KR101889198B1 - Apparatus and method for monitoring equipment - Google Patents

Abstract

According to an embodiment of the present invention, a facility monitoring apparatus and method are disclosed. The facility monitoring apparatus according to an embodiment of the present invention includes a sensor unit for sensing a state of an equipment and outputting a sensing signal, and a control unit for receiving the sensing signal, setting information related to setting of the equipment, Calculating a production statistical value by applying a statistical technique to a production deviation, which is a difference between the setting information and the product information, and calculating a production statistic value by applying a statistical technique to the sensing signal, Statistical value and the production deviation statistic value and the steady state standard data, and if it is determined that the equipment is abnormal as a result of the primary determination, the sensing statistic value and the production deviation statistical value And a state diagnosing and analyzing device for secondarily determining whether or not the equipment is abnormal based on the abnormal state standard data .

Description

[0001] Apparatus and method for monitoring equipment [

The present invention relates to an apparatus and a method for monitoring equipment installed in a factory, collecting and analyzing information related to facilities, and diagnosing facilities.

If problems arise in the facilities installed in the factory, problems arise such as a problem in the quality of the product to be produced or an inability to produce the product. In addition, in order to produce a better quality product, facilities placed in factories are heavily loaded.

For example, a number of rolling mills exist in factories producing steel and the like. However, in the current steel market, it is required to produce products having high quality, high tensile strength, high value-added, and physical properties not existing in the past, and the load on the rolling mills is increasing. Due to such an increase in load, problems such as facility breakage of rolling mills are increasing.

In addition, in the case of commercial devices for monitoring the state of the existing rolling mill equipment, the system is configured by using sensor technology simply by using sensor technology. Specifically, in the case of a conventional monitoring apparatus, a simple sensor and hardware devices are used to monitor the sensor signal, analyze only the frequency waveform, and indirectly determine the state of the equipment.

Published Patent Publication No. 2014-0087077

According to an embodiment of the present invention, there is provided a facility monitoring apparatus capable of more effectively monitoring factory facilities.

According to one embodiment of the present invention, there is provided a facility monitoring method capable of more effectively monitoring factory facilities.

The facility monitoring apparatus according to an embodiment of the present invention includes a sensor unit for sensing a state of an equipment and outputting a sensing signal, and a control unit for receiving the sensing signal, setting information related to setting of the equipment, Calculating a production statistical value by applying a statistical technique to a production deviation, which is a difference between the setting information and the product information, and calculating a production statistic value by applying a statistical technique to the sensing signal, Statistical value and the production deviation statistic value and the steady state standard data, and if it is determined that the equipment is abnormal as a result of the primary determination, the sensing statistic value and the production deviation statistical value And a state diagnosing and analyzing device for secondarily determining whether or not the equipment is abnormal based on the abnormal state standard data .

The apparatus for monitoring a facility according to an embodiment of the present invention is a rolling mill, and the sensor unit senses the vibration of the rolling mill and outputs the sensing signal. And information on the thickness of the product passed through the rolling mill can be input as the product information.

The apparatus for monitoring and diagnosing a facility monitoring apparatus according to an embodiment of the present invention includes system time information of a rolling controller for controlling the rolling mill, control information for each rolling mill, mechanical characteristics of each rolling mill, Physical characteristic information can be additionally input and stored.

The apparatus for diagnosing and analyzing conditions of a facility monitoring apparatus according to an embodiment of the present invention extracts a first outlier distance that is a Mahalanobis distance between each of the sensing statistical value and the production deviation statistical value and the steady state standard data , And firstly determines whether the facility is abnormal according to the first outlier distance, extracts a second outlier distance, which is a Mahalanobis distance between the sensing statistic value and the production deviation statistical value and the abnormal condition standard data, And secondarily determine whether the facility is abnormal according to the second outlier value.

The apparatus for monitoring and analyzing conditions of a facility monitoring apparatus according to an embodiment of the present invention may be configured such that a first number of times the first outlier distance is calculated to be equal to or greater than a first threshold value and a second number of times that the second outlier distance is less than a second threshold value And the abnormality degree of the facility can be determined based on the sum of the first number and the second number.

The apparatus for monitoring and analyzing conditions of a facility monitoring apparatus according to an embodiment of the present invention includes a data collection server for collecting the sensing signal, the setting information, and the product information, And a data processing server that receives the product information and calculates the sensing statistic value, the production deviation statistic value, the first outlier distance, and the second outlier distance, and determines whether the facility is abnormal, An algorithm library for applying the statistical technique, information on an algorithm library, and setting information on the operation of the algorithm library to the data processing server A data processing algorithm management server provided with And a rolling mill evaluation information database for mapping the sensing result of the data processing server, the sensing signal, the setting information, and the product information and storing the information.

The apparatus for diagnosing and analyzing conditions of a facility monitoring apparatus according to an embodiment of the present invention stores information on the sensing statistic value, the production deviation statistic value, the first outlier distance, and the second outlier distance, And a data predicting and life predicting server for extracting a time interval between times when the genital equipments are judged to be abnormal by using the information and evaluating the life and reliability of the facility based on the time interval.

The apparatus for monitoring and analyzing conditions of a facility monitoring apparatus according to an embodiment of the present invention may include a post-processing unit for inputting a result of post-processing verification to determine whether the facility is abnormal by the data processing apparatus, And the data processing algorithm management server may update the data stored in the standard database for facility condition evaluation using the verification result.

The facility monitoring apparatus according to an embodiment of the present invention may include a sensor monitoring unit, a first controller for outputting the setting information, a second controller for outputting the product information, and a communication And a communication protocol management device for setting and managing the setting items.

The facility monitoring apparatus according to an embodiment of the present invention may further include a virtual sensor simulator for generating and outputting a virtual sensing signal instead of the sensing signal according to a user's setting.

A facility monitoring method according to an embodiment of the present invention includes the steps of inputting a sensing signal indicating a state of a facility and calculating a sensing statistic value by applying a statistical technique to the sensing signal, Calculating a production deviation statistic value by applying a statistical technique to a production deviation which is a difference between the setting information and the product information, receiving the product information produced by the facility, calculating the production statistical value and the production deviation statistics Determining whether there is an abnormality in the facility using each of the sensing statistical value, the production deviation statistic value, and the abnormal condition standard data; And a step of judging.

The first determining step of the facility monitoring method according to an embodiment of the present invention may include extracting a first outlier distance that is a Mahalanobis distance between each of the sensing statistical value and the production deviation statistical value and the steady- And it is possible to determine whether the facility is abnormal according to the first outlier distance.

The secondarily determining step of the facility monitoring method according to an embodiment of the present invention extracts a second outlier distance that is a Mahalanobis distance between each of the sensing statistical value and the production deviation statistical value and the abnormal condition standard data And it is possible to determine whether the facility is abnormal according to the second outlier distance.

A facility monitoring method according to an embodiment of the present invention calculates a first number of times that the first outlier distance is calculated to be equal to or greater than a first threshold value and a second number of times that the second outlier distance is less than a second threshold value And determining the state of the facility according to the sum of the first number and the second number.

Thus, according to an apparatus monitoring apparatus and method according to an embodiment of the present invention, information related to the apparatus, for example, information collected by the sensor, information related to the operation of the apparatus, information And the like, and can register and manage an algorithm library for processing them dynamically. In addition, it is possible to provide an environment capable of processing and analyzing real-time data processing information that is processed in real time and processed in real time by processing and displaying the information and providing a large-capacity data processing function. In addition, it is possible to know in advance the occurrence of a problem problem of the facilities and the judgment of the constituent facility, and to predict the occurrence of the problems (for example, a fault and breakage of the facility) Can also help.

1 schematically shows a system including a facility monitoring apparatus according to the present invention.
FIG. 2 is a diagram schematically showing the configuration of an embodiment of the apparatus 200 for diagnosing and analyzing conditions according to an embodiment of the present invention shown in FIG.
3 is a view for explaining a process of collecting data by a data collection server of the facility monitoring apparatus according to an embodiment of the present invention.
FIG. 4 is a diagram for schematically explaining a data processing method by the data processing server 130 in the facility monitoring apparatus according to an embodiment of the present invention.
5 is a flowchart illustrating an apparatus monitoring method according to an embodiment of the present invention.
FIG. 6 is an operational flowchart for explaining step S100 of FIG. 5, which is interpreted by the statistical method of the facility monitoring method according to an embodiment of the present invention, in more detail.
FIG. 7 is an operational flowchart for explaining the step S200 of calculating the deviation of the facility monitoring method according to the embodiment of the present invention shown in FIG. 5 in more detail.
8 is a diagram for explaining the operation of the data processing algorithm management server.
9 is a view for explaining the operation of the large-capacity data post-processing and the life predicting server.
10 is a diagram for explaining the operation of the data post-analysis and verification apparatus.

Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings. However, the embodiments of the present invention can be modified into various other forms, and the scope of the present invention is not limited to the embodiments described below. Further, the embodiments of the present invention are provided to more fully explain the present invention to those skilled in the art.

FIG. 1 is a schematic view of a system including a facility monitoring apparatus according to an embodiment of the present invention, and is a diagram illustrating a case where a rolling facility is monitored. The rolling facility monitored by the facility monitoring apparatus according to an embodiment of the present invention may include a plurality of rolling mills 110, a rolling controller 120, and a rolling Supervisory Control Computer (SCC) 130. A facility monitoring apparatus according to an embodiment of the present invention may include sensors 140, a state diagnostic and analysis apparatus 200, a virtual sensor simulator 300, and a communication protocol management apparatus 400.

The plurality of rolling mills 110 can roll the steel sheet to a predetermined thickness under the control of the rolling controller 120.

The rolling controller 120 may be a programmable logic controller (PLC) and may control a plurality of rolling mills 110 according to various setup information received from the rolling SCC 130. The rolling controller 120 also includes control and characteristic information (i_con (i), i_con (i_con)) including at least one of the control information for each rolling mill, the mechanical characteristic signals for each rolling mill, the physical characteristic signals for each rolling mill, Can be output. The mechanical characteristics of the mills and the physical characteristics of the mills are measured by the speed of the mill, the motor current of each mill, the roll force of each mill, the roll gap of each mill, the tension information of the mills on and off, A bending rate, and an intermediate roll bending (IMR) ratio. The production status information of the product that rolls the mill may include at least one of strip thickness information and strip thickness deviation information. In addition, the rolling controller 120 may additionally output information on time used in the rolling controller 120, i.e., system time information.

The rolling SCC 130 outputs various setup information and the like to the rolling controller 120. Further, the rolling SCC 130 outputs setting information (i_set) including at least one of rolling mill setup information and rolling mill setup information. The information about the rolling operation and the mill setup information includes at least one of the currently produced coil product number, the steel grade information, the strip thickness and width existing information, the thickness setup information, the product deformation resistance information, the Forward slip information, Or more.

The sensors 140 may acquire status information on the status of the equipment and output a sensing signal sen including the status information. When the facility monitoring apparatus according to an embodiment of the present invention monitors the rolling facility, the sensors 140 may be a vibration sensor that detects and outputs the vibration of the rolling mill.

The state diagnostics and analysis apparatus 200 receives the sensing signal sen from the sensors 140 and receives control information from the rolling controller 120 for each rolling mill, mechanical characteristic signals for each rolling mill, physical characteristic signals for each rolling mill, (I_con), including at least one of the production state information of the product past the rolling mill, and at least one of the rolling mill information from the rolling SCC 130, and at least one or more of the rolling mill setup information (I_set), and interprets the received information using a statistical technique to determine whether the equipment is abnormal.

The state diagnostic and analysis apparatus 200 includes sensing signal data including system time information of the rolling controller 120 and vibration per rolling mill, control information for each rolling mill, mechanical characteristics information per rolling mill, , Information on the production status of the product passed through the mill, information about the rolling operation, and information about the mill setup information can be collectively managed. In addition, the state diagnosis and analysis apparatus 200 can determine the abnormality of the rolling mills by applying a statistical analysis technique to the collectively collected data, and can determine the reliability by determining the final conclusion through various steps have. Furthermore, the reliability of the judgment can be further improved by using the result of the determination and analysis of the state diagnostic and analysis apparatus 200 and the result of the post-test on the result of the determination. In addition, the state diagnostic and analysis apparatus 200 can include a plurality of servers, such as a server that executes an algorithm, a algorithm that is used for judgment, and a server that can update other set values, so that the set values can be changed without interruption It is possible to cope with environmental changes more flexibly.

The specific configuration and operation of the condition diagnosis and analysis apparatus 200 will be described later with reference to Figs. 2 to 10. Fig.

The virtual sensor simulator 300 can generate and output the virtual sensing signal v_sen using the sensing signal stored in the state diagnosis and analysis apparatus 200. [ For example, the virtual sensor simulator 300 may randomly extract stored sensing signals to generate a virtual sensing signal v_sen and output the generated virtual sensing signal v_sen.

The virtual sensor simulator 300 may operate according to the operation mode set by the user or may not operate. The facility monitoring apparatus according to an embodiment of the present invention further includes a virtual sensor simulator 300, thereby enhancing the reliability of facility evaluation. For example, when the user determines that there is a problem with the operation of the sensors 140, a signal obtained by randomly extracting an existing sensing signal instead of the sensing signal sen inputted from the sensors 140 (v_sen)), it is possible to assure the reliability of the facility evaluation even if an error occurs in the operation of the sensors 140. In other words, when there is a problem in an actual sensor, the state diagnostic and analysis apparatus 200 performs a predetermined algorithm on an existing stored and managed sensing signal, thereby enhancing the consistency.

The communication protocol management device 400 may be configured to communicate with the state diagnostics and analysis device 200 through an interface scheme, protocol scheme, communication scheme A professional format, and IP address and port information, and outputs the communication setting information (com_set) including the set information to the status diagnosis and analysis apparatus 200. [ The above-described interface method, protocol method, communication professional format, and IP address and port information can be set and stored by setting a user-defined communication protocol setting information file. The facility monitoring apparatus according to an embodiment of the present invention further includes the communication protocol management apparatus 400 so that communication can be performed by applying the changed contents even when the setting information for communication during operation is changed.

FIG. 2 is a diagram schematically showing the configuration of an embodiment of the apparatus 200 for diagnosing and analyzing a facility monitoring apparatus according to an embodiment of the present invention shown in FIG. 1, The state diagnostic and analysis apparatus 200 includes a data collection server 202, a data processing server 204, a data processing algorithm management server 206, a data storage management server 208, a mass data postprocessing and life prediction server 210 A rolling stock assessment database 216, a data and information display 218, and a system watchdog management device 220. The standard database 212, the data post-analysis and verification device 214, the rolling mill evaluation information database 216, .

The data collection server 202 receives at least one of a sensing signal sen indicating the status of the facility to be monitored, operation status of the facility to be monitored, information on products produced by the facility, It can be collected collectively. 1, the data collection server 202 receives the sensing signal sen from the sensors 140 (Fig. 1), and the rolling controller (Fig. 1 From the rolling SCC (130 in FIG. 1), the rolling operation information and the rolling mill information from the rolling SCC (120 in FIG. 1), at least one of the control information for each rolling mill, the mechanical / physical characteristic signal information, At least one of the setup information may be received. Also, as described above, the current system time information can be received from the rolling controller (120 in FIG. 1), and the received current system time information can be utilized as the reference system time information.

1) from the communication protocol management apparatus (400 in FIG. 1), and data collection and communication necessity setting information from the rolling machine controller 120, rolling SCC 130 Receives it, imports it, recognizes it, and operates as set up. At this time, the data collection server 202 has an interface method for interfacing with the sensors (140 in FIG. 1), communication protocols and professional setting information for communicating with the rolling controller (120 in FIG. 1) From the communication protocol management device (400 in FIG. 1) a user defined communication protocol setting information file including a communication protocol for communicating with the communication protocol 130 and related specialized setting information. With this arrangement, it is possible to dynamically determine the interfaces, the protocol environment, and the communication professional form for communicating with the sensors (140 in FIG. 1), the rolling controller (120 in FIG. 1), the rolling SCC (eh 1 dml 130). In addition, even if the information for communication is changed during system operation, the changed contents can be reflected again without stopping the system, so that flexibility of the apparatus can be provided. That is, even if the interface and the communication environment in which the data collection server 202 collects data are changed frequently by the field environment, the changed environment can be reflected without system redevelopment and / or system shutdown.

The data collection server 202 periodically receives an Alive message request from the system watchdog management apparatus 220 and transmits an Alive message to the system watchdog management apparatus 220 when the data collection server 202 continuously operates. That is, the data collection server 202 may periodically inform the system watchdog management apparatus 220 that it is operating.

The data collection server 202 also maintains data storage queue areas for storing data received from the sensors (140 in FIG. 1), the rolling controller (120 in FIG. 1), and the rolling SCC And simultaneously stores the corresponding data in the corresponding cue areas simultaneously with the data collection.

Also, the data collection server 202 can manage the time information received from the rolling controller (120 in FIG. 1), the rolling operation information, the information on the product to be produced, and the data such as the production speed. In addition, various signal data to which the system information of the rolling controller (120 in FIG. 1) is mapped may be transmitted to the data processing server and the data storage management server.

The data processing server 204 may receive various information from the data collection processing server 202, evaluate and monitor the status of the equipment through processing of the information, and determine whether there is a problem with the equipment. To this end, the data processing server 204 may receive data processing algorithm libraries and user-defined data processing algorithm operation setting information and the like from the data processing algorithm management server 206. Specifically, when monitoring the mills, the data processing server 204 receives sensing signal information from the data collection processing server 202, control information for each mill, physical / mechanical signaling information for each mill, Production status information, rolling operation information, rolling mill setup information, and the like. The data processing algorithm library for monitoring, evaluating, and diagnosing the state of the rolling mill equipment by processing the received information and operation setting information of the data processing algorithm library are received from the data processing algorithm management server 206. The data processing server 204 imports and applies the received algorithm library and its operation setting information, etc. to drive the driven algorithm, and receives data from the data collection server 202 to create input data in the rolling mill facility condition evaluation algorithm library items Queuing the received information, and evaluating and diagnosing the condition of the mill equipment. The operation for evaluating the equipment condition of the mill by the data processing server 204 can be dynamically changed during system operation by the real-time data processing algorithm library and operation setting information set by the user, and if necessary, It may also work.

The specific rolling machine equipment condition evaluation operation by the data processing server 204 will be described later with reference to Figs.

The data processing algorithm management server 206 manages the algorithm libraries for evaluating and diagnosing the state of the rolling mill equipment in real time by processing the data by the data processing server 204, Algorithm library and operation setting information document file. The definition of the document file can be performed through the definition tool, and the user can define the document file at any time using the definition tool. If the defined contents are to be reflected in the system, the contents of the corresponding configuration information are transferred to the data processing server and can be reflected during system operation and can be reflected without system stoppage. Concretely, the execution step information of the data processing algorithm, the library information to be executed step by step, the input and output information of the executed libraries, the maximum limit values of the outlier distance used in the data processing server, And allows the user to set the facility status character information (Normal etc.) for each count count exceeding the maximum limit value. Such configuration information files can be defined by languages that can efficiently express configuration information files such as XML (eXtensible Markup Language). And, the configuration for such an execution step can be changed flexibly without limitation.

The data storage management server 208 may store and manage the processing result information collected or processed in the data collection server 202 and / or the data processing server 204 in the rolling machine evaluation information database 216. The data storage management server 208 periodically checks the size capacity of the rolling mill evaluation information database 216 and can delete old data when the used capacity exceeds the maximum size capacity set by the user. Also, the user can set the data size to be deleted periodically, regardless of the size, and set the data deletion to be executed according to the period. These contents can be defined in the operation configuration file of the data storage management server.

The mass data post-processing and life prediction server 210 accumulates and manages various data generated by algorithms driven for evaluating and diagnosing the state of the rolling mill equipment in real time from the data processing server 203, And post-processing to identify long-term trends in equipment life at the process line level included in current facilities. That is, the mass data post-processing and life prediction server 210 can predict reliability information and life-time-related information of the current facility line by using the data received from the data processing server 203. To this end, the mass data post-processing and lifetime prediction server 210 may use machine learning, data mining, or artificial intelligence algorithms. For example, when using a neural network, which is one of artificial intelligence, for example, the input values are the received information, and the output value is the time at which the rolling process line occurred, i.e., MTBF (Mean Time Between Failure) If new input values arrive, the MTBF can be updated to estimate the expected time of the fault. The post-treatment may also include an assessment of the level of driving current whether the operating operators are operating correctly or not. Also, the algorithm for obtaining the MTBF can be used equally, and the output value is the operating level of the process line. This level of operation is obtained by initially inputting the operation level information from the operator, mapping the input values according to the respective operation level information, constructing the artificial intelligence network, and then using the constructed artificial intelligence network To estimate and assess current operational levels. That is, the mass data post-processing and life prediction server 210 can evaluate the reliability of the rolling process line in a comprehensive manner. Information data extracted through the mass data post-processing and life prediction server 210, that is, post-processing result information, process line life prediction information, and reliability information may be transmitted to the data and information display device 218 and displayed.

The standard database 212 for evaluating the rolling mill equipment condition can store basic data for evaluating the rolling mill equipment condition. Specifically, in the standard database 212 for evaluating the equipment condition of the rolling mill, abnormality data groups, which are data groups in the case where the state of the equipment is normal and data groups in the case where the state of the equipment is abnormal, are stored and managed. In other words, the vibration sensor signal information for each rolling mill, the control information for each rolling mill and the physical / mechanical characteristics information when the rolling mill equipment is in a normal state, the product production status information about the rolling mill, the rolling operation information, The vibration sensor signal information for each rolling mill, the control information for each rolling mill, the physical / mechanical characteristics information, the product production status information on the rolling mill, the rolling operation information, and the rolling mill setup information are also managed in accordance with the coil information It is managed according to product coil information by production speed. These groups of normal and abnormal standard data may be managed and added or deleted by the data processing algorithm management server 206. That is, as described above, the data processing algorithm management server 206 receives the coil product number when the rolling mill equipment state is normal, the abnormal state, and the coil product number In the rolling machine evaluation information database 216, control and physical / mechanical characteristic signal information for each rolling mill, product production status information past the rolling mill, rolling mill and mill setup information, and statistical Processing method (RMS, etc.), processing result information, and Mahalanobis distance information can all be analyzed and stored in the standard database 212 for the rolling mill equipment condition evaluation. The total number and size of the normal data group and the abnormal data group can be set and limited in the real-time data processing algorithm library and the operation setting information file, and when the limit is reached, Mode or mode information that can no longer be added, and so on, so that the latest normal data group and abnormal data group can be managed. The data processing algorithm server 206 reads the corresponding configuration file at startup, applies it, operates it, and re-applies it without stopping the system operation. In addition, for initial start-up, the user can input normal data signals to the standard database 212 for rolling stock facility evaluation through the data processing algorithm management server 206, which is provided through the tool.

The data post-analysis and verification device 214 can provide an environment for the user to inquire and analyze the information data stored and managed by the data storage management server 208. [ At the time of inquiry of the information data, the user can inquire, search and analyze data by various conditions such as data type, data storage time, product product information, etc. using the data post-analysis and verification device 214. Also, in this analysis environment, the signal data information collected by the data collection server 202 and the information of the signal data processed by the data processing server 204 are stored and can be analyzed and analyzed. In the analysis, it provides an environment that can analyze data by applying RMS, Peak-to-Peak, average and so on, digital signal processing techniques such as FFT and statistical processing techniques. Also, the data processing server 204 in the verification environment can input the feedback on the extracted rolling mill equipment condition evaluation result by the user. For example, if the evaluation result of the rolling mill equipment condition is normal but abnormal, the number of the coil product is transmitted to the data processing algorithm server 206. The data processing algorithm server 206 uses the number of the corresponding coil product to calculate data on the coil product determined to be abnormal in the rolling machine evaluation information database 216, Vibration sensor signal information received via the data storage management server 208 and stored via the data storage management server 208 and control and physical / mechanical characteristic signal information per rolling machine, product production status information past the rolling mill, And the data related to the coil product among the rolling machine setup information are collected and classified into the abnormal data group in the standard database 212 for evaluating the rolling mill equipment condition. In the opposite case, if the result of the rolling mill equipment condition evaluation is determined to be abnormal but the user normally inputs the data through the data post-analysis and verification unit 214, the data processing algorithm management server 206 sets the coil product number of the data, And the data processing algorithm management server 206 classifies the coil product number and the data determined as normal into the normal data group. By such an operation, it is possible to improve the consistency of the algorithm by feedback of the evaluation result of the rolling mill equipment condition dynamically.

The rolling mill evaluation information database 216 may store and manage the information collected through the data collection server 202 and the processing result information processed by the data processing server 204. In the rolling machine evaluation information database 216, corresponding information transmitted by the data storage management server 208 is stored and managed.

The data and information display device 218 includes information data transmitted from the data collection server 202, i.e., sensing signal information, control information for each rolling mill, physical / mechanical characteristic signal information for each rolling mill, The rolling operation information and the rolling set-up information, and the rolling condition evaluation information generated by the data processing server 204 through the algorithm driven to evaluate and diagnose the condition of the equipment in real time, can do. That is, the data and information display device 218 can perform as an HMI role. In addition, the data and information display device 218 can also provide a function of managing the user management and access authority information and enabling the user to execute the screen by authentication. Then, the data processing server 204 can be initialized by transmitting a command for resetting the count of the total sum values of Mahalanobis distance information that is an outliers distance exceeding the maximum limit value. Also, the data and information display device 218 can send and receive Alive messages to and from the system watchdog management device 220 in the same manner as other server devices. The data and information display device 218 receives the information data received from the large-capacity data post-processing and life prediction server 210, that is, the post-processing result information, the process line life prediction information, and the reliability information, Can be displayed.

The system watchdog management apparatus 220 periodically exchanges alive messages with other servers and devices to confirm whether the configuration servers of the current system are normally operating. If a problem arises, the problem status display is basic on the screen so that it can be notified to the system administrator and the user or the operating driver, and furthermore, the mail or character can be transmitted. The period in which the system watchdog management apparatus 220 exchanges alive messages with other components can be dynamically set by the user. If the period is adjusted according to the communication load, it can be reflected without stopping the system have. In addition, the corresponding configuration file includes counterpart node information, that is, an IP number and a communication port number, to which an Alive message should be exchanged, so that the counterpart nodes to which such an Alive message is to be exchanged can be set and reflected dynamically. This information can be described in the user-defined system watchdog management configuration information file.

3 is a view for explaining a process of collecting data by a data collection server of the facility monitoring apparatus according to an embodiment of the present invention.

First, information on the necessity of communication with the sensors 140, the rolling SCC 130, and the rolling controller 120, and the like, and the communication interface and / or protocol operation information are set by the communication protocol management apparatus 400 , The set information may be transmitted to the data collection server 202 (S1010 and S1020)

Next, the data collection server 202 receives (S1030 and S1040) information such as vibration per rolling machine obtained by the sensors 140 from the sensors 140 according to the set information, or the rolling SCC 130 (S1050 and S1060) of outputting the rolling operation information and / or the rolling set-up information (S1050 and S1060), or by controlling the rolling machine control information outputted by the rolling controller 120, mechanical / physical characteristic information per rolling mill, Information can be received.

The data collection server 202 may perform various mapping operations on the received information (S1090). For example, the rolling operation information such as the received sensing signal, the control information for each rolling mill, the mechanical / physical characteristic information, and the coil number of the manufactured product may be matched or the received sensing signal, the control information for each rolling mill, Or map the product distance information to the received sensing signal, control information per rolling machine, and mechanical / physical property information.

The data collection server 202 may output the mapped data to the data storage management server 208 and / or the data processing server 204.

FIG. 4 is a diagram for schematically explaining a data processing method by the data processing server 130 in the facility monitoring apparatus according to an embodiment of the present invention.

First, the data processing algorithm management server 206 sets the operation setting information of the data processing algorithm libraries, and outputs the set operation setting information and the execution target data processing algorithm library to the data processing server 204 (S2010 and S2020).

Next, the information mapped in the data collection server 202, that is, the mapping signal, the control information for each rolling mill, the mechanical / physical characteristic information for each rolling mill, the product production status information passed through the rolling mill, the rolling operation information, Operation setup information, and the like are transmitted to the data processing server 204 (S2040, S2050, and S2060).

Next, the data processing server 204 executes the one-time rolling machine condition evaluation algorithm and outputs the executed result to the data storage management server (S2070, S2080).

The data storage management server 208 stores the received information (S2090).

Although not shown, the data processing server 204 may output the result of executing the one-time rolling mill condition evaluation algorithm to the data display device 218. [

5 is a flowchart illustrating an apparatus monitoring method according to an embodiment of the present invention.

First, the sensed information, the control information for each rolling mill, and the physical / mechanical characteristic signal information for each rolling mill are analyzed statistically to determine whether the state of the equipment is abnormal (S100). It is possible to accumulate the data for a predetermined time (for example, one second) each time the information is stored. The analysis through the statistical method is performed by receiving data from the data processing algorithm management server as basic statistical processing techniques such as mean, standard deviation, RMS (Root Mean Square), Peak-to-Peak, Skewness, Kurtosis, Crest Factor and K- Can be performed by first extracting them using the libraries that implement the statistical processing technique.

Next, in step S200, the production state information of the product passing through the rolling mills and the actual rolling operation and setup information are obtained using the production state information of the product, the rolling operation information, and the rolling mill setup information. For example, it is possible to obtain the difference in thickness between the products being produced and the difference in thickness between the originally produced product and the actual produced thickness. The average, standard deviation, RMS, Peak-to-Peak, Skewness, Kurtosis, Crest Factor, and K-Factor values for the deviation are extracted after accumulating data generated for each predetermined time (for example, one second) can do.

Next, the abnormal value distance between the statistical values extracted at step S100 and the statistical values for the deviation or deviation extracted at step S200 and the steady state data may be calculated, and it may be determined based on the calculated distance (step S300 ). The differential distance may be Mahalanobis distance. When searching for Mahalanobis distance, other steady-state data can be used for different mill speeds. That is, the standard data groups in the steady state can be stored in the standard database for evaluating the state of the equipment according to the rolling machine speed. In step S300, appropriate steady state data are taken from the standard database And can calculate the Mahalanobis distance between the corresponding steady state data and the currently extracted statistical value or deviation. That is, the normal standard data can be classified and managed in detail based on the operating speed standard in the standard database for evaluating the condition of the mill equipment. If it is determined that the rolling mill equipment condition evaluation result is normal, the information on the product coil that determines the steady state is transmitted to the data processing algorithm management server so that the related information data can be managed in the standard data group . The Mahalanobis distance is an algorithm for evaluating an abnormal value of data, and is an algorithm for determining how far the outlier distance from actual normal standard data is.

Next, the abnormal value distance between the statistical values for the statistical values extracted in step S100 and the deviation or deviation extracted in step S200 and the abnormal state data may be calculated, and it may be determined based on the calculated distance (step S400 ). The step S400 may be executed when it is determined that the equipment status is abnormal as a result of the determination in the step S300. The differential distance may be Mahalanobis distance. That is, the abnormal value data (i.e., the outlier data group) stored in the standard database for evaluating the equipment condition of the rolling mill, such as statistical values or deviations calculated based on the present data, and the outliers are re-evaluated. If the outlier distance calculated in step S400 is less than or equal to an arbitrary threshold value, it can be determined that there is an abnormality in the facility.

In addition, the product coil information in which the current rolling mill equipment condition information is evaluated as an abnormal value may be transmitted to the data processing algorithm management server so that it can be managed as the abnormal standard data group in the rolling stock facility condition evaluation database.

Next, if it is determined in step S300 and step S400 that the current state of the rolling mill is abnormal, it is determined whether or not the equipment is abnormal based on the number of times the total value of the Mahalanobis distances exceeds the maximum limit value Step S500). The sum of the Mahalanobis distances of the statistical processing results of the data calculated in step S100 and the sum of the Mahalanobis distances of the statistical processing results of the data calculated in step S200, that is, Mahalanobis distance information extracted in step S300 Counts the number of times the total sum values exceed the maximum limit value (threshold value), and accumulates the counted value. Based on the accumulated count value, the resume diagnosis status information can be mapped. For example, when the maximum limit value is not exceeded, it is normal. When the number is more than 10, (Abnormal), when it exceeds 20, (Bad) Etc., can be represented by characters.

The maximum limit value may be defined by the user through the data processing algorithm management server and received and applied when the data processing server is operating. These maximum limit values can be dynamically re-reflected when changed during system operation. Also, the rolling mill condition diagnosis mapping information can be set in real time by the user through the data processing algorithm management server and can be re-reflected during dynamic reflection or operation.

5, steps S100 and S200 may be performed simultaneously.

The step-by-step operation can be changed by setting the data processing algorithm management server to modify according to the real-time data processing algorithm library and operation setting information set by the user and dynamically set whether to perform the system operation or not, So that it can be operated.

FIG. 6 is an operational flowchart for explaining step S100 of FIG. 5, which is interpreted by the statistical method of the facility monitoring method according to an embodiment of the present invention, in more detail.

First, data on the sensing signal information is patched (step S110).

Next, the control information for each rolling machine and the mechanical / physical characteristic signal information for each rolling machine are patched (step S120).

Next, the fetched data is accumulated for a predetermined time (for example, 1 second) (step S130)

Next, a basic statistical processing technique library for each accumulated data is executed, and statistical processing technique result information for each accumulated data is extracted (step S140). The statistical processing technique result information can be at least one of mean, standard deviation, RMS, peak-to-peak, skewness, and kurtosis.

FIG. 7 is an operational flowchart for explaining the step S200 of calculating the deviation of the facility monitoring method according to the embodiment of the present invention shown in FIG. 5 in more detail.

First, data on product production information past the rolling mill is fetched (step S210).

Next, data on the rolling operation information and / or setup information for the rolling operation is fetched (step S220).

Next, the deviation between the data on the product production information and the data on the basis of the operation information and the setup information is extracted (step S230).

Next, the basic statistical processing techniques for the extracted deviations are executed and libraries related to the averaging processing technique are extracted. The statistical processing technique result information can be at least one of mean, standard deviation, RMS, peak-to-peak, skewness, and kurtosis.

The facility monitoring method according to one embodiment of the present invention illustrated in FIGS. 5 to 7 can be performed by the state diagnosis and analysis apparatus of FIG. More specifically, the facility monitoring method according to one embodiment of the present invention described in Figs. 5 to 7 can be performed by the data processing server (204 in Fig. 2) of the state diagnosis and analysis apparatus.

8 is a diagram for explaining the operation of the data processing algorithm management server.

The user creates a library of data processing algorithms, information about data processing algorithm libraries, and configuration information for operation of the data processing algorithm libraries. The data processing algorithm management server 206 can store and manage the result created by the user, and can generate information on the result as a user-defined data processing algorithm operation setting information file. The data processing algorithm management server 206 may output the processing algorithm library, information on the data processing algorithm libraries, and setting information for operation of the data processing algorithm libraries to the data processing server 204.

9 is a view for explaining the operation of the large-capacity data post-processing and the life predicting server.

First, the data processing server 204 outputs the result of processing using the statistical technique to the received data, the analysis result thereof, and the like to the large capacity post-processing and life prediction server 210 (S3010, S3020).

The mass post-processing and life prediction server 210 stores and manages the data received from the data processing server 204. Also, the large capacity post-processing and lifetime prediction server 210 generates lifetime information of equipment and operation level evaluation information of facility drivers by executing a mean time between failure (MTBF) prediction algorithm and an operation level evaluation algorithm, (S3030, S3040).

10 is a diagram for explaining the operation of the data post-analysis and verification apparatus.

First, the user inputs information for data consistency evaluation to the data post-analysis and verification device 214 (S4010, S4020). For example, the user may enter feedback on the results for equipment conditions analyzed by the data processing server 204.

The data post-analysis and verification device 214 acquires information (e.g., coil information) about the product whose data consistency is evaluated and outputs the information to the data processing algorithm management server 206 (S4030, S4040). For example, the analysis result of the equipment status by the data processing server 204 is normal, but in reality, the information of the abnormality-determined product or the analysis result of the equipment status by the data processing server 204 is abnormal The information about the product determined to be normal may be output to the data processing algorithm management server 206 in practice.

The data processing algorithm management server 206 may inquire and collect data related to the information on the product input from the rolling machine evaluation information database 216 (S4050 and S4060). The data includes information about sensing signals related to the product, control information for each mill, physical / mechanical characteristics for each mill, production status information of the products past the mill, information about rolling operations, And the like.

The data processing algorithm management server 206 may classify the collected data into a normal standard data group or an abnormal standard data group according to the consistency evaluation result described above and store the result in the standard database 212 for evaluating the condition of the equipment condition (S4070, S4080).

Through this process, the data stored and managed in the rolling machine evaluation information database 216 may be added to the standard database 212 for evaluating the condition of the rolling mill equipment condition with normal standard data or abnormal standard data by the user, The normal standard data or the abnormal standard data stored in the evaluation standard database 212 may be deleted. Then, the normal standard data and the abnormal standard data stored in the standard database 212 for evaluating the equipment condition of the rolling mill can be used as standard data when the algorithm is driven by the data processing server.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, It will be obvious to those of ordinary skill in the art.

110: rolling mills 120: rolling controller
130: Rolling SCC 140: Sensors
200: state diagnosis and analysis apparatus 300: virtual sensor simulator
400: communication protocol management device

Claims (14)

A sensor unit for sensing a state of the equipment and outputting a sensing signal; And
(RMS), a Peak-to-Peak (RMS) value, and a Peak-to-Peak ), A skewness, a kurtosis, a crest factor, and a K-factor as a sensing statistic value, and generates at least one of a production statistic value, which is a difference between the setting information and the product information, Peak-to-Peak, Skewness, Kurtosis, Crest Factor, and K-Factor, as well as mean, standard deviation, Root Mean Square (RMS) And generating a production deviation statistical value, and firstly determining whether the equipment is abnormal based on the sensing statistic value and the production deviation statistical value and the steady state standard data, If it is judged to be abnormal, the sensing statistical value and the production deviation statistical value, And a status diagnosis and analysis device for secondarily determining whether the equipment is abnormal based on the standard data,
The state diagnostic and analysis device
A first abnormal value distance which is a Mahalanobis distance between each of the sensing statistical value and the production deviation statistical value and the steady state standard data is extracted and the abnormality of the facility is primarily determined according to the first abnormal value distance ,
Extracting a second outlier distance that is a Mahalanobis distance between each of the sensing statistical value and the production deviation statistical value and the abnormal state standard data and secondarily determining whether the equipment is abnormal according to the second outlier distance Facility monitoring device.
The method according to claim 1,
The facility is a rolling mill,
Wherein the sensor unit senses the vibration of the rolling mill and outputs the sensing signal,
Wherein the state diagnosis and analysis apparatus inputs information on the target thickness as the setting information and inputs information on the thickness of the product passed through the rolling mill as the product information.
3. The apparatus of claim 2,
System information of the rolling controller for controlling the rolling mill, control information for each rolling mill, mechanical characteristics of the rolling mill, and physical characteristics of the rolling mill.
delete 2. The apparatus of claim 1,
A first number of times the first ideal value distance is calculated to be equal to or greater than a first threshold value and a second number of times the second ideal value distance is calculated to be equal to or less than a second threshold value, And an abnormality determination unit for determining an abnormality of the facility based on the value of the facility monitoring apparatus.
delete 2. The apparatus of claim 1,
And stores information on the sensed statistical value, the production deviation statistic value, the first outlier distance, and the second outlier distance, and extracts a time interval between times when the genital organ is judged to be abnormal using the stored information And further evaluating the lifetime and reliability of the facility based on the time interval.
2. The apparatus of claim 1,
And updates the steady state standard data and the abnormal state standard data using a result of the post-verification to determine whether the facility is abnormal.
The facility monitoring apparatus according to claim 1,
A communication controller for setting and managing communication setting items related to communication between each of the sensors, the first controller for outputting the setting information, and the second controller for outputting the product information, A further facility monitoring device.
The facility monitoring apparatus according to claim 1,
And a virtual sensor simulator for generating and outputting a virtual sensing signal instead of the sensing signal according to a setting of a user.
1. A facility monitoring method performed by a state diagnostic and analysis apparatus comprising a plurality of servers,
The state diagnosis and analysis apparatus inputs a sensing signal indicating the state of the facility and calculates an average, a standard deviation, a root mean square (RMS), a peak-to-peak, Generating at least one of a skewness, a kurtosis, a crest factor, and a K-factor as a sensing statistic value;
Wherein the state diagnosis and analysis apparatus receives the setting information of the facility and the product information produced by the facility, and calculates a production deviation average, a standard deviation, a root mean square (RMS) difference between the setting information and the product information, , One or more of a peak-to-peak, a skewness, a kurtosis, a crest factor, and a K-factor as a production deviation statistic value step;
The condition diagnosis and analysis apparatus may include a step of primarily determining whether the equipment is abnormal using the sensing statistical value and the production deviation statistical value and the steady state standard data, And
Wherein the state diagnosis and analysis device includes a step of secondarily determining whether an abnormality of the equipment is abnormal using the sensing statistical value and the production deviation statistical value and the abnormal condition standard data,
Wherein the first determining step comprises: extracting a first outlier distance that is a Mahalanobis distance between each of the sensing statistic value and the production deviation statistical value and the steady state standard data, If it is determined that the abnormality is present,
Wherein the secondarily determining step comprises the steps of: extracting a second outlier distance that is a Mahalanobis distance between the sensing statistic value and the production deviation statistical value and the abnormal condition standard data; A facility monitoring method to determine anomaly.
delete delete 12. The method of claim 11,
A first number of times the first ideal value distance is calculated to be equal to or greater than a first threshold value and a second number of times the second ideal value distance is calculated to be equal to or less than a second threshold value, And determining the state of the facility based on the value of the facility.

Images