CN115659185A - Method and device for processing time sequence data of operation and maintenance system - Google Patents

Method and device for processing time sequence data of operation and maintenance system Download PDF

Info

Publication number
CN115659185A
CN115659185A CN202211093088.7A CN202211093088A CN115659185A CN 115659185 A CN115659185 A CN 115659185A CN 202211093088 A CN202211093088 A CN 202211093088A CN 115659185 A CN115659185 A CN 115659185A
Authority
CN
China
Prior art keywords
time sequence
time
sequence
data
processing
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
Application number
CN202211093088.7A
Other languages
Chinese (zh)
Inventor
陈超
严川
张博
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Cloudwise Beijing Technology Co Ltd
Original Assignee
Cloudwise Beijing Technology Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Cloudwise Beijing Technology Co Ltd filed Critical Cloudwise Beijing Technology Co Ltd
Publication of CN115659185A publication Critical patent/CN115659185A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/245—Query processing
    • G06F16/2458—Special types of queries, e.g. statistical queries, fuzzy queries or distributed queries
    • G06F16/2474—Sequence data queries, e.g. querying versioned data
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/28—Databases characterised by their database models, e.g. relational or object models
    • G06F16/284—Relational databases
    • G06F16/285—Clustering or classification

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Computational Linguistics (AREA)
  • Software Systems (AREA)
  • Probability & Statistics with Applications (AREA)
  • Fuzzy Systems (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
  • Testing And Monitoring For Control Systems (AREA)

Abstract

The embodiment of the invention provides a method and a device for processing time sequence data of an operation and maintenance system, wherein the method comprises the following steps: acquiring time series data of a target object of an operation and maintenance system; carrying out periodic detection processing on the time sequence data to obtain a detection result of whether the time sequence data is a periodic time sequence; if the detection result shows that the time sequence data is a cycle time sequence, removing the cycle of the cycle time sequence from the time sequence data to obtain a target time sequence; and according to the target time sequence, carrying out detection, classification and/or prediction processing on the abnormal indexes of the target object to obtain a processing result, and outputting the processing result. The embodiment of the invention can realize the abnormal detection, classification or early warning treatment of the target object of the operation and maintenance system, and improve the operation performance of the operation and maintenance system.

Description

Method and device for processing time sequence data of operation and maintenance system
Technical Field
The present invention relates to the field of time series data processing technologies, and in particular, to a method and an apparatus for processing time series data of an operation and maintenance system.
Background
With the development of large data, time series data appears in more and more fields, such as internet of things, medical treatment, finance and the like. In an operation and maintenance scene, various kpi indexes are required to be subjected to automatic anomaly detection, classification or early warning treatment so as to monitor whether a system fails or not. In most cases, the upper and lower bounds of the index are set by means of a static threshold or k-sigma, and if the index exceeds the upper bound or is lower than the lower bound, the index is considered as abnormal. When the mode meets periodic indexes, the detection accuracy is poor.
As shown in fig. 1, the index (time series) includes two abnormalities a and B. When the detection is performed using k-sigma, only B-anomalies can be detected as shown in fig. 2.
The conventional data period detection method cannot well process time sequence data which simultaneously covers periods and trend components, so that inaccurate period detection is caused, the time sequence data cannot be effectively analyzed, and further, target objects in an operation and maintenance system cannot be effectively subjected to effective abnormal detection, classification or early warning treatment, and the performance of the operation and maintenance system is influenced.
Disclosure of Invention
The invention provides a method and a device for processing time sequence data of an operation and maintenance system. Based on the periodic detection of the time series data, the abnormity detection, classification or early warning processing of the target object of the operation and maintenance system is realized, and the operation performance of the operation and maintenance system is improved.
To solve the above technical problem, an embodiment of the present invention provides the following solutions:
a method for processing time series data of an operation and maintenance system comprises the following steps:
acquiring time sequence data of a target object of the operation and maintenance system;
carrying out periodic detection processing on the time sequence data to obtain a detection result of whether the time sequence data is a periodic time sequence;
if the detection result shows that the time sequence data is a cycle time sequence, removing the cycle of the cycle time sequence from the time sequence data to obtain a target time sequence;
and according to the target time sequence, carrying out detection, classification and/or prediction processing on the abnormal indexes of the target object to obtain a processing result, and outputting the processing result.
Optionally, the performing a periodic detection process on the time series data to obtain a detection result of whether the time series data is a periodic time series includes:
processing the time sequence data to obtain an autocorrelation sequence of the time sequence data;
processing the self-correlation sequence to obtain a similarity difference sequence of the self-correlation sequence;
obtaining a candidate period T of the time sequence data according to the similarity difference sequence;
obtaining the noise degree according to the candidate period T and the time sequence data
Figure 394680DEST_PATH_IMAGE001
And the specific gravity of the trend component
Figure 868386DEST_PATH_IMAGE002
；
According to the noise level
Figure 712846DEST_PATH_IMAGE001
And specific gravity of trend component
Figure 332132DEST_PATH_IMAGE002
And obtaining a detection result of whether the time series data are periodic time series or not according to a corresponding threshold value.
Optionally, processing the time-series data to obtain an autocorrelation sequence of the time-series data includes:
according to an autocorrelation function, carrying out autocorrelation calculation on the time sequence data to obtain an autocorrelation sequence of the time sequence data; the autocorrelation function is:
Figure 335860DEST_PATH_IMAGE003
；
wherein the time-series data is
Figure 121414DEST_PATH_IMAGE004
N represents the length of time-series data X;
Figure 577803DEST_PATH_IMAGE005
is the average of the time-series data X,
Figure 457903DEST_PATH_IMAGE006
；
Figure 191504DEST_PATH_IMAGE007
indicates the t-th point in the time-series data,
Figure 7013DEST_PATH_IMAGE008
。
optionally, the autocorrelation sequence of the time-series data is:
Figure 75332DEST_PATH_IMAGE009
；
wherein, note
Figure 900069DEST_PATH_IMAGE010
，
Figure 488176DEST_PATH_IMAGE011
，
Figure 599220DEST_PATH_IMAGE012
；
Figure 30201DEST_PATH_IMAGE013
And
Figure 268416DEST_PATH_IMAGE014
is composed of
Figure 101243DEST_PATH_IMAGE015
I is the length of the subsequence;
then
Figure 648768DEST_PATH_IMAGE016
。
Optionally, obtaining a similarity difference sequence of the auto-correlation sequence includes:
computing subsequence
Figure 442411DEST_PATH_IMAGE013
And
Figure 733584DEST_PATH_IMAGE014
obtaining similarity difference sequences M through similarity difference under each length;
wherein,
Figure 155338DEST_PATH_IMAGE017
，
Figure 358918DEST_PATH_IMAGE018
a function is calculated for the similarity difference.
Optionally, obtaining the candidate period T of the time series data according to the similarity difference sequence includes:
according to the similarity difference sequence, order
Figure 764491DEST_PATH_IMAGE019
By the formula
Figure 859355DEST_PATH_IMAGE020
Obtaining a candidate period T of the time sequence data; wherein j is the index of the element M in the similarity difference sequence M.
Optionally, according to the waiting timeSelecting the period T and the time sequence data to obtain the noise degree
Figure 276561DEST_PATH_IMAGE001
And specific gravity of trend component
Figure 431468DEST_PATH_IMAGE002
The method comprises the following steps:
according to
Figure 58758DEST_PATH_IMAGE021
Obtaining the noise level
Figure 442466DEST_PATH_IMAGE001
Or, according to
Figure 838812DEST_PATH_IMAGE022
Obtaining the noise level
Figure 899041DEST_PATH_IMAGE001
(ii) a Wherein,
Figure 888994DEST_PATH_IMAGE023
；
according to
Figure 122398DEST_PATH_IMAGE024
Obtaining the specific gravity of the trend component
Figure 779775DEST_PATH_IMAGE002
Or according to
Figure 745326DEST_PATH_IMAGE025
Obtaining the specific gravity of the trend component
Figure 347209DEST_PATH_IMAGE002
Or, according to
Figure 72719DEST_PATH_IMAGE026
Obtaining the specific gravity of the trend component
Figure 443658DEST_PATH_IMAGE002
。
Optionally, according to said noise level
Figure 580110DEST_PATH_IMAGE001
And specific gravity of trend component
Figure 403709DEST_PATH_IMAGE002
And a corresponding threshold, determining the time series data as a periodic time series, comprising:
if it is
Figure 932911DEST_PATH_IMAGE027
And, in addition,
Figure 282990DEST_PATH_IMAGE028
if the time sequence data X is a periodic time sequence, the period length is T;
wherein,
Figure 465709DEST_PATH_IMAGE029
a preset noise threshold value,
Figure 183130DEST_PATH_IMAGE030
Is a preset trend threshold.
The embodiment of the present invention further provides a device for processing time series data of an operation and maintenance system, where the device includes:
the acquisition module is used for acquiring time series data of a target object of the operation and maintenance system;
the processing module is used for carrying out periodic detection processing on the time sequence data to obtain a detection result of whether the time sequence data is a periodic time sequence; if the detection result shows that the time sequence data is a cycle time sequence, removing the cycle of the cycle time sequence from the time sequence data to obtain a target time sequence; and according to the target time sequence, carrying out detection, classification and/or prediction processing on the abnormal indexes of the target object to obtain a processing result, and outputting the processing result.
Embodiments of the present invention also provide a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the method as described above.
The scheme of the invention at least comprises the following beneficial effects:
according to the scheme, time sequence data of the target object of the operation and maintenance system are obtained; carrying out periodic detection processing on the time sequence data to obtain a detection result of whether the time sequence data is a periodic time sequence; if the detection result shows that the time sequence data is a cycle time sequence, removing the cycle of the cycle time sequence from the time sequence data to obtain a target time sequence; and according to the target time sequence, carrying out detection, classification and/or prediction processing on the abnormal indexes of the target object to obtain a processing result, and outputting the processing result. Based on the periodic detection of the time series data, the abnormity detection, classification or early warning processing of the target object of the operation and maintenance system is realized, and the operation performance of the operation and maintenance system is improved.
Drawings
FIG. 1 is a schematic diagram of time-series data in a method for processing time-series data;
FIG. 2 is a schematic diagram illustrating the effect of k-sigma anomaly detection performed on the time-series data shown in FIG. 1;
FIG. 3 is a schematic flow chart of a method for processing time-series data according to an embodiment of the present invention;
FIG. 4 is a diagram illustrating raw data in a time series data cycle test according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of a period component extracted from the time-series data of FIG. 4 according to an embodiment of the present invention (where a is a schematic diagram of the period component, and b is a schematic diagram of the comparison between the period component and the original data);
FIG. 6 is a schematic diagram of the target time sequence after the period is removed from the time sequence data of FIG. 4 according to an embodiment of the present invention (where a is a schematic diagram of the obtained residual sequence, and b is a schematic diagram of a comparison between the original data, the period component, and the residual sequence);
fig. 7 is a schematic diagram of a detection effect obtained by the time series data of fig. 4 according to the residual sequence in the embodiment of the present invention (where a is a schematic diagram of an abnormal point obtained by comparing the residual sequence with a threshold, b is a schematic diagram of an abnormal point obtained by comparing the original data, the residual sequence and the threshold, and c is a schematic diagram of an abnormal point obtained by comparing the original data with the threshold);
FIG. 8 is a schematic structural diagram of an apparatus for processing time-series data according to an embodiment of the present invention;
fig. 9 is a schematic diagram of an autocorrelation sequence obtained by processing time-series data according to an embodiment of the present invention.
Detailed Description
Exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the invention are shown in the drawings, it should be understood that the invention may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art.
As shown in fig. 3, the present invention provides a method for detecting time-series data, including:
step 31, acquiring time series data of a target object of the operation and maintenance system, such as index data of database call delay shown in fig. 4, and recording time lengths required for database query at different times;
step 32, performing periodic detection processing on the time series data to obtain a detection result of whether the time series data is a periodic time series;
step 33, if the detection result indicates that the time sequence data is a periodic time sequence, removing a period of the periodic time sequence from the time sequence data to obtain a target time sequence, and as shown in fig. 6, removing a period component from a database calling delay index to obtain a residual sequence;
and step 34, according to the target time sequence, carrying out detection, classification and/or prediction processing on the abnormal indexes of the target object to obtain a processing result, and outputting the processing result.
The embodiment acquires time series data of a target object of the operation and maintenance system; carrying out periodic detection processing on the time sequence data to obtain a detection result of whether the time sequence data is a periodic time sequence; if the detection result shows that the time sequence data is a cycle time sequence, removing the cycle of the cycle time sequence from the time sequence data to obtain a target time sequence; and according to the target time sequence, carrying out detection, classification and/or prediction processing on the abnormal indexes of the target object to obtain a processing result, and outputting the processing result. Based on the periodic detection of the time series data, the abnormity detection, classification or early warning processing of the target object of the operation and maintenance system is realized, and the operation performance of the operation and maintenance system is improved.
In an alternative embodiment of the present invention, step 32 may include:
step 321, processing the time series data to obtain an autocorrelation sequence of the time series data; specifically, according to an autocorrelation function, performing autocorrelation calculation on the time-series data to obtain an autocorrelation sequence of the time-series data, as shown in fig. 9;
here, the autocorrelation function is:
Figure 499710DEST_PATH_IMAGE003
；
wherein the time-series data is
Figure 579662DEST_PATH_IMAGE004
N represents the length of the time-series data X;
Figure 339807DEST_PATH_IMAGE005
is the average of the time-series data X,
Figure 137999DEST_PATH_IMAGE006
；
Figure 992692DEST_PATH_IMAGE007
indicating the t-th point in the sequence,
Figure 927149DEST_PATH_IMAGE008
，
Figure 858196DEST_PATH_IMAGE031
represents the mathematical expectation of X;
the autocorrelation sequence of the time series data is:
Figure 268318DEST_PATH_IMAGE009
；
wherein, remember
Figure 536488DEST_PATH_IMAGE010
，
Figure 466398DEST_PATH_IMAGE011
，
Figure 692980DEST_PATH_IMAGE012
；
Figure 336537DEST_PATH_IMAGE013
And
Figure 408399DEST_PATH_IMAGE014
is composed of
Figure 723973DEST_PATH_IMAGE015
I is the length of the subsequence;
then
Figure 714932DEST_PATH_IMAGE016
；
For example, for time series
Figure 240591DEST_PATH_IMAGE032
Mean value of
Figure 991510DEST_PATH_IMAGE033
The sequence length is 12, and the autocorrelation sequence of the sequence X can be calculated according to the autocorrelation function
Figure 145279DEST_PATH_IMAGE034
；
Step 322, processing the autocorrelation sequence to obtain a similarity difference sequence of the autocorrelation sequence;
step 323, obtaining a candidate period T of the time sequence data according to the similarity difference sequence;
step 324, obtaining the noise degree according to the candidate period T and the time sequence data
Figure 182505DEST_PATH_IMAGE001
And specific gravity of trend component
Figure 70827DEST_PATH_IMAGE002
；
Step 325, according to the noise level
Figure 484491DEST_PATH_IMAGE001
And specific gravity of trend component
Figure 758346DEST_PATH_IMAGE002
And obtaining a detection result of whether the time series data are periodic time series or not according to a corresponding threshold value.
In an alternative embodiment of the present invention, step 322 may include:
computing subsequence
Figure 904157DEST_PATH_IMAGE013
And
Figure 997884DEST_PATH_IMAGE014
obtaining similarity difference sequences M through similarity difference under each length;
wherein,
Figure 949659DEST_PATH_IMAGE035
，
Figure 94333DEST_PATH_IMAGE018
are likeThe degree difference calculation function may be, for example, similarity difference measurement modes such as MSE, RMSE, MAE, DTW, and the like.
In an implementation example, the similarity difference calculation function may be an average absolute error
Figure 597995DEST_PATH_IMAGE036
Wherein
Figure 54384DEST_PATH_IMAGE037
represents the above-mentioned subsequence
Figure 950796DEST_PATH_IMAGE013
，
Figure 809031DEST_PATH_IMAGE038
Represents the above-mentioned subsequence
Figure 14753DEST_PATH_IMAGE014
。
For example, for time series
Figure 833804DEST_PATH_IMAGE032
The autocorrelation sequence of which is
Figure 392962DEST_PATH_IMAGE039
Using the average absolute error as the similarity measure, then
Figure 230337DEST_PATH_IMAGE040
，
Figure 216747DEST_PATH_IMAGE041
(ii) a Similarity difference sequence
Figure 523095DEST_PATH_IMAGE042
。
In an alternative embodiment of the present invention, step 323 may include:
step 3231, ordering according to the similarity difference sequence
Figure 10577DEST_PATH_IMAGE019
By the formula
Figure 577824DEST_PATH_IMAGE020
Obtaining a candidate period T of the time sequence data; wherein j is the index of the element M in the similarity difference sequence M.
In an alternative embodiment of the present invention, step 324 may include:
step 3241, obtaining noise level according to the time sequence data
Figure 610502DEST_PATH_IMAGE001
And specific gravity of trend component
Figure 794359DEST_PATH_IMAGE002
。
In particular, according to
Figure 554374DEST_PATH_IMAGE021
Obtaining the degree of noise
Figure 382652DEST_PATH_IMAGE001
Or according to
Figure 710865DEST_PATH_IMAGE022
Obtaining the degree of noise
Figure 975493DEST_PATH_IMAGE001
(ii) a Here, ,
Figure 680144DEST_PATH_IMAGE023
period of time-series data
Figure 97350DEST_PATH_IMAGE043
When the number of the carbon atoms is 7,
Figure 862044DEST_PATH_IMAGE044
。
in particular, according to
Figure 613968DEST_PATH_IMAGE024
Obtaining the specific gravity of the trend component
Figure 732097DEST_PATH_IMAGE002
Or according to
Figure 394022DEST_PATH_IMAGE025
Obtaining the specific gravity of the trend component
Figure 188672DEST_PATH_IMAGE002
Or according to
Figure 568838DEST_PATH_IMAGE026
Obtaining the specific gravity of the trend component
Figure 225078DEST_PATH_IMAGE002
。
In an alternative embodiment of the present invention, step 325 may include:
if it is
Figure 741510DEST_PATH_IMAGE027
And, in addition,
Figure 972640DEST_PATH_IMAGE028
then, the time series data X is considered as a periodic time series with a period length T.
Here, based on T, the noise level of the sequence is calculated
Figure 449889DEST_PATH_IMAGE001
And specific gravity of trend component
Figure 300033DEST_PATH_IMAGE002
，
Figure 795606DEST_PATH_IMAGE001
For describing the degree of noise on a sequence, the smaller the value, the smaller the noise, and the more significant the period.
Figure 807424DEST_PATH_IMAGE002
Description of the inventionThe smaller the degree, value, of the trending component on the sequence, the fewer the trending component, and the more pronounced the period.
If it is
Figure 506390DEST_PATH_IMAGE027
And, in addition,
Figure 284859DEST_PATH_IMAGE028
then, the time series data X is considered as a periodic time series with a period length T.
Wherein j is the index of the element M in the similarity difference sequence M;
Figure 510304DEST_PATH_IMAGE029
a preset noise threshold value,
Figure 568390DEST_PATH_IMAGE030
Is a preset trend threshold.
For example:
Figure 410444DEST_PATH_IMAGE045
；
Figure 473164DEST_PATH_IMAGE046
(ii) a In the M sequence, 0 is the smallest
Figure 553115DEST_PATH_IMAGE047
If the corresponding index j is 6, then pass the formula
Figure 313261DEST_PATH_IMAGE020
Obtaining the candidate period of the time series data
Figure 236086DEST_PATH_IMAGE043
Is 7.
In an alternative embodiment of the present invention, step 34 may include at least one of:
step 341, performing anomaly detection processing on preset time sequence data according to the target time sequence to obtain a detection result; as shown in fig. 7, the fixed threshold detection on the residual sequence is more accurate;
step 342, classifying preset time sequence data according to the target time sequence to obtain a classification result;
step 343, according to the target time sequence, carrying out prediction processing on preset time sequence data to obtain a prediction processing result;
according to the embodiment of the invention, whether the sequence is a periodic sequence can be judged through periodic detection of the time sequence data of the target object of the operation and maintenance system, and the preset sequence data is classified; the period length of the sequence can be automatically acquired; cycles can be eliminated from the sequence to facilitate subsequent anomaly detection and prediction processing.
In an operation and maintenance scene, the method is used for carrying out cycle detection and extraction on time sequence data of a target object, cycles are removed from an original sequence, abnormal detection, classification or prediction processing and the like are carried out on the removed data, if abnormal detection is carried out, as shown in fig. 4, the time sequence data of the target object is index data (unit: millisecond) of database calling delay, the index data is displayed in a week from 2020-05-15 to 2020-05-22, and an operation and maintenance person carries out analysis according to experience: from the index trend analysis, the data has obvious day period, and the analysis can know that in the vicinity of 2020-05-1812. The reason for analyzing by the operation and maintenance personnel may be that the operation and maintenance personnel are affected by a fault, so that the failure rate of database calling at the moment is high, and further, the whole sudden drop of calling delay is caused. When a fixed threshold value is adopted to process the data, the suspected abnormal points are not determined to be abnormal, but are determined to be normal data, and the data are inconsistent with the analysis of operation and maintenance personnel. By periodically detecting and extracting the data, the method of the present invention can extract periodic components from the index data, and obtain the processing result as shown in fig. 5 (a). Comparing the cycle component with the original index data, the abnormal values of 2020-05-18. Then, the periodic components are removed from the original sequence to obtain the processing result shown in fig. 6, and the residual sequence shown in fig. 6 (a) is obtained. By this processing, it can be seen from fig. 6 (b) that the abnormal data is more vivid on the residual sequence, and then the residual sequence is detected by using a fixed threshold or the like, so that the abnormality can be effectively identified, and the processing result shown in fig. 7 is obtained, so that the method of the present invention can identify the abnormality (the abnormal point of 2020-05-18 00.
According to the method provided by the embodiment of the invention, the period of the periodic time sequence is removed from the time sequence data to obtain a target time sequence based on the periodic detection of the time sequence data; according to the target time sequence, carrying out detection, classification and/or prediction processing on abnormal indexes of the target object to obtain a processing result, and outputting the processing result; the abnormity detection, classification or early warning processing of the target object of the operation and maintenance system is realized, and the operation performance of the operation and maintenance system is improved.
As shown in fig. 8, an embodiment of the present invention further provides a device 80 for processing time series data of an operation and maintenance system, where the device 80 includes:
the obtaining module 81 is configured to obtain time series data of a target object of the operation and maintenance system;
the processing module 82 is configured to perform periodic detection processing on the time series data to obtain a detection result of whether the time series data is a periodic time series; if the detection result shows that the time sequence data is a cycle time sequence, removing the cycle of the cycle time sequence from the time sequence data to obtain a target time sequence; and according to the target time sequence, carrying out detection, classification and/or prediction processing on the abnormal indexes of the target object to obtain a processing result, and outputting the processing result.
Optionally, the performing a periodic detection process on the time series data to obtain a detection result of whether the time series data is a periodic time series includes:
processing the time sequence data to obtain an autocorrelation sequence of the time sequence data;
processing the self-correlation sequence to obtain a similarity difference sequence of the self-correlation sequence;
obtaining a candidate period T of the time sequence data according to the similarity difference sequence;
obtaining the noise degree according to the candidate period T and the time sequence data
Figure 700566DEST_PATH_IMAGE001
And specific gravity of trend component
Figure 41548DEST_PATH_IMAGE002
；
According to the noise level
Figure 831650DEST_PATH_IMAGE001
And the specific gravity of the trend component
Figure 241772DEST_PATH_IMAGE002
And obtaining a detection result of whether the time series data is a periodic time series or not according to a corresponding threshold value.
Optionally, processing the time-series data to obtain an autocorrelation sequence of the time-series data includes:
according to an autocorrelation function, carrying out autocorrelation calculation on the time sequence data to obtain an autocorrelation sequence of the time sequence data; the autocorrelation function is:
Figure 650887DEST_PATH_IMAGE003
；
wherein the time-series data is
Figure 705431DEST_PATH_IMAGE004
N represents the length of the time-series data X;
Figure 791067DEST_PATH_IMAGE005
is the average of the time-series data X,
Figure 563851DEST_PATH_IMAGE006
；
Figure 511079DEST_PATH_IMAGE007
indicates the t-th point in the time-series data,
Figure 420129DEST_PATH_IMAGE008
。
optionally, the autocorrelation sequence of the time-series data is:
Figure 942246DEST_PATH_IMAGE048
；
wherein, note
Figure 812113DEST_PATH_IMAGE010
，
Figure 953244DEST_PATH_IMAGE011
，
Figure 107014DEST_PATH_IMAGE012
；
Figure 144240DEST_PATH_IMAGE013
And
Figure 32562DEST_PATH_IMAGE014
is composed of
Figure 570859DEST_PATH_IMAGE015
I is the length of the subsequence;
then the
Figure 720081DEST_PATH_IMAGE049
。
Optionally, obtaining a similarity difference sequence of the auto-correlation sequence includes:
computing subsequence
Figure 69154DEST_PATH_IMAGE013
And
Figure 162881DEST_PATH_IMAGE014
obtaining similarity difference sequences M through similarity difference under each length;
wherein,
Figure 645815DEST_PATH_IMAGE035
，
Figure 259330DEST_PATH_IMAGE018
a function is calculated for the similarity difference.
Optionally, obtaining the candidate period T of the time series data according to the similarity difference sequence includes:
according to the similarity difference sequence, order
Figure 903938DEST_PATH_IMAGE019
By the formula
Figure 750540DEST_PATH_IMAGE020
Obtaining a candidate period T of the time sequence data; wherein j is the index of the element M in the similarity difference sequence M.
Optionally, a noise level is obtained according to the candidate period T and the time-series data
Figure 381372DEST_PATH_IMAGE001
And specific gravity of trend component
Figure 833082DEST_PATH_IMAGE002
The method comprises the following steps:
according to
Figure 648592DEST_PATH_IMAGE021
Obtaining the noise level
Figure 716911DEST_PATH_IMAGE001
Or, according to
Figure 417013DEST_PATH_IMAGE022
Obtaining the noise level
Figure 129754DEST_PATH_IMAGE001
(ii) a Wherein,
Figure 506378DEST_PATH_IMAGE023
；
according to
Figure 937359DEST_PATH_IMAGE024
Obtaining the specific gravity of the trend component
Figure 909995DEST_PATH_IMAGE002
Or, according to
Figure 867455DEST_PATH_IMAGE025
Obtaining the specific gravity of the trend component
Figure 24767DEST_PATH_IMAGE002
Or, according to
Figure 818411DEST_PATH_IMAGE026
Obtaining the specific gravity of the trend component
Figure 719371DEST_PATH_IMAGE002
。
Optionally, according to the degree of noise
Figure 531338DEST_PATH_IMAGE001
And specific gravity of trend component
Figure 496DEST_PATH_IMAGE002
And a corresponding threshold, determining the time series data as a periodic time series, comprising:
if it is
Figure 140491DEST_PATH_IMAGE027
And, in addition,
Figure 704196DEST_PATH_IMAGE028
if the time sequence data X is a periodic time sequence, the period length is T;
wherein,
Figure 511615DEST_PATH_IMAGE029
a preset noise threshold value,
Figure 73046DEST_PATH_IMAGE030
Is a preset trend threshold.
It should be noted that the apparatus is an apparatus corresponding to the above method, and all the implementations in the above method embodiment are applicable to the embodiment of the apparatus, and the same technical effects can be achieved.
Embodiments of the present invention also provide a computing device, comprising: the system comprises a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface complete mutual communication through the communication bus; the memory is used for storing at least one executable instruction, and the executable instruction causes the processor to execute the corresponding operation of the method.
Embodiments of the present invention also provide a computer-readable storage medium storing instructions that, when executed on a computer, cause the computer to perform the method as described above.
Those of ordinary skill in the art will appreciate that the various illustrative elements and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware or combinations of computer software and electronic hardware. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems, apparatuses and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
In the embodiments provided in the present invention, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the units is only one logical division, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.
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 position, or may be distributed on multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of the embodiment.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit.
The functions may be stored in a computer-readable storage medium if they are implemented in the form of software functional units and sold or used as separate products. 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: various media capable of storing program codes, such as a U disk, a removable hard disk, a ROM, a RAM, a magnetic disk, or an optical disk.
Furthermore, it is to be noted that in the device and method of the invention, it is obvious that the individual components or steps can be decomposed and/or recombined. These decompositions and/or recombinations are to be regarded as equivalents of the present invention. Also, the steps of performing the series of processes described above may naturally be performed chronologically in the order described, but need not necessarily be performed chronologically, and some steps may be performed in parallel or independently of each other. It will be understood by those skilled in the art that all or any of the steps or elements of the method and apparatus of the present invention may be implemented in any computing device (including processors, storage media, etc.) or network of computing devices, in hardware, firmware, software, or any combination thereof, which can be implemented by those skilled in the art using their basic programming skills after reading the description of the present invention.
The object of the invention is thus also achieved by a program or a set of programs running on any computing device. The computing device may be a general purpose device as is well known. The object of the invention is thus also achieved solely by providing a program product comprising program code for implementing the method or the apparatus. That is, such a program product also constitutes the present invention, and a storage medium storing such a program product also constitutes the present invention. It is to be understood that the storage medium may be any known storage medium or any storage medium developed in the future. It is further noted that in the apparatus and method of the present invention, it is apparent that each component or step can be decomposed and/or recombined. These decompositions and/or recombinations are to be regarded as equivalents of the present invention. Also, the steps of executing the series of processes described above may naturally be executed chronologically in the order described, but need not necessarily be executed chronologically. Some steps may be performed in parallel or independently of each other.
While the foregoing is directed to the preferred embodiment of the present invention, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention as defined in the appended claims.

Claims (10)

1. A method for processing time series data of an operation and maintenance system is characterized by comprising the following steps:
acquiring time sequence data of a target object of the operation and maintenance system;
carrying out periodic detection processing on the time sequence data to obtain a detection result of whether the time sequence data is a periodic time sequence;
if the detection result shows that the time sequence data is a cycle time sequence, removing the cycle of the cycle time sequence from the time sequence data to obtain a target time sequence;
and according to the target time sequence, carrying out detection, classification and/or prediction processing on the abnormal indexes of the target object to obtain a processing result, and outputting the processing result.
2. The method for processing time series data of an operation and maintenance system according to claim 1, wherein performing a periodic detection process on the time series data to obtain a detection result of whether the time series data is a periodic time series includes:
processing the time sequence data to obtain an autocorrelation sequence of the time sequence data;
processing the autocorrelation sequence to obtain a similarity difference sequence of the autocorrelation sequence;
obtaining a candidate period T of the time sequence data according to the similarity difference sequence;
obtaining a noise degree noise _ level and a trend component proportion trend _ level according to the candidate period T and the time sequence data;
and obtaining a detection result of whether the time sequence data is a periodic time sequence or not according to the noise degree noise _ level, the trend component specific gravity trend _ level and a corresponding threshold value.
3. The method for processing time series data of an operation and maintenance system according to claim 2, wherein processing the time series data to obtain an autocorrelation sequence of the time series data comprises:
according to an autocorrelation function, carrying out autocorrelation calculation on the time sequence data to obtain an autocorrelation sequence of the time sequence data; the autocorrelation function is:
Figure 239216DEST_PATH_IMAGE001
；
wherein the time-series data is
Figure 606612DEST_PATH_IMAGE002
N represents the length of the time-series data X;
Figure 268538DEST_PATH_IMAGE003
is the average of the time-series data X,
Figure 813920DEST_PATH_IMAGE004
；
Figure 866190DEST_PATH_IMAGE005
indicates the t-th point in the time-series data,
Figure 912643DEST_PATH_IMAGE006
。
4. the method for processing time series data of an operation and maintenance system according to claim 3, wherein the autocorrelation sequence of the time series data is:
Figure 819288DEST_PATH_IMAGE007
；
wherein, note
Figure 597888DEST_PATH_IMAGE008
，
Figure 58825DEST_PATH_IMAGE009
，
Figure 643390DEST_PATH_IMAGE010
；
Figure 889695DEST_PATH_IMAGE011
And
Figure 901513DEST_PATH_IMAGE012
is composed of
Figure 193955DEST_PATH_IMAGE013
I is the length of the subsequence;
then
Figure 441265DEST_PATH_IMAGE014
。
5. The method for processing time series data of an operation and maintenance system according to claim 4, wherein the obtaining of the similarity difference sequence of the autocorrelation sequence comprises:
computing subsequence
Figure 197869DEST_PATH_IMAGE011
And
Figure 255954DEST_PATH_IMAGE012
obtaining similarity difference sequences M through similarity difference under each length;
wherein,
Figure 566850DEST_PATH_IMAGE015
，
Figure 696480DEST_PATH_IMAGE016
a function is calculated for the similarity difference.
6. The method for processing time series data of an operation and maintenance system according to claim 5, wherein obtaining the candidate period T of the time series data according to the similarity difference sequence comprises:
according to the similarity difference sequence, order
Figure 166645DEST_PATH_IMAGE017
By the formula
Figure 785845DEST_PATH_IMAGE018
Obtaining a candidate period T of the time sequence data; wherein j is the index of the element M in the similarity difference sequence M.
7. The method for processing time series data of an operation and maintenance system according to claim 6, wherein obtaining a noise level noise _ level and a trend component specific gravity trend _ level according to the candidate period T and the time series data comprises:
according to
Figure 459403DEST_PATH_IMAGE019
Obtain the noise level noise _ level, or, according to
Figure 189461DEST_PATH_IMAGE020
Obtaining the noise degree noise _ level; wherein,
Figure 592761DEST_PATH_IMAGE021
；
according to
Figure 241917DEST_PATH_IMAGE022
Obtaining the trend component specific gravity trend _ level, or, according to
Figure 527405DEST_PATH_IMAGE023
The trend component specific gravity trend _ level is obtained, or, according to
Figure 202100DEST_PATH_IMAGE024
The trend component specific gravity trend _ level is obtained.
8. The method for processing time series data of an operation and maintenance system according to claim 7, wherein determining the time series data as a periodic time series according to the noise level noise _ level, the trend component specific gravity trend _ level, and a corresponding threshold value comprises:
if it is
Figure 912435DEST_PATH_IMAGE025
And, in addition,
Figure 607859DEST_PATH_IMAGE026
if the time sequence data X is a periodic time sequence, the period length is T;
wherein,
Figure 583905DEST_PATH_IMAGE027
a noise threshold is preset,
Figure 531133DEST_PATH_IMAGE028
Is a preset trend threshold.
9. An apparatus for processing time series data of an operation and maintenance system, the apparatus comprising:
the acquisition module is used for acquiring time series data of a target object of the operation and maintenance system;
the processing module is used for carrying out periodic detection processing on the time sequence data to obtain a detection result of whether the time sequence data is a periodic time sequence; if the detection result shows that the time sequence data is a cycle time sequence, removing the cycle of the cycle time sequence from the time sequence data to obtain a target time sequence; and according to the target time sequence, carrying out detection, classification and/or prediction processing on the abnormal indexes of the target object to obtain a processing result, and outputting the processing result.
10. A computer-readable storage medium having stored thereon instructions which, when executed on a computer, cause the computer to perform the method of any one of claims 1 to 8.
CN202211093088.7A 2021-12-08 2022-09-08 Method and device for processing time sequence data of operation and maintenance system Pending CN115659185A (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202111485747.7A CN113886470A (en) 2021-12-08 2021-12-08 Method, device and equipment for processing time series data
CN2021114857477 2021-12-08

Publications (1)

Publication Number Publication Date
CN115659185A true CN115659185A (en) 2023-01-31

Family

ID=79015795

Family Applications (2)

Application Number Title Priority Date Filing Date
CN202111485747.7A Pending CN113886470A (en) 2021-12-08 2021-12-08 Method, device and equipment for processing time series data
CN202211093088.7A Pending CN115659185A (en) 2021-12-08 2022-09-08 Method and device for processing time sequence data of operation and maintenance system

Family Applications Before (1)

Application Number Title Priority Date Filing Date
CN202111485747.7A Pending CN113886470A (en) 2021-12-08 2021-12-08 Method, device and equipment for processing time series data

Country Status (1)

Country Link
CN (2) CN113886470A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116652396A (en) * 2023-08-01 2023-08-29 南通大学 Safety early warning method and system for laser inner carving machine

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20230341832A1 (en) * 2022-04-26 2023-10-26 Hitachi, Ltd. Versatile anomaly detection system for industrial systems

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116652396A (en) * 2023-08-01 2023-08-29 南通大学 Safety early warning method and system for laser inner carving machine
CN116652396B (en) * 2023-08-01 2023-10-10 南通大学 Safety early warning method and system for laser inner carving machine

Also Published As

Publication number Publication date
CN113886470A (en) 2022-01-04

Similar Documents

Publication Publication Date Title
CN115659185A (en) Method and device for processing time sequence data of operation and maintenance system
US10373065B2 (en) Generating database cluster health alerts using machine learning
JP6361175B2 (en) Abnormality diagnosis apparatus and program
EP2015186B1 (en) Diagnostic systems and methods for predictive condition monitoring
EP3552067B1 (en) Methods and systems for discovery of prognostic subsequences in time series
CN105259895B (en) A kind of detection of industrial process small fault and separation method and its monitoring system
EP2905665A2 (en) Information processing apparatus, diagnosis method, and program
JP2017196523A5 (en) Brain activity analysis system, brain activity analysis device, and brain activity analysis program
CN108549914B (en) Abnormal SER/SOE event identification method based on Apriori algorithm
AU2002246994A1 (en) Diagnostic systems and methods for predictive condition monitoring
CN110008247B (en) Method, device and equipment for determining abnormal source and computer readable storage medium
EP3333757B1 (en) Predictive anomaly detection
CN112581719B (en) Semiconductor packaging process early warning method and device based on time sequence generation countermeasure network
EP4160339A1 (en) Abnormality/irregularity cause identifying apparatus, abnormality/irregularity cause identifying method, and abnormality/irregularity cause identifying program
CN116304909A (en) Abnormality detection model training method, fault scene positioning method and device
EP4160341A1 (en) Abnormal modulation cause identifying device, abnormal modulation cause identifying method, and abnormal modulation cause identifying program
CN114297264A (en) Method and system for detecting abnormal segments of time sequence signal
CN116974268B (en) Intelligent monitoring and early warning method for control system circuit
JP2009110262A (en) Filter device, filtering method for filter device and filtering program
CN117591860A (en) Data anomaly detection method and device
US11449781B2 (en) Plant abnormality prediction system and method
CN113571092B (en) Engine abnormal sound identification method and related equipment thereof
US20220253051A1 (en) Method and assistance system for detecting an abnormal behaviour of a device
CN114580982B (en) Method, device and equipment for evaluating data quality of industrial equipment
JP7449248B2 (en) Support devices, support methods and support programs

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