WO2019063812A1 - METHOD AND DEVICE FOR DETECTING ANOMALIES OF DISCREET PRODUCTION EQUIPMENT - Google Patents
METHOD AND DEVICE FOR DETECTING ANOMALIES OF DISCREET PRODUCTION EQUIPMENT Download PDFInfo
- Publication number
- WO2019063812A1 WO2019063812A1 PCT/EP2018/076507 EP2018076507W WO2019063812A1 WO 2019063812 A1 WO2019063812 A1 WO 2019063812A1 EP 2018076507 W EP2018076507 W EP 2018076507W WO 2019063812 A1 WO2019063812 A1 WO 2019063812A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- variable
- characteristic value
- type
- production equipment
- discrete
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 105
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 103
- 230000005856 abnormality Effects 0.000 title claims abstract description 90
- 230000002159 abnormal effect Effects 0.000 claims abstract description 23
- 238000005070 sampling Methods 0.000 claims abstract description 17
- 125000004122 cyclic group Chemical group 0.000 claims description 23
- 230000009467 reduction Effects 0.000 claims description 19
- 238000000605 extraction Methods 0.000 claims description 8
- 230000003044 adaptive effect Effects 0.000 claims description 7
- 238000000354 decomposition reaction Methods 0.000 claims description 7
- 230000001052 transient effect Effects 0.000 claims description 3
- 238000004886 process control Methods 0.000 claims 1
- 238000001514 detection method Methods 0.000 description 29
- 230000008569 process Effects 0.000 description 28
- 238000010586 diagram Methods 0.000 description 7
- 230000006870 function Effects 0.000 description 4
- 238000007781 pre-processing Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000004891 communication Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000004806 packaging method and process Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 239000000047 product Substances 0.000 description 1
- 239000013589 supplement Substances 0.000 description 1
- 230000017105 transposition Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/0703—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation
- G06F11/0751—Error or fault detection not based on redundancy
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/418—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM]
- G05B19/4184—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM] characterised by fault tolerance, reliability of production system
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
- G05B23/0205—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults
- G05B23/0218—Electric testing or monitoring by means of a monitoring system capable of detecting and responding to faults characterised by the fault detection method dealing with either existing or incipient faults
- G05B23/0224—Process history based detection method, e.g. whereby history implies the availability of large amounts of data
- G05B23/024—Quantitative history assessment, e.g. mathematical relationships between available data; Functions therefor; Principal component analysis [PCA]; Partial least square [PLS]; Statistical classifiers, e.g. Bayesian networks, linear regression or correlation analysis; Neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F11/00—Error detection; Error correction; Monitoring
- G06F11/07—Responding to the occurrence of a fault, e.g. fault tolerance
- G06F11/0703—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation
- G06F11/0706—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation the processing taking place on a specific hardware platform or in a specific software environment
- G06F11/0736—Error or fault processing not based on redundancy, i.e. by taking additional measures to deal with the error or fault not making use of redundancy in operation, in hardware, or in data representation the processing taking place on a specific hardware platform or in a specific software environment in functional embedded systems, i.e. in a data processing system designed as a combination of hardware and software dedicated to performing a certain function
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2413—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on distances to training or reference patterns
- G06F18/24133—Distances to prototypes
- G06F18/24137—Distances to cluster centroïds
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/2433—Single-class perspective, e.g. one-against-all classification; Novelty detection; Outlier detection
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/80—Management or planning
Definitions
- the present application generally relates to the field of abnormality detection, and in particular relates to a method and device for detecting abnormalities of discrete production equipment .
- OPC-UA Object Linking and Embedding (OLE) for Process Control-Unified Architecture
- HET Industry Internet of things
- the amount of variable data obtained by use of the OPC-UA technique is so large that a large amount of communication overheads will be consumed.
- a lot of variable data of the accessed variable data is useless for equipment abnormality detection and thus lowers the detection efficiency.
- the prior equipment abnormality detection solutions are generally customized for specific fields (for example, jet engines) and these detection solutions are generally not applicable to other fields.
- profound know-how is required to analyze and determine specific variables in the customized detection solutions.
- an alarm is generally based on a threshold during an abnormality detection, and in this situation, it is impossible to use historical data to improve the detection accuracy .
- the present application provides a method and device for detecting abnormalities of discrete production equipment.
- the VOI space of discrete production equipment, the characteristic value of the variable type and the variable abnormality related characteristic value are obtained, the time sequence signal of each variable in said VOI space is sampled and the corresponding characteristic value is extracted, the type of the variable is determined based on the extracted characteristic value, and an equipment abnormality determination is made based on the extracted characteristic value of each variable and the variable abnormality related characteristic value of the corresponding type during a detection.
- equipment abnormalities can efficiently be detected.
- a method for detecting abnormalities of discrete production equipment comprises: for each variable in the VOI space of said discrete production equipment, sampling the value of the variable within the production period of said production equipment to obtain the time sequence signal of the variable, extracting the characteristic value of the variable from the sampled time sequence signal of the variable, determining the type of the variable based on the extracted characteristic value of the variable and the characteristic value of the variable type, and determining whether said discrete production equipment is abnormal based on the extracted characteristic value of each variable and the variable abnormality related characteristic value of the determined type of the variable.
- said VOI space is obtained by acquiring random cyclic variables from the variables which are obtained during normal operation of said discrete production equipment.
- extracting the characteristic value of the variable from the sampled time sequence signal of the variable can comprise: extracting the characteristic value of the variable by applying a dimensionality reduction technique to the sampled time sequence signal of the variable .
- said dimensionality reduction technique can include one of the following techniques: discrete Fourier transform, discrete wavelet transform, singular value decomposition, piecewise linear approximation, adaptive piecewise constant approximation, piecewise aggregate
- said characteristic value of the variable type and said variable abnormality related characteristic value can be obtained by use of the following method: use the K-means algorithm to perform clustering operations for the extracted characteristic value of said variable from the sampled time sequence signal of said variable during normal operation of said discrete production equipment to form a plurality of types, for each type, calculate the mean and standard deviation of the characteristic values of the type, and use the calculated mean as the characteristic value of the variable type and the calculated standard deviation as the variable abnormality related characteristic value of the type.
- determining whether said discrete production equipment is abnormal based on the extracted characteristic value of each variable and the variable abnormality related characteristic value of the determined type of the variable can comprise: calculating the Euclidean distance of the variable relative to said characteristic value of the variable type based on the extracted characteristic value of the variable and the characteristic value of the variable type, and determining said discrete production equipment is abnormal when the calculated Euclidean distance is greater than a preset multiple of the variable abnormality related characteristic value of the type of the variable.
- the time sequence signal of said variable is obtained by an edge device by use of the OPC-UA technique.
- the time sequence signal of said variable is concurrently obtained by a plurality of edge devices by use of the OPC-UA technique.
- a device for detecting abnormalities of discrete production equipment comprises: a sampling unit, used to sample the values of each variable in the VOI space of said discrete production equipment within the production period of said production equipment to obtain the time sequence signal of the variable, an extraction unit, used to extract the characteristic value of the variable from the sampled time sequence signal of the variable, a type determining unit, used to determine the type of the variable based on the extracted characteristic value of the variable and the characteristic value of the variable type, and an abnormality determining unit, used to determine whether said discrete production equipment is abnormal based on the extracted characteristic value of each variable and the variable abnormality related characteristic value of the determined type of the variable.
- said device can further comprise a VOI space acquiring unit, used to obtain the VOI space by acquiring random cyclic variables from the variables which are obtained during normal operation of said discrete production equipment.
- said extraction unit is used to extract the characteristic value of the variable by applying a dimensionality reduction technique to the sampled time sequence signal of the variable.
- said dimensionality reduction technique can include one of the following techniques: discrete Fourier transform, discrete wavelet transform, singular value decomposition, piecewise linear approximation, adaptive piecewise constant approximation, piecewise aggregate
- said device can further comprise a clustering unit, used to use the K-means algorithm to perform clustering operations for the extracted characteristic value of said variable from the sampled time sequence signal of said variable during normal operation of said discrete production equipment to form a plurality of types, and a calculating unit, used to calculate the mean and standard deviation of the characteristic values of each type and use the calculated mean as the characteristic value of the variable type and the calculated standard deviation as the variable abnormality related characteristic value of the type of the variable.
- a clustering unit used to use the K-means algorithm to perform clustering operations for the extracted characteristic value of said variable from the sampled time sequence signal of said variable during normal operation of said discrete production equipment to form a plurality of types
- a calculating unit used to calculate the mean and standard deviation of the characteristic values of each type and use the calculated mean as the characteristic value of the variable type and the calculated standard deviation as the variable abnormality related characteristic value of the type of the variable.
- said abnormality determining unit can further be used to calculate the Euclidean distance of the variable relative to said characteristic value of the variable type based on the extracted characteristic value of the variable and the characteristic value of the variable type, and determine said discrete production equipment is abnormal when the calculated Euclidean distance is greater than a preset multiple of the variable abnormality related characteristic value of the type of the variable.
- said device can further comprise a storage unit, used to store variable information in said VOI space, characteristic values of variable types and variable abnormality related characteristic values.
- a computing device comprises one or more processors, and a memory used to store instructions.
- said instructions When said instructions are executed by said one or more processors, said one or more processors will execute the above-mentioned method for detecting discrete production equipment.
- a non-transient machine readable storage medium stores executable instructions. When said instructions are executed, said machine will execute the above-mentioned method for detecting discrete production equipment.
- the VOI space of discrete production equipment, the characteristic value of the variable type and the variable abnormality related characteristic value are obtained in advance, and the time sequence signal of each variable in said VOI space is sampled and the corresponding characteristic value is extracted, the type of the variable is determined based on the extracted characteristic value, and an equipment abnormality determination is made based on the extracted characteristic value of each variable and the variable abnormality related characteristic value of the corresponding type during a detection.
- equipment abnormalities can efficiently be detected.
- the number of variables to be processed can be reduced by acquiring random cyclic variables from the variables which are obtained during normal operation of said discrete production equipment to obtain the VOI space, and thus the processing efficiency is improved and the occupied storage space is reduced.
- a dimensionality reduction technique is applicable to a wider range of signals, for example, burst signals, by using the
- Figure 1 shows the flowchart of the process used to obtain the VOI space in the present application.
- Figure 2 shows the flowchart of an example of the process used to remove non-cyclic variables in the present application.
- Figure 3 shows the flowchart of an example of the process used to remove fixed cyclic variables in the present application.
- Figure 4 shows the flowchart of an example of the process used to remove derived variables in the present application.
- Figure 5 shows the flowchart of an example of the process used to obtain the characteristic value and the variable abnormality related characteristic value of a variable type in the present application .
- Figure 6 shows the flowchart of the method for detecting abnormalities of discrete production equipment in the present application .
- Figure 7 is a block diagram for an example of the device for detecting abnormalities of discrete production equipment in the present application.
- Figure 8 is a block diagram for another example of the device for detecting abnormalities of discrete production equipment in the present application.
- Figure 9 is a block diagram for a further example of the device for detecting abnormalities of discrete production equipment in the present application.
- the term “comprise” and its variants are open terms and mean “include but are not limited to” .
- the term “based on” means “at least partially based on”.
- the term “one embodiment” means “at least one embodiment”.
- the term “another embodiment” means “at least one other embodiment”.
- the term “first” and “second” can refer to different identical objects. Other definitions, explicit or implicit, can be included below. Unless otherwise specified in the context, the definition of a term is consistent in the whole description.
- discrete production equipment in this document refers to the production equipment having the following characteristics : (1) the production equipment has a production period (that is to say, the production equipment operates cyclically) , in other words, the production equipment repeats specific operations all the time and has an obvious cycle start signal and an obvious cycle end signal; (2) the production equipment involves limited processes and produces limited types of products. Said discrete production equipment can include a packaging machine, for example .
- OPC-UA is a new generation technology provided based on the OPC Foundation. Through the OPC-UA, all required information can be accessed by each authorized person for each authorized application at any place at any time. This function is independent of the original application, programming language and operating system of the manufacturer.
- the OPC-UA is a supplement to the OPC industry standard in use and provides some important features, including independence, scalability, high-reliability and Internet accessibility of the platform.
- the OPC-UA independent of DCOM, is a service-oriented architecture (SOA) .
- SOA service-oriented architecture
- the technique permits a single OPC-UA server to consolidate data, alarms and events, and historical information into its address space and use a unified service to externally provide an interface for them.
- variable B refers to the variables derived from other variables. For example, for variable A and variable B, if variable B can be expressed based on the expression of variable A or variable B can be derived from variable A, then variable B is a derived variable of variable A.
- characteristic value of variable type is a characteristic value used to represent the characteristic of the type of a variable, for example, the mean of the time sequence signals of the variables included in the type.
- variable abnormality related characteristic value refers to a characteristic value related to the abnormal state of a variable, for example, the standard deviation of the time sequence signals of the variables included in the type.
- detection device Said device for detecting abnormalities of discrete production equipment is called detection device hereinafter.
- Said preprocessing process includes the acquisition of the VOI space and the acquisitions of the characteristic value of a variable type and the variable abnormality related characteristic value. Said preprocessing process is performed online or offline.
- Figure 1 shows the flowchart of the process used to obtain the VOI space in the present application.
- removing writable variables can comprise first determining whether a writable variable exists in the obtained variable set. For example, determine whether a variable is a writable variable by reading the read and write attribute information of the variable . If the variable is a writable variable, remove it from the obtained variable set. Then, the process goes to block 120.
- removing constant variables can comprise determining whether a constant variable exists in the first remaining variable set and if a constant variable exists, removing said constant variable.
- FIG. 130 shows the flowchart of an example of the process used to remove non-cyclic variables in the present application.
- Euclidean distance D mn ( ⁇ (
- a preset threshold for example, 0.1
- FIG. 140 shows the flowchart of an example of the process used to remove fixed cyclic variables in the present application.
- FIG. 150 shows the flowchart of an example of the process used to remove derived variables in the present application.
- Figure 1 shows only an example of the method for obtaining the VOI space. In other examples, the sequence of the steps of the method shown in Figure 1 can be changed .
- Figure 5 shows the flowchart of an example of the process used to obtain the characteristic value and the variable abnormality related characteristic value of a variable type in the present application .
- extracting the characteristic value of the variable from the sampled time sequence signal of the variable can comprise: extracting the characteristic value of the variable by applying a dimensionality reduction technique to the sampled time sequence signal of the variable.
- said dimensionality reduction technique can include one of the following techniques: discrete Fourier transform, discrete wavelet transform, singular value decomposition, piecewise linear approximation, adaptive piecewise constant approximation, piecewise aggregate approximation, Chebyshev polynomials, and symbolic approximation.
- the K-means algorithm After extracting the characteristic value of the variable, in block 230, use the K-means algorithm to perform clustering operations for the extracted characteristic value, for example, ⁇ dxii, dx ⁇ 2 ... dx ⁇ N ⁇ , to form a plurality of types. Then, in block 240, for each type, calculate the mean and standard deviation of the characteristic values of the type, and use the calculated mean as the characteristic value of the variable type and the calculated standard deviation as the variable abnormality related characteristic value of the type.
- the above-mentioned process can further comprise the operation in block 250.
- block 250 store the calculated characteristic value of the variable type and variable abnormality related characteristic value in the detection device.
- FIG. 6 shows the flowchart of the method for detecting abnormalities of discrete production equipment in the present application .
- extracting the characteristic value of the variable from the sampled time sequence signal of the variable can comprise: extracting the characteristic value of the variable by applying a dimensionality reduction technique to the sampled time sequence signal of the variable .
- said dimensionality reduction technique can include one of the following techniques: discrete Fourier transform, discrete wavelet transform, singular value decomposition, piecewise linear approximation, adaptive piecewise constant approximation, piecewise aggregate
- said dimensionality reduction technique is discrete wavelet transform.
- determining the type of a variable based on the extracted characteristic value of the variable and characteristic value of the variable type can comprise : calculating the Euclidean distance of the characteristic value of the variable relative to each stored characteristic value (for example, mean) of the variable type and determining the type of the characteristic value of the variable type corresponding to the calculated Euclidean distance less than a preset threshold to be the type of the variable
- the Euclidean distances calculated relative to a plurality of means are all less than a preset threshold, the variable is considered as an undistinguishable variable and needs to be removed from said VOI space.
- determining whether said discrete production equipment is abnormal can comprise: calculating the Euclidean distance of the extracted characteristic value of the variable relative to the characteristic value (for example, mean) of the variable type, and then comparing the calculated Euclidean distance with the variable abnormality related characteristic value (for example, standard deviation) of the type of the variable to determine whether discrete production equipment is abnormal. For example, if the calculated Euclidean distance is greater than a preset multiple (for example, the multiple 3) of the variable abnormality related characteristic value, the discrete production equipment is considered abnormal. Otherwise, the discrete production equipment is not considered abnormal. If it is determined that an abnormality occurs in block 340, the process goes to block 360 and the process ends. Preferably, the result that the discrete production equipment is abnormal can also be returned to the user in this situation. If it is determined that no abnormality occurs to the variable in block 340, then, the process goes to block 350.
- block 350 determine whether there are any variables that are not detected in said VOI space. If there are variables that are not detected, select one variable that is not detected and go to block 310 to execute the above-mentioned detection process for the variable. If there are no variables that are not detected, the process goes to block 360 and the process ends. Preferably, the result that the discrete production equipment is normal can also be returned to the user in this situation.
- time sequence signals of variables are obtained by use of the OPC-UA technique in the description above.
- the time sequence signals of variables can be obtained by use of other suitable techniques.
- the time sequence signals of variables can be obtained by one edge device, or can concurrently be obtained by a plurality of edge devices.
- FIG. 7 is a block diagram for an example of the device (referred to as detection device 700 hereinafter) for detecting
- the detection device 700 comprises a sampling unit 710, an extraction unit 720, a type determining unit 730, and an abnormality determining unit 740.
- the sampling unit 710 is used to sample the values of each variable in the VOI space of said discrete production equipment within the production period of said discrete production equipment to obtain the time sequence signal of the variable.
- the extraction unit 720 is used to extract the characteristic value of the variable from the sampled time sequence signal of the variable.
- the extraction unit 720 can extract the characteristic value of the variable by applying a dimensionality reduction technique to the sampled time sequence signal of the variable .
- Said dimensionality reduction technique can include one of the following techniques : discrete Fourier transform, discrete wavelet transform, singular value decomposition, piecewise linear approximation, adaptive piecewise constant approximation, piecewise aggregate approximation, Chebyshev polynomials, and symbolic approximation.
- said dimensionality reduction technique is discrete wavelet transform.
- the type determining unit 730 is used to determine the type of the variable based on the extracted characteristic value of the variable and the characteristic value of the variable type. For example, when the characteristic value of the variable type is the mean of the time sequence signals, the type determining unit 730 can be used to calculate the Euclidean distance of the characteristic value of the variable relative to each stored characteristic value (for example, each mean) of the variable type and determine the type of the characteristic value of the variable type corresponding to the calculated Euclidean distance less than a preset threshold to be the type of the variable.
- the abnormality determining unit 740 is used to determine whether said discrete production equipment is abnormal based on the extracted characteristic value of each variable and the variable abnormality related characteristic value of the determined type of the variable.
- the abnormality determining unit 740 can be used to calculate the Euclidean distance of the variable relative to said characteristic value of the variable type based on the extracted characteristic value of the variable and the characteristic value of the variable type, and determine said discrete production equipment is abnormal when the calculated Euclidean distance is greater than a preset multiple of the variable abnormality related characteristic value of the type of the variable.
- FIG 8 is a block diagram for another example of the device (referred to as detection device 800 hereinafter) for detecting abnormalities of discrete production equipment in the present application.
- the detection device 800 shown in Figure 8 is a modification to the detection device 700 shown in Figure 7.
- the detection device 800 further comprises a VOI space acquiring unit 703, a clustering unit 705 and a calculating unit 707.
- the VOI acquiring unit 703 is used to obtain the VOI space by acquiring random cyclic variables from the variables which are obtained during normal operation of said discrete production equipment.
- the VOI space is acquired in advance by removing one or more of the following variables from the variables obtained during the normal operation of said discrete production equipment: writable variable, constant variable, non-cyclic variable, fixed cyclic variable and derived variable.
- the clustering unit 705 is used to use the K-means algorithm to perform clustering operations for the extracted characteristic value of said variable from the sampled time sequence signal of said variable during the normal operation of said discrete production equipment to form a plurality of types. Then, the calculating unit 707 calculates the mean and standard deviation of the characteristic values of the type, and uses the calculated mean as the characteristic value of the variable type and the calculated standard deviation as the variable abnormality related characteristic value of the type.
- the detection device 800 further comprises a storage unit 709, which is used to store variable information in said VOI space, characteristic values of variable types and variable abnormality related characteristic values.
- the device for detecting abnormalities of discrete production equipment in the present application can comprise one or more of the VOI space acquiring unit 703, the clustering unit 705, the calculating unit 707 and the storage unit 709.
- Figure 9 is a block diagram for a further example of the device (referred to as detection device 900 hereinafter) for detecting abnormalities of discrete production equipment in the present application.
- the detection device 900 comprises one or more processors 910 and a memory 920.
- Computer executable instructions are stored in the memory 920.
- processors 910 are used to sample the value of each variable in the VOI space of said discrete production equipment within the production period of said discrete production equipment to obtain the time sequence signal of the variable, extract the
- characteristic value of the variable from the sampled time sequence signal of the variable, determine the type of the variable based on the extracted characteristic value of the variable and the characteristic value of the variable type, and determine whether said discrete production equipment is abnormal based on the extracted characteristic value of each variable and the variable abnormality related characteristic value of the determined type of the variable.
- processors 910 when the computer executable instructions stored in the memory 920 are executed, one or more processors 910 will execute various operations and functions described in the embodiments of the present application.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Life Sciences & Earth Sciences (AREA)
- Quality & Reliability (AREA)
- Automation & Control Theory (AREA)
- Probability & Statistics with Applications (AREA)
- Manufacturing & Machinery (AREA)
- Mathematical Physics (AREA)
- General Factory Administration (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Testing And Monitoring For Control Systems (AREA)
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710911412.4A CN109582482A (zh) | 2017-09-29 | 2017-09-29 | 用于检测离散型生产设备的异常的方法及装置 |
CN201710911412.4 | 2017-09-29 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2019063812A1 true WO2019063812A1 (en) | 2019-04-04 |
Family
ID=63762495
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/EP2018/076507 WO2019063812A1 (en) | 2017-09-29 | 2018-09-28 | METHOD AND DEVICE FOR DETECTING ANOMALIES OF DISCREET PRODUCTION EQUIPMENT |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN109582482A (zh) |
WO (1) | WO2019063812A1 (zh) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117149500A (zh) * | 2023-10-30 | 2023-12-01 | 安徽思高智能科技有限公司 | 基于指标数据和日志数据的异常根因获得方法及系统 |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140358465A1 (en) * | 2013-05-31 | 2014-12-04 | Samsung Sds Co., Ltd. | Yield analysis system and method using sensor data of fabrication equipment |
WO2017086963A1 (en) * | 2015-11-19 | 2017-05-26 | Siemens Aktiengesellschaft | Anomaly detection in multiple correlated sensors |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP5501903B2 (ja) * | 2010-09-07 | 2014-05-28 | 株式会社日立製作所 | 異常検知方法及びそのシステム |
US9984334B2 (en) * | 2014-06-16 | 2018-05-29 | Mitsubishi Electric Research Laboratories, Inc. | Method for anomaly detection in time series data based on spectral partitioning |
CN105205113A (zh) * | 2015-09-01 | 2015-12-30 | 西安交通大学 | 一种时序数据异常变化过程的挖掘系统及方法 |
CN105205112A (zh) * | 2015-09-01 | 2015-12-30 | 西安交通大学 | 一种时序数据异常特征的挖掘系统及方法 |
-
2017
- 2017-09-29 CN CN201710911412.4A patent/CN109582482A/zh active Pending
-
2018
- 2018-09-28 WO PCT/EP2018/076507 patent/WO2019063812A1/en active Application Filing
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20140358465A1 (en) * | 2013-05-31 | 2014-12-04 | Samsung Sds Co., Ltd. | Yield analysis system and method using sensor data of fabrication equipment |
WO2017086963A1 (en) * | 2015-11-19 | 2017-05-26 | Siemens Aktiengesellschaft | Anomaly detection in multiple correlated sensors |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117149500A (zh) * | 2023-10-30 | 2023-12-01 | 安徽思高智能科技有限公司 | 基于指标数据和日志数据的异常根因获得方法及系统 |
CN117149500B (zh) * | 2023-10-30 | 2024-01-26 | 安徽思高智能科技有限公司 | 基于指标数据和日志数据的异常根因获得方法及系统 |
Also Published As
Publication number | Publication date |
---|---|
CN109582482A (zh) | 2019-04-05 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Wen et al. | RobustSTL: A robust seasonal-trend decomposition algorithm for long time series | |
AU2019201857B2 (en) | Sparse neural network based anomaly detection in multi-dimensional time series | |
CN106104496B (zh) | 用于任意时序的不受监督的异常检测 | |
US11165799B2 (en) | Anomaly detection and processing for seasonal data | |
Nouhi et al. | Prediction of surface roughness of various machining processes by a hybrid algorithm including time series analysis, wavelet transform and multi view embedding | |
US11435736B2 (en) | Cause determination of anomalous events | |
JP2019500680A5 (zh) | ||
WO2013188326A1 (en) | Discriminative hidden kalman filters for classification of streaming sensor data in condition monitoring | |
CN111144548A (zh) | 抽油机井工况的识别方法及装置 | |
WO2020246944A1 (en) | Method and system for attack detection in a sensor network of a networked control system | |
Borjas et al. | Subspace identification for industrial processes | |
JP2014041565A (ja) | 時系列データ解析装置、方法、及びプログラム | |
Liao et al. | Anomaly detection for semiconductor tools using stacked autoencoder learning | |
CN112583944B (zh) | 一种更新域名证书的处理方法和装置 | |
Das et al. | Data mining and control loop performance assessment: The multivariate case | |
EP4273564A1 (en) | Apparatus, system and method for detecting anomalies in a grid | |
US20150332152A1 (en) | Systems And Methods For Generating Models For Physical Systems Using Sentences In A Formal Grammar | |
CN111488947A (zh) | 电力系统设备的故障检测方法与装置 | |
WO2019063812A1 (en) | METHOD AND DEVICE FOR DETECTING ANOMALIES OF DISCREET PRODUCTION EQUIPMENT | |
Ma et al. | Fault prognosis technology for non‐Gaussian and nonlinear processes based on KICA reconstruction | |
Hua et al. | Mechanical fault diagnosis using color image recognition of vibration spectrogram based on quaternion invariable moment | |
CN108304778A (zh) | 一种基于压缩域的振动信号特征提取方法 | |
US8620987B2 (en) | Multi-regime detection in streaming data | |
CN111897851A (zh) | 异常数据的确定方法、装置、电子设备及可读存储介质 | |
US20200372385A1 (en) | Compensating for out-of-phase seasonality modes in time-series signals to facilitate prognostic-surveillance operations |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 18782375 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 18782375 Country of ref document: EP Kind code of ref document: A1 |