KR102236802B1 - Device and method for feature extraction of data for diagnostic models - Google Patents

Abstract

The present invention relates to an apparatus for extracting characteristics of data for a diagnosis model and a method thereof. The apparatus includes: a machine tool operation data extraction part extracting machine tool operation data created based on vibrations generated when a machine tool is operated, and including normal data and abnormal data; a machine tool operation data classification part creating classification data by classifying the machine tool operation data through a plurality of classification standards; a plot area determination part forming plot lines by connecting each individual datum included in the classification data through a plot and determining the inside of the outermost plot line as a plot area; and a classification standard determination part determining classification standards satisfying a similarity value, which is calculated by performing overlapping between plot areas with respect to each of the normal data and abnormal data, at no more than a specific standard. Therefore, the present invention is capable of determining the most suitable classification standard for the creation of data.

Description

Device and method for extracting features of data for diagnostic models {DEVICE AND METHOD FOR FEATURE EXTRACTION OF DATA FOR DIAGNOSTIC MODELS}

The present invention relates to a technology for extracting features of diagnostic model data, and more particularly, an apparatus and method for extracting features of diagnostic model data capable of extracting more efficient classification criteria in the process of generating diagnostic model data. It is about.

In the development of the existing learning-based diagnosis model, the feature data extracted from the data was reduced by using a Principal Component Analysis (PCA) technique to reduce the learning time and the number of data. However, according to this dimensionality reduction technique, the capacity of data could not be reduced in a way that all data should be used.

Machine learning is an algorithm technology that classifies/learns the features of input data by itself, and element technology is a technology that simulates functions such as cognition and judgment of the human brain using machine learning algorithms. It consists of technical fields such as prediction, knowledge expression, and motion control.

Motion control is a technology that controls autonomous driving of a vehicle and movement of a robot, and includes movement control (navigation, collision, travel), operation control (behavior control), and the like.

Korean Patent Registration No. 10-1068319 (2011.09.21) is for use in a position control apparatus including an iterative learning circuit, an exposure apparatus, a device manufacturing method, and a position control apparatus having an iterative learning circuit including a learning filter. It relates to an iterative learning method, and more specifically, a detection unit configured to detect a position of a control target, a subtraction unit configured to subtract the output of the detection unit from a target value, and a deviation between the output of the detection unit and the target value is input And an iterative learning control circuit for feeding forward a control input to the control object, and a parameter calculation unit configured to calculate a parameter variation of the control object.

Korean Patent Laid-Open Publication No. 10-2013-0065624 (2013.06.19) relates to an abnormality detection device of a periodic drive system, a processing device having a periodic drive system, an abnormality detection method of a periodic drive system, and a computer program. A detection means for detecting sound from a periodic drive system used for processing in, data acquisition means for acquiring time series data that fluctuates with time from the detected sound, and time series data from the time series data acquired by the data acquisition means A deterministic value calculation means for calculating a plurality of values representing determinism as an indicator of whether determinism or probabilistic values or intermediate changes in the process of calculating the values at each predetermined time, and calculation of a value representing the plurality of calculated determinism or the value thereof Probability distribution calculation means for calculating a probability distribution of an intermediate change amount in a process, and a determination means for determining whether or not the periodic drive system is abnormal based on the value representing the determinism or the probability distribution of the intermediate change amount in the calculation process of the value Includes.

Korean Patent Registration No. 10-1068319 (2011.09.21) Korean Patent Publication No. 10-2013-0065624 (2013.06.19)

An embodiment of the present invention is to provide an apparatus and method for extracting features of diagnostic model data capable of determining an appropriate classification criterion for a diagnostic model by dividing normal data and abnormal data to extract classification data when determining a classification criterion.

An embodiment of the present invention is to provide an apparatus and method for extracting features of diagnostic model data capable of generating classification data by displaying machine tool operation data in a Cartesian coordinate system through a plurality of classification criteria in the process of determining the classification criteria. do.

An embodiment of the present invention is to provide an apparatus and method for extracting features of diagnostic model data capable of determining a classification criterion most suitable for generating training data based on an area of a plot area.

Among the embodiments, the device for extracting features of the diagnostic model data includes a machine tool operation data extracting unit that is generated based on vibrations generated when the machine tool is operated and extracts machine tool operation data including normal data and abnormal data, A machine tool operation data classification unit that classifies the machine tool operation data through a plurality of classification criteria to generate classification data, and forms a plot line by connecting each individual data included in the classification data with a plot. A plot area determination unit that determines the inside of the formed plot line as a plot area and a classification criterion that satisfies the calculated similarity value below a specific criterion by performing overlapping between the plot areas for each of the normal data and the abnormal data It may include a classification criterion determining unit.

In an embodiment, the machine tool operation data classification unit may generate the classification data by displaying the machine tool operation data in a Cartesian coordinate system through the plurality of classification criteria.

In an embodiment, the machine tool operation data classification unit may generate the classification data by combining the plurality of classification criteria into different types and displaying the machine tool operation data in a Cartesian coordinate system.

In an embodiment, the plot area determiner may determine a plot area for each of the normal data and the classification data for the abnormal data, and determine an area of each plot area for the normal data and the abnormal data.

In one embodiment, the classification criterion determining unit performs overlapping between plot areas for each of the normal data and the abnormal data, and sets the outermost plot line of the normal data and the abnormal data to a data boundary between the normal data and the abnormal data. Can be determined by

In one embodiment, the classification criterion determining unit

The similarity value can be calculated according to the following [Equation].

[Equation]

(Where, A is the area of the plot area of abnormal data, and B is the area of the plot area of normal data)

In an embodiment, the classification criterion determiner may determine a classification criterion for calculating the similarity value as a minimum value.

Among the embodiments, the method for extracting features of the diagnostic model data includes a machine tool operation data extraction step of extracting machine tool operation data including normal data and abnormal data generated based on vibrations generated when the machine tool is operated, Machine tool operation data classification step of generating classification data by classifying the machine tool operation data through a plurality of classification criteria, forming a plot line by connecting each individual data included in the classification data with a plot, and forming a plot line at the outermost side. A plot area determining step of determining the inside of the formed plot line as a plot area, and performing overlapping between the plot areas for each of the normal data and the abnormal data to determine a classification criterion that satisfies the calculated similarity value below a specific criterion It may include a step of determining the classification criteria.

In an embodiment, in the step of extracting machine tool operation data, the classification data may be generated by displaying the machine tool operation data in a Cartesian coordinate system through the plurality of classification criteria.

In an embodiment, the determining of a plot area may determine a plot area for each of the normal data and the classification data for the abnormal data, and determine an area of each plot area for the normal data and the abnormal data.

In one embodiment, in the determining of the classification criterion, the normal data and the abnormal data are overlapped with each other, while performing overlapping between the normal data and the abnormal data, the outermost plot line of the normal data and the abnormal data is determined as data between the normal data and the abnormal data. It is determined as a boundary, and the similarity value can be calculated according to the following [Equation].

[Equation]

(Where, A is the area of the plot area of abnormal data, and B is the area of the plot area of normal data)

The disclosed technology can have the following effects. However, since it does not mean that a specific embodiment should include all of the following effects or only the following effects, it should not be understood that the scope of the rights of the disclosed technology is limited thereby.

The apparatus and method for extracting features of diagnostic model data according to an embodiment of the present invention may determine an appropriate classification criteria for a diagnostic model by dividing normal data and abnormal data to extract classification data when determining a classification criteria.

An apparatus and method for extracting features of diagnostic model data according to an embodiment of the present invention generate classification data by displaying machine tool operation data in a Cartesian coordinate system through a plurality of classification criteria in the process of determining the classification criteria. can do.

The apparatus and method for extracting features of diagnostic model data according to an embodiment of the present invention may determine a classification criterion most suitable for generating training data based on an area of a plot area.

1 is a diagram illustrating a system for extracting features of data for a diagnostic model according to an embodiment of the present invention.
FIG. 2 is a block diagram illustrating a physical configuration of an apparatus for extracting features of data for a diagnostic model in FIG. 1.
FIG. 3 is a block diagram illustrating a functional configuration of the device for extracting features of diagnostic model data in FIG. 1.
FIG. 4 is a flowchart illustrating a process of determining a classification criterion performed by the apparatus for extracting features of diagnostic model data in FIG. 1.

Since the description of the present invention is merely an embodiment for structural or functional description, the scope of the present invention should not be construed as being limited by the embodiments described in the text. That is, since the embodiments can be variously changed and have various forms, the scope of the present invention should be understood to include equivalents capable of realizing the technical idea. In addition, since the object or effect presented in the present invention does not mean that a specific embodiment should include all or only such effects, the scope of the present invention should not be understood as being limited thereto.

Meanwhile, the meaning of terms described in the present application should be understood as follows.

Terms such as "first" and "second" are used to distinguish one component from other components, and the scope of rights is not limited by these terms. For example, a first component may be referred to as a second component, and similarly, a second component may be referred to as a first component.

When a component is referred to as being "connected" to another component, it should be understood that although it may be directly connected to the other component, another component may exist in the middle. On the other hand, when it is mentioned that a component is "directly connected" to another component, it should be understood that there is no other component in the middle. On the other hand, other expressions describing the relationship between components, that is, "between" and "just between" or "neighboring to" and "directly neighboring to" should be interpreted as well.

Singular expressions are to be understood as including plural expressions unless the context clearly indicates otherwise, and terms such as "comprises" or "have" refer to implemented features, numbers, steps, actions, components, parts, or It is to be understood that it is intended to designate that a combination exists and does not preclude the presence or addition of one or more other features or numbers, steps, actions, components, parts, or combinations thereof.

In each step, the identification code (e.g., a, b, c, etc.) is used for convenience of explanation, and the identification code does not describe the order of each step, and each step has a specific sequence clearly in the context. Unless otherwise stated, it may occur differently from the stated order. That is, each of the steps may occur in the same order as specified, may be performed substantially simultaneously, or may be performed in the reverse order.

The present invention can be implemented as computer-readable code on a computer-readable recording medium, and the computer-readable recording medium includes all types of recording devices storing data that can be read by a computer system. . Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tapes, floppy disks, and optical data storage devices. Further, the computer-readable recording medium is distributed over a computer system connected by a network, so that computer-readable codes can be stored and executed in a distributed manner.

All terms used herein have the same meaning as commonly understood by one of ordinary skill in the field to which the present invention belongs, unless otherwise defined. Terms defined in commonly used dictionaries should be construed as having meanings in the context of related technologies, and cannot be construed as having an ideal or excessively formal meaning unless explicitly defined in the present application.

1 is a diagram illustrating a system for extracting features of data for a diagnostic model according to an embodiment of the present invention.

Referring to FIG. 1, a system 100 for extracting features of data for a diagnostic model may include a user terminal 110, a device 130 for extracting features of data for a diagnostic model, and a database 150.

The user terminal 110 may correspond to a computing device capable of confirming data extracted and generated by the feature extraction device 130 of the diagnostic model data, and may be implemented as a smartphone, a laptop computer, or a computer, and is limited thereto. It can be implemented in various devices such as a tablet PC. The user terminal 110 may be connected to the device 130 for extracting features of the diagnostic model data through a network, and the plurality of user terminals 110 may be connected to the device 130 for extracting features of the data for the diagnostic model at the same time.

The apparatus 130 for extracting features of diagnostic model data may be implemented as a computer or a server corresponding to a program capable of determining a classification criterion. The device 130 for extracting features of the diagnostic model data may be wirelessly connected to the user terminal 110 through Bluetooth, WiFi, a communication network, and the like, and may exchange data with the user terminal 110 through a network.

In an embodiment, the apparatus 130 for extracting features of diagnostic model data may store necessary information in the process of determining a classification criterion in connection with the database 150. Meanwhile, unlike FIG. 1, the apparatus 130 for extracting features of data for a diagnostic model may be implemented by including the database 150 therein. The device 130 for extracting features of the diagnostic model data may be implemented including a processor, a memory, a user input/output unit, and a network input/output unit, which will be described in more detail with reference to FIG. 2.

The database 150 may correspond to a storage device that stores various pieces of information necessary in the process of determining a classification criterion by the feature extraction device 130 of the diagnostic model data. The database 150 may store machine tool operation data, classification data, and similar values, and is not limited thereto, and may store data required to drive the feature extraction device 130 for diagnostic model data.

FIG. 2 is a block diagram illustrating a physical configuration of an apparatus 130 for extracting features of diagnostic model data in FIG. 1.

Referring to FIG. 2, the apparatus 130 for extracting features of diagnostic model data may be implemented by including a processor 210, a memory 230, a user input/output unit 250, and a network input/output unit 270.

The processor 210 may execute a procedure for performing each operation in the process of determining the classification criteria, and manage the memory 230 that is read or written throughout the process, and the volatile memory in the memory 230 You can schedule the synchronization time between memory and non-volatile memory. The processor 210 may control the overall operation of the device 130 for extracting features of the diagnostic model data, and is electrically connected to the memory 230, the user input/output unit 250, and the network input/output unit 270. You can control the data flow. The processor 210 may be implemented as a central processing unit (CPU) of the device 130 for extracting features of diagnostic model data.

The memory 230 is implemented as a nonvolatile memory such as a solid state drive (SSD) or a hard disk drive (HDD), and is used to store all of the data required for the feature extraction device 130 for diagnostic model data. It may include, and may include a main memory device implemented as a volatile memory such as random access memory (RAM).

The user input/output unit 250 may include an environment for receiving a user input and an environment for outputting specific information to a user. For example, the user input/output unit 250 may include an input device including an adapter such as a touch pad, a touch screen, an on-screen keyboard, or a pointing device, and an output device including an adapter such as a monitor or a touch screen. In an embodiment, the user input/output unit 250 may correspond to a computing device connected through a remote connection, and in such a case, the feature extraction device 130 of data for a diagnostic model may be performed as a server.

The network input/output unit 270 includes an environment for connecting to an external device or system through a network, for example, a local area network (LAN), a metropolitan area network (MAN), a wide area network (WAN), and a VAN ( Value Added Network) may include an adapter for communication.

FIG. 3 is a block diagram illustrating a functional configuration of the feature extraction apparatus 130 for diagnostic model data in FIG. 1.

Referring to FIG. 3, a feature extraction device 130 for diagnostic model data includes a machine tool operation data extraction unit 310, a machine tool operation data classification unit 330, a plot area determination unit 350, and a classification criterion determination unit. 370 and a control unit 390 may be included.

The machine tool operation data extraction unit 310 is generated based on vibrations generated when the machine tool is operated and may extract machine tool operation data including normal data and abnormal data. Normal data is machine tool motion data that is generated when the machine tool operates without errors. The abnormal data is the motion data of the machine tool that is generated when an error occurs in the machine tool. Machine tool operation data is a position over time. For example, the machine tool operation data extraction unit 310 may extract a position over time for each of the X-axis, Y-axis, and Z-axis directions. For example, the machine tool operation data extraction unit 310 may extract a position for each axis with respect to a motion with respect to a bearing.

The machine tool operation data classification unit 330 may generate classification data by classifying machine tool operation data through a plurality of classification criteria. For example, the machine tool operation data classification unit 330 may generate classification data according to classification criteria of asymmetry, standard deviation, average, and square root of square mean.

In an embodiment, the machine tool operation data classification unit 330 may generate classification data by displaying machine tool operation data in a Cartesian coordinate system through a plurality of classification criteria. For example, the machine tool operation data classification unit 330 is an X-axis of the machine tool operation data extracted according to the machine tool operation data extraction unit 310 according to the classification criteria of asymmetry, standard deviation, average and square root mean square. Classification data can be generated by displaying in a Cartesian coordinate system consisting of and Y-axis.

In an embodiment, the machine tool operation data classification unit 330 may generate classification data by combining a plurality of classification criteria into different types and displaying the machine tool operation data on a Cartesian coordinate system. For example, the machine tool operation data classification unit 330 may generate classification data by displaying machine tool operation data in a Cartesian coordinate system based on a kurtosis value in the X-axis and asymmetry in the Y-axis as a classification standard. For another example, the machine tool operation data classification unit 330 may generate classification data by displaying normal data and abnormal data of a machine tool on the same orthogonal coordinate system based on the same classification criteria. In addition, the machine tool operation data classification unit 330 may generate a plurality of classification data for normal data and abnormal data by combining the classification criteria into different types. For example, looking at [Figure 1],

[Picture 1]

Normal data and abnormal data may be located in the same orthogonal coordinate system according to a plurality of classification criteria.

The plot area determiner 350 may form a plot line by connecting each individual data included in the classification data as a plot, and determine the inside of the plot line formed at the outermost side as the plot area. Classification data is a set of individual data. The individual data is one data generated according to a plurality of classification criteria for the position of the machine tool at a specific time according to the movement of the machine tool. The plot area determination unit 350 may connect individual data with a plot line according to the progress of time for the movement of the machine tool changed according to the change of time. The plot area determiner 350 may set the inside of the plot line formed at the outermost part of the plot line as the plot area. For example, referring to [Figure 2],

[Picture 2]

The plot area determiner 350 may connect between individual data for normal data and abnormal data with a plot line.

In an embodiment, the plot area determiner 350 may determine a plot area for each of the normal data and classification data for the abnormal data, and determine an area of each plot area for the normal data and the abnormal data. The plot area determination unit 350 may determine a plot area for normal data and abnormal data of a machine tool, and determine an area of each plot area to generate an index that can be compared with the normal data and the abnormal data.

The classification criterion determiner 370 may determine a classification criterion that satisfies the calculated similarity value below a specific criterion by performing overlapping between plot areas for each of the normal data and the abnormal data. The classification criterion determiner 370 may perform overlapping between the plot regions of the normal data and the abnormal data in order to detect a similarity between the plot regions of the normal data and the abnormal data.

In one embodiment, the classification criterion determination unit 370 determines the outermost plot line of the normal data and the abnormal data as a data boundary between the normal data and the abnormal data while performing overlapping between the plot regions for each of the normal data and the abnormal data. I can. For example, if there is an overlap between the normal data and the plot area for the abnormal data due to overlapping between the normal data and the plot area for the abnormal data, the classification criterion determination unit 370 The outer plot line can be determined as a data boundary between normal and abnormal data.

In an embodiment, the classification criterion determiner 370 may calculate a similarity value according to [Equation 1] below.

[Equation 1]

(Where, A is the area of the plot area of abnormal data, and B is the area of the plot area of normal data)

For example, the classification criteria determiner 370 may calculate a similarity value for classification data generated according to the same classification criteria for normal data and abnormal data according to [Equation 1].

In an embodiment, the classification criterion determiner 370 may determine a classification criterion for calculating a similarity value as a minimum value. For example, the classification criterion determination unit 370 calculates similar values for classification data for normal data and abnormal data generated according to the same classification criteria, and for normal data and abnormal data generated according to different classification criteria. A classification criterion having a minimum value can be determined by calculating and comparing similar values for classification data. For example, a plurality of plot regions may be generated based on the same normal data and abnormal data, and similar values may be compared, and a classification criterion having the smallest similar value may be determined. Referring to [Table 1],

[Table 1]

Since the similarity value according to the classification criteria for kurtosis and asymmetry among the similarity value calculation results is the smallest, the classification criteria determination unit 370 determines the classification criteria such as kurtosis value and asymmetry for the classification criteria for the diagnostic model of the machine tool. I can.

The control unit 390 controls the overall operation of the feature extraction device 130 of the diagnostic model data, and the machine tool operation data extraction unit 310, the machine tool operation data classification unit 330, and the plot area determination unit 350 And it is possible to manage the control flow or data flow between the classification criterion determining unit 370.

FIG. 4 is a flowchart illustrating a process of determining a classification criterion performed by the feature extraction apparatus 130 of data for a diagnostic model in FIG. 1.

Referring to FIG. 4, the feature extraction device 130 of the diagnostic model data is generated based on the vibration generated when the machine tool is operated through the machine tool operation data extraction unit 310, and includes normal data and abnormal data. It is possible to extract the operation data of the machine tool (S410).

The apparatus 130 for extracting features of diagnostic model data may generate classification data by classifying machine tool operation data through a plurality of classification criteria through the machine tool operation data classification unit 330 (S430).

The diagnostic model data feature extraction device 130 connects each individual data included in the classification data with a plot through the plot area determination unit 350 to form a plot line. It may be determined as a plot area (S450).

The apparatus 130 for extracting features of the diagnostic model data is classified by performing overlapping between the plot regions for each of the normal data and the abnormal data through the classification criterion determination unit 370 to satisfy the calculated similarity value below a specific criterion. The criteria can be determined (S470).

Although the above has been described with reference to preferred embodiments of the present invention, those skilled in the art will variously modify and change the present invention within the scope not departing from the spirit and scope of the present invention described in the following claims. You will understand that you can do it.

100: Data feature extraction system for diagnostic models
110: user terminal
130: Feature extraction device for diagnostic model data
150: database
210: processor 230: memory
250: user input/output unit 270: network input/output unit
310: machine tool operation data extraction unit
330: machine tool operation data classification unit
350: plot area determination unit 370: classification criterion determination unit
390: control unit

Claims (11)

A machine tool operation data extracting unit that is generated based on the vibration generated when the machine tool is operated and extracts machine tool operation data including normal data and abnormal data;
A machine tool operation data classification unit for generating classification data by classifying the machine tool operation data through a plurality of classification criteria;
A plot area determination unit configured to form a plot line by connecting individual data included in the classification data as a plot, and to determine an inside of the plot line formed at the outermost side as a plot area; And
And a classification criterion determining unit for determining a classification criterion that satisfies a similarity value calculated by performing overlapping between plot areas for each of the normal data and the abnormal data.
The method of claim 1, wherein the machine tool operation data classification unit
And generating the classification data by displaying the machine tool operation data in a Cartesian coordinate system through the plurality of classification criteria.
The method of claim 2, wherein the machine tool operation data classification unit
And generating the classification data by combining the plurality of classification criteria into different types and displaying the machine tool operation data in a Cartesian coordinate system.
The method of claim 1, wherein the plot area determination unit
And determining a plot area for each of the normal data and the classification data for the abnormal data, and determining an area of each plot area for the normal data and the abnormal data.
The method of claim 1, wherein the classification criterion determining unit
And determining an outermost plot line of the normal data and the abnormal data as a data boundary between the normal data and the abnormal data while performing overlapping between the plot regions for each of the normal data and the abnormal data Data feature extraction device.
The method of claim 5, wherein the classification criterion determining unit
A feature extraction device for diagnostic model data, characterized in that calculating the similarity value according to the following [Equation].
[Equation]

(Where, A is the area of the plot area of abnormal data, and B is the area of the plot area of normal data)
The method of claim 1, wherein the classification criterion determining unit
And determining a classification criterion for calculating the similarity value as a minimum value.
A machine tool operation data extraction step of extracting machine tool operation data including normal data and abnormal data, generated based on vibrations generated when the machine tool is operated;
A machine tool operation data classification step of classifying the machine tool operation data through a plurality of classification criteria to generate classification data;
A plot area determining step of forming a plot line by connecting each individual data included in the classification data with a plot, and determining an inside of the plot line formed at the outermost side as a plot area; And
And a classification criterion determining step of determining a classification criterion that satisfies a similarity value calculated by performing overlapping between plot areas for each of the normal data and the abnormal data.
The method of claim 8, wherein the step of extracting the machine tool operation data
And generating the classification data by displaying the machine tool operation data in a Cartesian coordinate system through the plurality of classification criteria.
The method of claim 8, wherein the determining of the plot area comprises:
And determining a plot area for each of the normal data and the classification data for the abnormal data, and determining an area of each plot area for the normal data and the abnormal data.
The method of claim 8, wherein the step of determining the classification criteria
Determining an outermost plot line of the normal data and the abnormal data as a data boundary between the normal data and the abnormal data while performing overlapping between the plot regions for each of the normal data and the abnormal data,
A method for extracting data for a diagnostic model, characterized in that the similarity value is calculated according to the following [Equation].
[Equation]

(Where, A is the area of the plot area of abnormal data, and B is the area of the plot area of normal data)

Images