the time series over a non-periodic time period, RtRepresentation refers to other factors that affect the time series data, such as the accuracy of the equipment used to collect the data.
Furthermore, the target circuit parameter timing samples may also be decomposed into an addition/multiplication hybrid model consisting of equation (2) and equation (3). During feature design, the workload of a single-week (month and quarter) time can be calculated by defining the workload as a period variable according to different degrees, namely, different long-term fault prediction periods can be set.
According to the obtained seasonal time sequence item StTrend time series term TtAnd an error term RtAnd parameters of the fault prediction model are formed, and the fault prediction model for long-term fault prediction is determined according to the parameters of the fault model.
Optionally, when long-term fault prediction is performed on the power supply screen, the collected circuit parameters of the target power supply screen in the test set and the prediction period of the long-term fault prediction are input into the fault prediction model, and the seasonal time sequence item S is obtainedtTrend time series term TtAnd an error term Rt。
The professional operation and maintenance personnel acquire the seasonal time sequence item StThe seasonal time sequence item when the power supply screen normally operates and the acquired trend time sequence item TtAnd a trend time sequence item when the power supply screen normally operates and an error item R obtained according to the trend time sequence itemtAnd comparing the error term with the error term when the power supply screen is normal to determine whether the power supply screen fails in the prediction period of the long-term failure prediction.
According to the method for predicting the faults of the track power supply screen, the seasonal time sequence item, the trend time sequence item and the error item which are required by long-time fault prediction of the power supply screen are obtained by performing frequency domain decomposition on the target circuit parameter time sequence sample, and the seasonal time sequence item, the trend time sequence item and the error item which are obtained are compared with the seasonal time sequence item, the trend time sequence item and the error item when the power supply screen normally operates, so that the long-time fault prediction of the power supply screen is realized.
Further, in one embodiment, the circuit parameters of the power supply panel include the following data:
voltage, current, frequency, phase and harmonics of the rail power supply;
voltage, current, frequency, phase and harmonics of the local power supply;
the phase angle of the local power supply and the phase angle of the rail power supply;
the suction release time and the return coefficient of the relay in the track power supply screen are determined;
track relay occupancy and idle state data;
voltage, current, phase, harmonics and frequency of the supply;
power on-off data between the transformer box and the signal end of the choke transformer;
voltage, current, phase, harmonic and frequency provided by a power transformer to the target power screen; and the temperature and humidity in the transformer box.
Optionally, as shown in fig. 2, when the power panel is subjected to fault prediction, because the circuit parameters depended on by the power panel are different when the corresponding fault types of the power panel are different, in an actual scene, the data acquisition module is often used to acquire the circuit parameters of the power panel from the outdoor device and the indoor device.
Optionally, the indoor device collects the following circuit parameters of the main collection power supply screen:
the rail power supply, the data of gathering include: voltage, current, frequency, phase and harmonics;
local power supply, the data gathered include: voltage, current, frequency, phase and harmonics;
the phase angle of the local power supply and the phase angle of the rail power supply;
the performance of a key relay of the track circuit, the suction release time, the return coefficient and the like;
track relay occupancy and idle state data;
optionally, the outdoor device collects the following circuit parameters of the main collection power supply screen:
the power quality of the power transmission end comprises voltage, current, phase, harmonic and frequency;
power supply on-off data between the XB box of the transformer and the BE signal end of the choke transformer;
the power quality provided by the power transformer BG to the equipment comprises voltage, current, phase, harmonic and frequency;
and the temperature and the humidity in the XB box of the transformer.
According to the fault prediction method of the track power supply screen, the circuit parameters of the power supply screen are acquired through the data acquisition module, abundant circuit parameter samples of the power supply screen can be provided for constructing a fault prediction model, and the established fault prediction model can reflect the running state of the power supply screen more comprehensively and accurately.
The fault prediction device of the track power supply panel provided by the invention is described below, and the fault prediction device of the track power supply panel described below and the fault prediction method of the track power supply panel described above can be referred to correspondingly.
Fig. 3 is a schematic structural diagram of a fault prediction apparatus of a rail power supply panel provided in the present invention, as shown in fig. 3, including: an abnormal sample labeling module 310, a target sample obtaining module 311, a fault prediction model determining module 312 and a fault prediction module 313;
an abnormal sample labeling module 310, configured to perform abnormal detection on the circuit parameter timing sequence sample based on an abnormal detection algorithm to obtain an abnormal sample, and perform fault type labeling on the abnormal sample;
a target sample obtaining module 311, configured to obtain a target circuit parameter sample according to the abnormal sample and the fault type marking data;
a fault prediction model determining module 312, configured to obtain a fault prediction model according to the target circuit parameter sample;
the fault prediction module 313 is used for inputting the circuit parameters and the prediction period of the target power supply screen in the test set into the fault prediction model to obtain the probability of the target power supply screen failing;
the circuit parameter time sequence sample is determined according to circuit parameters of a target power supply screen in a training set;
the circuit parameters of the target power supply screen in the training set are collected before the circuit parameters of the target power supply screen in the testing set are collected.
The fault prediction device of the track power supply screen provided by the invention has the advantages that the circuit parameters of the track power supply screen are subjected to abnormal detection based on the abnormal sample marking module 310, the abnormal sample is obtained, the fault type marking is carried out on the abnormal sample, the time sequence characteristic method is combined, the target sample obtaining module 311 is utilized to integrate the abnormal sample and the fault type marking data to obtain the target circuit parameter sample, the fault prediction model determining module 312 is utilized to construct an accurate power supply screen fault prediction model, the fault prediction module 313 can construct the accurate fault prediction model based on the accurate fault prediction model to realize the automatic prediction of the power supply screen fault, the accuracy of the power supply screen fault prediction is improved, compared with the prior art that operation and maintenance personnel predict the power supply screen fault according to the prior experience, the efficiency of the power supply screen fault prediction is improved, and the operation and maintenance personnel need to carry out fault prediction on different power supply screens independently when carrying out fault prediction on different power supply screens The method has low universality, can realize the fault prediction of different power supply screens by constructing a fault prediction model, and improves the applicability of the fault prediction of the power supply screens.
Fig. 4 is a schematic physical structure diagram of an electronic device provided in the present invention, and as shown in fig. 4, the electronic device may include: a processor (processor)410, a communication interface 411, a memory (memory)412 and a bus (bus)413, wherein the processor 410, the communication interface 411 and the memory 412 complete communication with each other through the bus 413. The processor 410 may call logic instructions in the memory 412 to perform the following method:
performing anomaly detection on the circuit parameter time sequence samples based on an anomaly detection algorithm to obtain abnormal samples, and performing fault type marking on the abnormal samples;
acquiring a target circuit parameter sample according to the abnormal sample and the data after the abnormal sample is subjected to fault type marking;
acquiring a fault prediction model according to the target circuit parameter sample;
inputting the circuit parameters and the prediction period of the target power supply screen in the test set into a fault prediction model to obtain the probability of the target power supply screen failing;
the circuit parameter time sequence sample is determined according to circuit parameters of a target power supply screen in a training set;
the acquisition of the circuit parameters of the target power supply panel in the training set is prior to the acquisition of the circuit parameters of the target power supply panel in the testing set.
In addition, the logic instructions in the memory may be implemented in the form of software functional units and may be stored in a computer readable storage medium when sold or used as a stand-alone product. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and the like.
Further, the present invention discloses a computer program product comprising a computer program stored on a non-transitory computer readable storage medium, the computer program comprising program instructions, which when executed by a computer, the computer is capable of executing the method for predicting failure of a rail power supply panel provided by the above-mentioned method embodiments, for example, the method comprises:
performing anomaly detection on the circuit parameter time sequence samples based on an anomaly detection algorithm to obtain abnormal samples, and performing fault type marking on the abnormal samples;
acquiring a target circuit parameter sample according to the abnormal sample and the data after the abnormal sample is subjected to fault type marking;
acquiring a fault prediction model according to the target circuit parameter sample;
inputting the circuit parameters and the prediction period of the target power supply screen in the test set into a fault prediction model to obtain the probability of the target power supply screen failing;
the circuit parameter time sequence sample is determined according to circuit parameters of a target power supply screen in a training set;
the acquisition of the circuit parameters of the target power supply panel in the training set is prior to the acquisition of the circuit parameters of the target power supply panel in the testing set.
In another aspect, the present invention also provides a non-transitory computer readable storage medium, on which a computer program is stored, the computer program being implemented by a processor to execute the method for predicting the failure of a rail power supply panel provided in the foregoing embodiments, for example, the method includes:
performing anomaly detection on the circuit parameter time sequence samples based on an anomaly detection algorithm to obtain abnormal samples, and performing fault type marking on the abnormal samples;
acquiring a target circuit parameter sample according to the abnormal sample and the data after the abnormal sample is subjected to fault type marking;
acquiring a fault prediction model according to the target circuit parameter sample;
inputting the circuit parameters and the prediction period of the target power supply screen in the test set into a fault prediction model to obtain the probability of the target power supply screen failing;
the circuit parameter time sequence sample is determined according to circuit parameters of a target power supply screen in a training set;
the acquisition of the circuit parameters of the target power supply panel in the training set is prior to the acquisition of the circuit parameters of the target power supply panel in the testing set.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A fault prediction method of a rail power supply screen is characterized by comprising the following steps:
performing anomaly detection on the circuit parameter time sequence sample based on an anomaly detection algorithm to obtain an anomaly sample, and performing fault type marking on the anomaly sample;
acquiring a target circuit parameter sample according to the abnormal sample and the data after the abnormal sample is subjected to fault type marking;
acquiring a fault prediction model according to the target circuit parameter sample;
inputting circuit parameters and a prediction period of a target power supply screen in a test set into a fault prediction model to obtain the probability of the target power supply screen failing;
wherein the circuit parameter timing sequence sample is determined according to circuit parameters of a target power supply screen in a training set;
and the acquisition of the circuit parameters of the target power supply screen in the training set is prior to the acquisition of the circuit parameters of the target power supply screen in the testing set.
2. The method of claim 1, wherein the circuit parameter timing samples are obtained by:
and according to the type of the circuit parameters of the target power supply screen in the training set, carrying out time sequence division on the circuit parameters of the target power supply screen in the training set at intervals of preset time so as to obtain the circuit parameter time sequence sample.
3. The method for predicting the fault of the rail power supply panel according to claim 1, wherein the performing the abnormal detection on the circuit parameter time sequence samples based on the abnormal detection algorithm to obtain the abnormal samples comprises:
obtaining the statistical characteristics of the circuit parameter time sequence sample, and determining a characteristic matrix of the circuit parameter time sequence sample according to the statistical characteristics;
anomaly detection is carried out on the feature matrix based on an isolated forest anomaly detection algorithm or a local outlier coefficient anomaly detection algorithm so as to obtain an anomaly score of the circuit parameter time sequence sample;
determining the abnormal sample according to the abnormal score;
wherein the statistical features include at least: mean, standard deviation, and entropy.
4. The method of claim 1, wherein the obtaining the fault prediction model based on the target circuit parameter samples comprises:
analyzing the target circuit parameter samples to determine the transaction time of each fault type;
acquiring a maximum value of the transaction time, and taking the maximum value as the prediction period;
scaling the target circuit parameter sample in an equal ratio according to the prediction period to obtain a prediction vector;
determining parameters of the fault prediction model according to the prediction vector;
determining the fault prediction model according to the parameters of the fault model;
and the prediction vector is used for representing the probability of the target power supply screen failing.
5. The method for predicting the fault of the rail power supply screen according to claim 4, wherein the obtaining the probability of the fault of the target power supply screen comprises:
acquiring a slope vector in a prediction period from the starting time to the current time according to the circuit parameters of the target power supply screen in the test set and the prediction period;
matching the slope vector with the prediction vector output by the fault prediction model, and determining the probability of the fault of the target power supply screen according to the matching result;
wherein the slope vector is determined from adjacent data points in the circuit parameter;
the prediction vector is an average value of slope vectors of target circuit parameter samples after equal ratio scaling.
6. The method for predicting the fault of the rail power supply panel as recited in claim 1, wherein the obtaining a fault prediction model according to the target circuit parameter sample comprises:
dividing the target circuit parameter samples at intervals of the preset time according to the target circuit parameter samples to obtain target circuit parameter time sequence samples;
performing frequency domain decomposition on the target circuit parameter time sequence sample to obtain a season time sequence item, a trend time sequence item and an error item corresponding to the target circuit parameter time sequence sample;
determining parameters of the fault prediction model according to the seasonal time sequence item, the trend time sequence item and the error item;
and determining the fault prediction model according to the parameters of the fault prediction model.
7. The method of any of claims 1-6, wherein the circuit parameters comprise:
voltage, current, frequency, phase and harmonics of the rail power supply;
voltage, current, frequency, phase and harmonics of the local power supply;
the phase angle of the local power supply and the phase angle of the rail power supply;
the suction release time and the return coefficient of the relay in the track power supply screen are determined;
track relay occupancy and idle state data;
voltage, current, phase, harmonics and frequency of the supply;
power on-off data between the transformer box and the signal end of the choke transformer;
voltage, current, phase, harmonic and frequency provided by a power transformer to the target power screen; and the temperature and humidity in the transformer box.
8. A failure prediction device for a rail power supply panel, comprising: the system comprises an abnormal sample labeling module, a target sample obtaining module, a fault prediction model determining module and a fault prediction module;
the abnormal sample labeling module is used for carrying out abnormal detection on the circuit parameter time sequence sample based on an abnormal detection algorithm so as to obtain an abnormal sample, and carrying out fault type labeling on the abnormal sample;
the target sample acquisition module is used for acquiring a target circuit parameter sample according to the abnormal sample and the data after the fault type marking is carried out on the abnormal sample;
the fault prediction model determining module is used for acquiring a fault prediction model according to the target circuit parameter sample;
the fault prediction module is used for inputting the circuit parameters and the prediction period of the target power supply screen in the test set into the fault prediction model to obtain the probability of the target power supply screen failing;
wherein the circuit parameter timing sequence sample is determined according to circuit parameters of a target power supply screen in a training set;
and the acquisition of the circuit parameters of the target power supply screen in the training set is prior to the acquisition of the circuit parameters of the target power supply screen in the testing set.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the computer program implements the steps of the method for predicting a failure of a rail power supply panel according to any one of claims 1 to 7.
10. A non-transitory computer readable storage medium having stored thereon a computer program, wherein the computer program when executed by a processor implements the steps of the method for predicting a failure of a rail power supply panel according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011535641.9A CN112578213A (en) | 2020-12-23 | 2020-12-23 | Fault prediction method and device for rail power supply screen |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011535641.9A CN112578213A (en) | 2020-12-23 | 2020-12-23 | Fault prediction method and device for rail power supply screen |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112578213A true CN112578213A (en) | 2021-03-30 |
Family
ID=75138925
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011535641.9A Pending CN112578213A (en) | 2020-12-23 | 2020-12-23 | Fault prediction method and device for rail power supply screen |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112578213A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116184988A (en) * | 2023-05-04 | 2023-05-30 | 中科航迈数控软件(深圳)有限公司 | Multi-mode data-based fault prediction method, device, equipment and storage medium |
| CN116388112A (en) * | 2023-05-26 | 2023-07-04 | 北京国电通网络技术有限公司 | Abnormal supply end power-off method, device, electronic equipment and computer readable medium |
| CN117574300A (en) * | 2023-12-15 | 2024-02-20 | 国网湖北省电力有限公司超高压公司 | Oil immersed transformer aging fault identification method based on data analysis |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103745229A (en) * | 2013-12-31 | 2014-04-23 | 北京泰乐德信息技术有限公司 | Method and system of fault diagnosis of rail transit based on SVM (Support Vector Machine) |
| CN109034470A (en) * | 2018-07-18 | 2018-12-18 | 国网冀北电力有限公司信息通信分公司 | A kind of power communication scene O&M failure prediction method |
| CN111353482A (en) * | 2020-05-25 | 2020-06-30 | 天津开发区精诺瀚海数据科技有限公司 | LSTM-based fatigue factor recessive anomaly detection and fault diagnosis method |
| CN112001440A (en) * | 2020-08-20 | 2020-11-27 | 苏州鸿哲智能科技有限公司 | Fault diagnosis logic algorithm and system |
-
2020
- 2020-12-23 CN CN202011535641.9A patent/CN112578213A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103745229A (en) * | 2013-12-31 | 2014-04-23 | 北京泰乐德信息技术有限公司 | Method and system of fault diagnosis of rail transit based on SVM (Support Vector Machine) |
| CN109034470A (en) * | 2018-07-18 | 2018-12-18 | 国网冀北电力有限公司信息通信分公司 | A kind of power communication scene O&M failure prediction method |
| CN111353482A (en) * | 2020-05-25 | 2020-06-30 | 天津开发区精诺瀚海数据科技有限公司 | LSTM-based fatigue factor recessive anomaly detection and fault diagnosis method |
| CN112001440A (en) * | 2020-08-20 | 2020-11-27 | 苏州鸿哲智能科技有限公司 | Fault diagnosis logic algorithm and system |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116184988A (en) * | 2023-05-04 | 2023-05-30 | 中科航迈数控软件(深圳)有限公司 | Multi-mode data-based fault prediction method, device, equipment and storage medium |
| CN116388112A (en) * | 2023-05-26 | 2023-07-04 | 北京国电通网络技术有限公司 | Abnormal supply end power-off method, device, electronic equipment and computer readable medium |
| CN116388112B (en) * | 2023-05-26 | 2023-09-12 | 北京国电通网络技术有限公司 | Abnormal supply end power-off method, device, electronic equipment and computer readable medium |
| CN117574300A (en) * | 2023-12-15 | 2024-02-20 | 国网湖北省电力有限公司超高压公司 | Oil immersed transformer aging fault identification method based on data analysis |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112578213A (en) | Fault prediction method and device for rail power supply screen | |
| US20200327435A1 (en) | Systems and methods for sequential power system model parameter estimation | |
| JP5734961B2 (en) | Apparatus and method for estimating model quality and adapting model in multivariable process control | |
| US10444746B2 (en) | Method for managing subsystems of a process plant using a distributed control system | |
| CN111680820B (en) | Distributed photovoltaic power station fault diagnosis method and device | |
| CN111783953A (en) | 24-point power load value 7-day prediction method based on optimized LSTM network | |
| JP2012528392A5 (en) | ||
| CN112035716A (en) | XGboost algorithm-based fan generator fault monitoring and diagnosis method | |
| CN110544180A (en) | Building energy-saving system based on energy consumption prediction and analysis diagnosis | |
| CN111380686A (en) | XGboost algorithm model-based fan main bearing fault monitoring and diagnosis method | |
| CN112734128A (en) | 7-day power load peak value prediction method based on optimized RBF | |
| CN111582551A (en) | Method and system for predicting short-term wind speed of wind power plant and electronic equipment | |
| JP2002532799A (en) | Case-based reasoning system and method and apparatus for sensor prediction, especially in technological processes of cement kilns | |
| CN112858804A (en) | Power supply screen fault diagnosis method and device for railway vehicle | |
| CN117008479B (en) | Carbon emission optimization control method and system based on biomass gasification furnace | |
| CN115577851A (en) | Energy consumption prediction method, device, equipment and storage medium | |
| CN115603464B (en) | Intelligent generation management system for power grid operation ticket based on digital twin | |
| CN108363738B (en) | Recommendation method for industrial equipment data analysis algorithm | |
| CN112286088A (en) | Method and application system for online application of power equipment fault prediction model | |
| CN113361737A (en) | Abnormity early warning method and system for photovoltaic module | |
| KR101884907B1 (en) | Big Data Analytics Based Reliability Prediction Method | |
| CN113988173A (en) | Fault diagnosis method, system, equipment and storage medium based on qualitative trend analysis and five-state Bayesian network | |
| CN115186935A (en) | Electromechanical device nonlinear fault prediction method and system | |
| CN114677009A (en) | IGEP-CNN algorithm-based power data evaluation method | |
| Züfle et al. | A framework for time series preprocessing and history-based forecasting method recommendation |
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 |