CN110543505B - Monitoring system based on time series data - Google Patents

Monitoring system based on time series data Download PDF

Info

Publication number
CN110543505B
CN110543505B CN201910844017.8A CN201910844017A CN110543505B CN 110543505 B CN110543505 B CN 110543505B CN 201910844017 A CN201910844017 A CN 201910844017A CN 110543505 B CN110543505 B CN 110543505B
Authority
CN
China
Prior art keywords
data
series data
index
time series
monitoring
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201910844017.8A
Other languages
Chinese (zh)
Other versions
CN110543505A (en
Inventor
李晓波
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.)
Beijing Yuanshan Intelligent Technology Co Ltd
Original Assignee
Beijing Yuanshan Intelligent 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 Beijing Yuanshan Intelligent Technology Co Ltd filed Critical Beijing Yuanshan Intelligent Technology Co Ltd
Priority to CN201910844017.8A priority Critical patent/CN110543505B/en
Publication of CN110543505A publication Critical patent/CN110543505A/en
Application granted granted Critical
Publication of CN110543505B publication Critical patent/CN110543505B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

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
    • G06F16/2465—Query processing support for facilitating data mining operations in structured databases
    • 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

Abstract

The invention provides a monitoring system based on time series data, which comprises a monitoring project dividing subsystem, a monitoring data acquisition subsystem, a monitoring data mining subsystem and a monitoring early warning subsystem, wherein the monitoring project dividing subsystem divides a monitoring project into a single-index monitoring project and a multi-index monitoring project, the monitoring data acquisition subsystem acquires the time series data of indexes in the single-index monitoring project and the multi-index monitoring project, the monitoring data mining subsystem carries out classification mining on the acquired time series data, and the monitoring early warning subsystem carries out monitoring and early warning on the monitoring project based on the classification mining result of the time series data. The monitoring system based on the time sequence data realizes the whole process from time sequence data acquisition to classification mining no matter a single index monitoring project or a multi-index monitoring project, and realizes effective monitoring on the monitoring project.

Description

Monitoring system based on time series data
Technical Field
The invention relates to the technical field of monitoring, in particular to a monitoring system based on time series data.
Background
With the progress of human society and the development of scientific technology, time series data has an increasing role, and various aspects in production and life, such as atmospheric environment monitoring, human health monitoring, national economy monitoring and the like, can be monitored through the time series data, however, how to systematically complete monitoring activities by using the time series data still does not have a complete solution at present.
Disclosure of Invention
In order to solve the technical problems, the invention provides a monitoring system based on time series data, which realizes the whole process from time series data acquisition to classification mining no matter a single index monitoring project or a multi-index monitoring project, and realizes effective monitoring on the monitoring project.
The invention particularly provides a monitoring system based on time series data, which comprises a monitoring project dividing subsystem, a monitoring data acquisition subsystem, a monitoring data mining subsystem and a monitoring early warning subsystem, wherein the monitoring project dividing subsystem divides a monitoring project into a single-index monitoring project and a multi-index monitoring project, the monitoring data acquisition subsystem acquires the time series data of indexes in the single-index monitoring project and the multi-index monitoring project, the monitoring data mining subsystem carries out classification mining on the acquired time series data, and the monitoring early warning subsystem carries out monitoring and early warning on the monitoring project based on the classification mining result of the time series data.
Optionally, the monitoring data mining subsystem includes a data dividing module, a data measuring module and a data classifying module, the data dividing module divides the time series data into single-index time series data and multi-index time series data, the data measuring module is configured to perform similarity measurement on the single-index time series data and the multi-index time series data, and the data classifying module performs classified mining on the single-index time series data and the multi-index time series data based on a result of the similarity measurement.
Optionally, the data measurement module includes a first measurement unit and a second measurement unit, the first measurement unit is configured to perform similarity measurement on single-index time-series data, and the second measurement unit is configured to perform similarity measurement on multi-index time-series data;
the first measurement unit is used for carrying out similarity measurement on the single-index time series data and comprises the following steps:
for single index time series data A1＝(a1，a2，…，an) And A2＝(b1，b2，…，bn) Wherein n represents single index time-series data A1And A2The first similarity factor of the single index time series is calculated by adopting the following formula:
Figure BDA0002194596640000021
in the formula, S1Representing a first similarity factor, a, of a time series of a single indexi、biRespectively represent single index time series data A1And A2The ith time-series data of (1); the larger the first similarity factor of the single index time series is, the single index time series data A is represented1And A2The higher the similarity.
Optionally, the first metric unit is configured to perform similarity measurement on the single-index time-series data, and further includes:
for single index time series data A1＝(a1，a2，…，an) And A3＝(c1，c2，…，cm) Wherein n and m respectively represent single index time-series data A1And A3If the number of data in (1) is not equal to m, then A is first added1And A3Single index time series data A extended to length d1′＝(a1，a2，…，an，an+1，…，ad) And A3′＝(c1，c2，…，cm，cmm+1，…，cd) Wherein d represents the extended single index time series data A1' and A3' the number of data, d > n and d > m,
Figure BDA0002194596640000022
Figure BDA0002194596640000023
ai、cirespectively represent single index time series data A1And A3The ith time-series data of (1), calculating a single index time-series second similarity factor using the following formula:
Figure BDA0002194596640000024
in the formula, S2Representing a second similarity factor of a time series of a single index, aj、cjRespectively represent the extended single index time series data A'1And A'3The jth time-series data in (a); the larger the second similarity factor of the single index time series is, the single index time series data A is represented1And A3The higher the similarity.
Optionally, the second metric unit is configured to perform similarity measurement on the multi-index time series data, and includes:
for multi-index time series data E1＝{X1，X2，…，Xn1And E2＝{Y1，Y2，…，Yn1In which Xi＝(xi1，xi2，…，xik)T，Yi＝(yi1，yi2，…，yik)TI-1, 2, …, n1, n1 represent multi-index time-series data E1And E2Length of (1), XiAnd YiRespectively represent multi-index time series data E1And E2K represents multi-index time-series data E1And E2The number of the medium indexes, k is more than or equal to 2, and a first similarity factor of the multi-index time sequence is calculated by adopting the following formula:
Figure BDA0002194596640000025
in the formula, F1Representing a multi-index time series first similarity factor; the larger the first similarity factor of the multi-index time series is, the more the multi-index time series data E is represented1And E2The higher the similarity.
Optionally, the second metric unit is configured to perform similarity measurement on the multi-index time series data, and further includes:
for multi-index time series data E1＝{X1，X2，…，Xn1And E3＝{Z1，Z2，…，Zn2In which Xi＝(xi1，xi2，…，xik)T，Zi＝(Zi1，Zi2，…，Zik)TN1 and n2 respectively represent multi-index time-series data E1And E3And n1 ≠ n2, XiAnd ZiRespectively representing multi-index time series dataE1And E3K represents multi-index time-series data E1And E3The number of the medium indexes, k is more than or equal to 2;
first, E is1And E3Extended to multi-index time series data E 'with length w'1＝{X1，X2，…，Xn1，Xn1+1，…，XwAnd E'3＝{Z1，Z2，…，Zn2，Zn2+1，…，ZwW > n1 and w > n2,
Figure BDA0002194596640000031
Figure BDA0002194596640000032
calculating a multi-index time series second similarity factor using the following formula:
Figure BDA0002194596640000033
in the formula, F2Representing a multi-index time series second similarity factor; the larger the second similarity factor of the multi-index time series is, the more the multi-index time series data E is represented1And E3The higher the similarity.
Optionally, the monitoring data mining subsystem performs classification mining on the acquired time series data, specifically:
respectively carrying out similarity measurement on the two time sequence data subjected to similarity measurement and the known standard time sequence data by adopting the same similarity measurement method, presetting a difference judgment threshold value, and if the difference value of the similarity measurement between the two time sequence data and the known standard time sequence data is greater than the preset difference judgment threshold value, indicating that the difference between the two time sequence data is greater, and at the moment, setting the difference factor of the two time sequence data as P1，P1If the difference value of the similarity measure between the two time data sequences and the known standard time sequence data is less than the preset difference judgmentIf the threshold is cut off, the difference between the two time data sequences is small, and at the moment, the difference factor of the two time data sequences is set as P2，P2＞1；
Multiplying the similarity measurement result of the two time sequence data by the difference factor of the two time sequence data to obtain an improved similarity measurement result of the time sequence data;
and classifying the time series data by adopting a k-nearest neighbor algorithm based on the similarity measurement result of the improved time series data, and mining the data based on the classification result of the time series data.
Optionally, the multi-index monitoring item is a multi-index mechanical equipment monitoring item, and the multi-index includes mechanical equipment vibration data, mechanical equipment temperature data, and mechanical equipment sound data.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
Drawings
The invention is further illustrated by means of the attached drawings, but the embodiments in the drawings do not constitute any limitation to the invention, and for a person skilled in the art, other drawings can be obtained on the basis of the following drawings without inventive effort.
FIG. 1 is a schematic view of the structure of the present invention.
Reference numerals:
the monitoring project division subsystem 1, the monitoring data acquisition subsystem 2, the monitoring data mining subsystem 3 and the monitoring early warning subsystem 4.
Detailed Description
The invention is further described with reference to the following examples.
With reference to fig. 1, an embodiment of the present invention provides a monitoring system based on time series data, including a monitoring project dividing subsystem 1, a monitoring data collecting subsystem 2, a monitoring data mining subsystem 3, and a monitoring and early warning subsystem 4, where the monitoring project dividing subsystem 1 divides a monitoring project into a single-index monitoring project and a multi-index monitoring project, the monitoring data collecting subsystem 2 collects time series data of indexes in the single-index monitoring project and the multi-index monitoring project, the monitoring data mining subsystem 3 performs classified mining on the collected time series data, and the monitoring and early warning subsystem 4 monitors and early warns the monitoring project based on a classified mining result of the time series data.
In the actual production life, various monitoring projects exist, some monitoring projects only need to collect time series data of a first variable, such as monitoring the body temperature of a fever patient, monitoring the concentration of PM2.5 in the air and monitoring a vibration signal of mechanical equipment, at the moment, the monitoring projects are used as single index monitoring projects, a sensor only needs to collect the time series data of a single index, some monitoring projects need to detect a plurality of variables, such as comprehensively monitoring the body health of blood pressure, blood sugar and blood fat and comprehensively monitoring the weather of temperature, humidity and air pressure, at the moment, the monitoring projects are multi-index monitoring projects, and the sensor needs to collect multi-index time;
for the acquisition of time sequence data, various classification modes can be adopted for classification, such as k-nearest neighbor classification, artificial neural network classification, Bayesian classification and the like, and human body temperature is taken as an example, and the time sequence data can be classified into low body temperature, normal fever, low fever and high fever according to the body temperature.
And acquiring the required data based on the classification result so as to finish data mining, and further monitoring and early warning the corresponding monitoring items according to the data classification mining result.
In the embodiment, the whole process from time sequence data acquisition to classification mining is realized no matter a single index monitoring project or a multi-index monitoring project, and effective monitoring on the monitoring project is realized.
Preferably, the monitoring data mining subsystem 3 includes a data dividing module, a data measuring module and a data classifying module, the data dividing module divides the time series data into single-index time series data and multi-index time series data, the data measuring module is used for performing similarity measurement on the single-index time series data and the multi-index time series data, and the data classifying module performs classified mining on the single-index time series data and the multi-index time series data based on a result of the similarity measurement.
In the data mining process, for the reason that the processing methods and the complexity of single-index time series data and multi-index time series data are different, in order to reduce the difficulty of data processing and improve the data processing efficiency, the monitoring data are firstly divided according to different monitoring items, then corresponding similarity measurement is carried out on different data types, and the corresponding data are classified and mined based on the similarity measurement method.
Preferably, the data measurement module comprises a first measurement unit and a second measurement unit, the first measurement unit is used for carrying out similarity measurement on single-index time-series data, and the second measurement unit is used for carrying out similarity measurement on multi-index time-series data;
the first measurement unit is used for carrying out similarity measurement on the single-index time series data and comprises the following steps:
for single index time series data A1＝(a1，a2，…，an) And A2＝(b1，b2，…，bn) Wherein n represents single index time-series data A1And A2The first similarity factor of the single index time series is calculated by adopting the following formula:
Figure BDA0002194596640000051
in the formula, S1Representing a first similarity factor, a, of a time series of a single indexi、biRespectively represent single index time series data A1And A2The ith time-series data of (1); the larger the first similarity factor of the single index time series is, the single index time series data A is represented1And A2The higher the similarity.
For the characteristics of time series data in the actual monitoring process, the similarity measurement is performed on the condition that the data lengths in the single-index time series data are the same, the data meeting the conditions can be effectively measured by determining the first similarity factor of the single-index time series, and the first similarity factor of the single-index time series can be used as a similarity measurement result for subsequent calculation.
Preferably, the first metric unit is configured to perform similarity measurement on the single-index time-series data, and further includes:
for single index time series data A1＝(a1，a2，…，an) And A3＝(c1，c2，…，cm) Wherein n and m respectively represent single index time-series data A1And A3If the number of data in (1) is not equal to m, then A is first added1And A3Single index time series data A extended to length d1′＝(a1，a2，…，an，an+1，…，ad) And A3′＝(c1，c2，…，cm，cm+1，…，cd) Wherein d represents the extended single index time series data A1' and A3' the number of data, d > n and d > m,
Figure BDA0002194596640000052
Figure BDA0002194596640000053
ai、cirespectively represent single index time series data A1And A3The ith time-series data of (1), calculating a single index time-series second similarity factor using the following formula:
Figure BDA0002194596640000054
in the formula, S2Representing a second similarity factor of a time series of a single index, aj、cjRespectively represent the extended single index time series data A'1And A'3The jth time-series data in (a); the larger the second similarity factor of the single index time series is, the single index time series data A is represented1And A3The higher the similarity.
In the calculation process of the second similarity factor of the single index time sequence, d is larger than the maximum value of n and m, so that the time sequence data with different lengths can be conveniently expanded.
In the preferred embodiment, the single index time series data with different lengths are expanded, so that the similarity measurement of the single index time series data with different lengths is realized, and the second similarity factor of the single index time series can be used as a similarity measurement result for subsequent calculation.
Preferably, the second metric unit is configured to perform similarity measurement on the multi-index time-series data, and includes:
for multi-index time series data E1＝{X1，X2，…，Xn1And E2＝{Y1，Y2，…，Yn1In which Xi＝(xi1，xi2，…，xik)T，Yi＝(yi1，yi2，…，yik)TI-1, 2, …, n1, n1 represent multi-index time-series data E1And E2Length of (1), XiAnd YiRespectively representMulti-index time series data E1And E2K represents multi-index time-series data E1And E2The number of the medium indexes, k is more than or equal to 2, and a first similarity factor of the multi-index time sequence is calculated by adopting the following formula:
Figure BDA0002194596640000061
in the formula, F1Representing a multi-index time series first similarity factor; the larger the first similarity factor of the multi-index time series is, the more the multi-index time series data E is represented1And E2The higher the similarity.
For the characteristics of time series data in the actual monitoring process, the similarity measurement is performed on the condition that the data lengths in the multi-index time series data are the same, the data meeting the conditions can be effectively measured by determining the first similarity factor of the multi-index time series, and the first similarity factor of the multi-index time series can be used as a similarity measurement result for subsequent calculation.
Preferably, the second metric unit is configured to perform similarity measurement on the multi-index time-series data, and further includes:
for multi-index time series data E1＝{X1，X2，…，Xn1And E3＝{Z1，Z2，…，Zn2In which Xi＝(xi1，xi2，…，xik)T，Zi＝(Zi1，Zi2，…，Zik)TN1 and n2 respectively represent multi-index time-series data E1And E3And n1 ≠ n2, XiAnd ZiRespectively represent multi-index time series data E1And E3K represents multi-index time-series data E1And E3The number of the medium indexes, k is more than or equal to 2;
first, E is1And E3Is extended to a length ofw multi-index time-series data E'1＝{X1，X2，…，Xn1，Xn1+1，…，XwAnd E'3＝{Z1，Z2，…，Zn2，Zn2+1，…，ZwW > n1 and w > n2,
Figure BDA0002194596640000071
Figure BDA0002194596640000072
calculating a multi-index time series second similarity factor using the following formula:
Figure BDA0002194596640000073
in the formula, F2Representing a multi-index time series second similarity factor; the larger the second similarity factor of the multi-index time series is, the more the multi-index time series data E is represented1And E3The higher the similarity.
In the calculation process of the second similarity factor of the multi-index time series, w is greater than the maximum value of n1 and n2, so that the advantage is that the time series data with different lengths can be conveniently expanded, in order to reduce the operation amount, the value of d is usually not much greater than the maximum value of n1 and n2, and in the process of expanding the time series data, the average value of the original time series data is taken as the value of the expanded time series data, so that the distortion degree of the data can be reduced to the greatest extent, and the accuracy of the similarity measurement result can be ensured.
In the preferred embodiment, the similarity measurement of the multi-index time series data sequence with different lengths is realized by expanding the multi-index time series data with different lengths, and the second similarity factor of the multi-index time series can be used as a similarity measurement result for subsequent calculation.
Preferably, the monitoring data mining subsystem performs classification mining on the acquired time series data, specifically:
the two time series data for similarity measurement are respectively subjected to similarity measurement with known standard time series data by adopting the same similarity measurement method, the two time series data for similarity measurement can be single index time series data with the same length, also can be single index time series data with different lengths, also can be multi-index time series data with the same length and multi-index time series data with different lengths, the known standard time series data can be obtained in various ways, can be directly obtained through the Internet, also can be determined empirically or can be determined according to sample data; the standard time-series data can be understood as monitoring data when the monitoring item is normal.
Presetting a difference judgment threshold value, wherein the difference judgment threshold value is larger than zero and can be adjusted according to the actual situation, if the similarity measurement difference value between the two time data sequences and the known standard time sequence data is larger than the preset difference judgment threshold value, the difference between the two time sequence data is larger, and at the moment, setting the difference factor of the two time sequence data as P1，P1< 1, if the difference value of the similarity measure between the two time data sequences and the known standard time sequence data is smaller than the preset difference judgment threshold value, it indicates that the difference between the two time data sequences is smaller, and at this time, the difference factor of the two time data is set as P2，P2Is more than 1; the specific value of the difference factor can be determined empirically, or by determining the similarity measure between two time data sequences and known standard time sequence data and determining the similarity measure by the similarity measure, for example, P11-difference of similarity measure between two time data sequences and known standard time sequence data, P2Similarity metric difference between 1+ two time data sequences and known standard time sequence data;
the difference judgment threshold value can judge the difference of the similarity metric value between the two time-series data to be compared and the standard time-series data, thereby judging a display of the two time-series data to be comparedThe difference of the characteristics is that when the difference between the characteristics is larger, P is passed1The original similarity metric is reduced, and when the difference between the two is larger, the P is passed2Amplifying the original similarity measurement value, namely multiplying the similarity measurement result of the two time sequence data by the difference factor of the two time sequence data to obtain the improved similarity measurement result of the time sequence data; by improving the similarity measurement result, a more accurate similarity measurement result can be obtained;
and classifying the time series data by adopting a k-nearest neighbor algorithm based on the similarity measurement result of the improved time series data, and mining the data based on the classification result of the time series data. Other classification algorithms can be adopted to classify the time series data, and the time series data are classified and mined according to classification results, so that corresponding monitoring items can be conveniently monitored subsequently.
Preferably, the multi-index monitoring item is a multi-index mechanical equipment monitoring item, and the multi-index monitoring item includes mechanical equipment vibration data, mechanical equipment temperature data and mechanical equipment sound data. The vibration data, the temperature data and the sound data of the mechanical equipment are used as indexes of a multi-index mechanical equipment monitoring project, various problems of the mechanical equipment in the operation process can be accurately reflected, and the mechanical equipment can be effectively monitored.
From the above description of embodiments, it is clear for a person skilled in the art that the embodiments described herein can be implemented in hardware, software, firmware, middleware, code or any appropriate combination thereof. For a hardware implementation, a processor may be implemented in one or more of the following units: an Application Specific Integrated Circuit (ASIC), a Digital Signal Processor (DSP), a Digital Signal Processing Device (DSPD), a Programmable Logic Device (PLD), a Field Programmable Gate Array (FPGA), a processor, a controller, a microcontroller, a microprocessor, other electronic units designed to perform the functions described herein, or a combination thereof. For a software implementation, some or all of the procedures of an embodiment may be performed by a computer program instructing associated hardware. In practice, the program may be stored on or transmitted over as one or more instructions or code on a computer-readable medium. Computer-readable media includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A storage media may be any available media that can be accessed by a computer. Computer-readable media can include, but is not limited to, RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer.
Finally, it should be noted that the above embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the protection scope of the present invention, although the present invention is described in detail with reference to the preferred embodiments, it should be understood by the ordinary technical destination in the art that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims (5)

1. A monitoring system based on time series data is characterized by comprising a monitoring project dividing subsystem, a monitoring data acquisition subsystem, a monitoring data mining subsystem and a monitoring early warning subsystem, wherein the monitoring project dividing subsystem divides a monitoring project into a single-index monitoring project and a multi-index monitoring project, the monitoring data acquisition subsystem acquires time series data of indexes in the single-index monitoring project and the multi-index monitoring project, the monitoring data mining subsystem carries out classification mining on the acquired time series data, and the monitoring early warning subsystem carries out monitoring and early warning on the monitoring project based on a classification mining result of the time series data;
the monitoring data mining subsystem comprises a data dividing module, a data measuring module and a data classifying module, wherein the data dividing module divides time sequence data into single-index time sequence data and multi-index time sequence data, the data measuring module is used for carrying out similarity measurement on the single-index time sequence data and the multi-index time sequence data, and the data classifying module carries out classified mining on the single-index time sequence data and the multi-index time sequence data based on a similarity measurement result;
the data measurement module comprises a first measurement unit and a second measurement unit, wherein the first measurement unit is used for carrying out similarity measurement on single-index time series data, and the second measurement unit is used for carrying out similarity measurement on multi-index time series data;
the first measurement unit is used for carrying out similarity measurement on the single-index time series data and comprises the following steps:
for single index time series data A1＝(a1，a2，…，an) And A2＝(b1，b2，…，bn) Wherein n represents single index time-series data A1And A2The first similarity factor of the single index time series is calculated by adopting the following formula:
Figure FDA0003454157060000011
in the formula, S1Representing a first similarity factor, a, of a time series of a single indexi、biRespectively represent single index time series data A1And A2The ith time-series data of (1); the larger the first similarity factor of the single index time series is, the single index time series data A is represented1And A2The higher the similarity is;
the first measurement unit is used for carrying out similarity measurement on the single-index time series data, and further comprises:
for single index time series data A1＝(a1，a2，…，an) And A3＝(c1，c2，…，cm) Wherein n and m respectively represent single index time-series data A1And A3If the number of data in (1) is not equal to m, then A is first added1And A3Single index time series data A extended to length d1′＝(a1，a2，…，an，an+1，…，ad) And A3′＝(c1，c2，…，cm，cm+1，…，cd) Wherein d represents the extended single index time series data A1' and A3' the number of data, d > n and d > m,
Figure FDA0003454157060000012
Figure FDA0003454157060000013
ai、cirespectively represent single index time series data A1And A3The ith time-series data of (1), calculating a single index time-series second similarity factor using the following formula:
Figure FDA0003454157060000021
in the formula, S2Representing a second similarity factor of a time series of a single index, aj、cjRespectively represent the extended single index time series data A'1And A'3The jth time-series data in (a); the larger the second similarity factor of the single index time series is, the single index time series data A is represented1And A3The higher the similarity.
2. The time-series data-based monitoring system of claim 1, wherein the second metric unit is configured to perform similarity measurement on multi-index time-series data, and comprises:
for multi-index time series data E1＝{X1，X2，…，Xn1And E2＝{Y1，Y2，…，Yn1In which Xi＝(xi1，xi2，…，xik)T，Yi＝(yi1，yi2，…，yik)TI-1, 2, …, n1, n1 represent multi-index time-series data E1And E2Length of (1), XiAnd YiRespectively represent multi-index time series data E1And E2K represents multi-index time-series data E1And E2The number of the medium indexes, k is more than or equal to 2, and a first similarity factor of the multi-index time sequence is calculated by adopting the following formula:
Figure FDA0003454157060000022
in the formula, F1Representing a multi-index time series first similarity factor; the larger the first similarity factor of the multi-index time series is, the more the multi-index time series data E is represented1And E2The higher the similarity.
3. The time-series data-based monitoring system of claim 2, wherein the second metric unit is configured to perform a similarity metric on multi-index time-series data, and further comprising:
for multi-index time series data E1＝{X1，X2，…，Xn1And E3＝{Z1，Z2，…，Zn2In which Xi＝(xi1，xi2，…，xik)T，Zi＝(zi1，zi2，…，zik)TN1 and n2 respectively represent multi-index time-series data E1And E3And n1 ≠ n2, XiAnd ZiRespectively represent multi-index time series data E1And E3K represents multi-index time-series data E1And E3The number of the medium indexes, k is more than or equal to 2;
first, E is1And E3Extended to multi-index time series data E 'with length w'1＝{X1，X2，…，Xn1，Xn1+1，…，XwAnd E'3＝{Z1，Z2，…，Zn2，Zn2+1，…，ZwW > n1 and w > n2,
Figure FDA0003454157060000023
Figure FDA0003454157060000024
calculating a multi-index time series second similarity factor using the following formula:
Figure FDA0003454157060000025
in the formula, F2Representing a multi-index time series second similarity factor; the larger the second similarity factor of the multi-index time series is, the more the multi-index time series data E is represented1And E3The higher the similarity.
4. The time-series data-based monitoring system of claim 3, wherein the monitoring data mining subsystem performs classification mining on the collected time-series data, specifically:
respectively carrying out similarity measurement on the two time sequence data subjected to similarity measurement and the known standard time sequence data by adopting the same similarity measurement method, presetting a difference judgment threshold value, and if the difference value of the similarity measurement between the two time sequence data and the known standard time sequence data is greater than the preset difference judgment threshold value, indicating that the difference between the two time sequence data is greater, and at the moment, setting the difference factor of the two time sequence data as P1，P1< 1 > if the difference between the similarity measures of the two time data sequences and the known standard time sequence data is less than the predetermined difference judgment thresholdThe difference factor of the two time series data is set as P2，P2＞1；
Multiplying the similarity measurement result of the two time sequence data by the difference factor of the two time sequence data to obtain an improved similarity measurement result of the time sequence data;
and classifying the time series data by adopting a k-nearest neighbor algorithm based on the similarity measurement result of the improved time series data, and mining the data based on the classification result of the time series data.
5. The time-series data-based monitoring system of claim 1, wherein the multi-index monitoring item is a multi-index mechanical equipment monitoring item, and the multi-index comprises mechanical equipment vibration data, mechanical equipment temperature data, and mechanical equipment sound data.
CN201910844017.8A 2019-09-06 2019-09-06 Monitoring system based on time series data Active CN110543505B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910844017.8A CN110543505B (en) 2019-09-06 2019-09-06 Monitoring system based on time series data

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910844017.8A CN110543505B (en) 2019-09-06 2019-09-06 Monitoring system based on time series data

Publications (2)

Publication Number Publication Date
CN110543505A CN110543505A (en) 2019-12-06
CN110543505B true CN110543505B (en) 2022-02-18

Family

ID=68712893

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910844017.8A Active CN110543505B (en) 2019-09-06 2019-09-06 Monitoring system based on time series data

Country Status (1)

Country Link
CN (1) CN110543505B (en)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5732184A (en) * 1995-10-20 1998-03-24 Digital Processing Systems, Inc. Video and audio cursor video editing system
CN103872782A (en) * 2014-03-31 2014-06-18 国家电网公司 Electric energy quality data comprehensive service system
CN104281891A (en) * 2014-10-13 2015-01-14 安徽华贞信息科技有限公司 Time-series data mining method and system
CN104281130A (en) * 2014-09-22 2015-01-14 国家电网公司 Hydroelectric equipment monitoring and fault diagnosis system based on big data technology
CN107203686A (en) * 2017-03-31 2017-09-26 苏州艾隆信息技术有限公司 medicine information difference processing method and system

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2172820A1 (en) * 2008-10-06 2010-04-07 Basf Se Method and system for automated analysis of process data
CN105719421B (en) * 2016-04-27 2017-11-28 丛静华 A kind of integrated forest fire protection information system excavated based on big data

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5732184A (en) * 1995-10-20 1998-03-24 Digital Processing Systems, Inc. Video and audio cursor video editing system
CN103872782A (en) * 2014-03-31 2014-06-18 国家电网公司 Electric energy quality data comprehensive service system
CN104281130A (en) * 2014-09-22 2015-01-14 国家电网公司 Hydroelectric equipment monitoring and fault diagnosis system based on big data technology
CN104281891A (en) * 2014-10-13 2015-01-14 安徽华贞信息科技有限公司 Time-series data mining method and system
CN107203686A (en) * 2017-03-31 2017-09-26 苏州艾隆信息技术有限公司 medicine information difference processing method and system

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
Detecting financial early warning signs in Istanbul Stock Exchange by data mining;Koyuncugil A S;《International Journal of Business Research》;20071231;第7卷(第3期);全文 *
基于大数据的输电线路故障预警模型设计;郑茂然;《南方电网技术》;20171231;第11卷(第04期);第30-37页 *

Also Published As

Publication number Publication date
CN110543505A (en) 2019-12-06

Similar Documents

Publication Publication Date Title
CN108392211B (en) Fatigue detection method based on multi-information fusion
US20140149325A1 (en) System monitor and method of system monitoring
US7818131B2 (en) Non-parametric modeling apparatus and method for classification, especially of activity state
US10332638B2 (en) Methods and systems for pre-symptomatic detection of exposure to an agent
JP6494775B2 (en) Data analysis processing method and elasticity detection apparatus for elasticity detection apparatus
CN109564586B (en) System monitor and system monitoring method
CN111919242A (en) System and method for processing multiple signals
CN111370124A (en) Health analysis system and method based on facial recognition and big data
CN113096792B (en) Intelligent health monitoring and early warning method and system based on machine learning
CN110543505B (en) Monitoring system based on time series data
JP2012505456A (en) Improvements in multi-parameter monitoring or improvements related to multi-parameter monitoring
CN104736066A (en) Biometric Sound Testing Device And Biometric Sound Testing Method
US11378944B2 (en) System analysis method, system analysis apparatus, and program
CN116106672A (en) Vehicle network resonance detection method and device based on data driving and engineering knowledge
CN113974566B (en) COPD acute exacerbation prediction method based on time window
CN113273992B (en) Signal processing method and device
Silva et al. Compression of Different Time Series Representations in Asphyxia Detection
US20230074474A1 (en) Parameter adjustment apparatus, inference apparatus, parameter adjustment method, and computer-readable storage medium storing a parameter adjustment program
CN115526882A (en) Medical image classification method, device, equipment and storage medium
CN111436939B (en) Method, system, device and medium for identifying sign signals based on deep learning
Maidel et al. A prediction model of in-patient deteriorations based on passive vital signs monitoring technology
CN117235650B (en) Method, device, equipment and medium for detecting high-altitude operation state
US20140280273A1 (en) Animal abnormality detecting device, and animal abnormality detecting method
CN117598700B (en) Intelligent blood oxygen saturation detection system and method
Khan et al. Automated classification of human lung sound signals using phase space representation of intrinsic mode function

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant