JP5115791B2 - Plant operation data management system - Google Patents

Description

  The present invention relates to a plant operation data management system that collects a desired range of continuous plant operation data held in a database or the like and facilitates analysis.

  Patent Document 1 discloses a plant information management system that collects trend data indicating the current operation status from a process control system and displays the trend data in a superimposed manner with reference trend data.

In the current manufacturing industry, not only the plant is operated safely and stably, but also the efficiency of plant operation from various angles is required due to intensifying competition. for that reason,
(1) It is necessary to construct and change the line in a shorter period of time.
(2) It is necessary to eliminate waste in order to fully operate more sophisticated and complicated lines.
(3) It is necessary to manufacture high-quality products and improve yield.
It is said that.

  In order to manufacture a particularly high quality product, results management represented by finished product inspection is performed. However, even though this can prevent the outflow of quality defects, it cannot prevent the occurrence of quality defects.

  Therefore, grasping the process behavior and the state of the apparatus quickly and accurately is one of the important tasks for the plant operator. However, many of these tasks are left to the judgment based on the experience of the operator, and there are many situations where the situation of the entire plant cannot be grasped objectively.

  In order to objectively grasp the situation of the whole plant, process management by managing equipment (Material), materials (Material), human operations (Man Power), and procedures (Method), especially called 4M It is important to collect data that facilitate (visualize) such management.

The main operations performed by operators to understand the plant status are as follows.
(1) The operator confirms the production plan and production results.
(2) The operator searches the trend of a specific instrument from the historical database based on past experience, and determines the trend.
(3) The operator confirms the operation performed in the past from the event history window.
(4) The operator finds a break by arranging the above information in time series. For example, the period from when the raw material is introduced until the product comes out is divided into one segment.
(5) The operator imagines the situation of the plant in the head based on past experience in units of divisions.
For example, how is the product as usual, how long can you achieve the plan?

  FIG. 23 is an image diagram of a conventional plant operation data search. Plant operation data is collected by an operator who manually accesses a data server and selects and searches a target database from various databases.

  FIG. 24 is a display screen example of searched plant operation data. The acquired plant operation data is centered on trend data displayed in a graph shown in FIG. 24A or event history data listed in time series shown in FIG.

  FIG. 25 is an image diagram that hierarchically displays knowledge generation from data to wisdom. The process from the various human data to the material for judging events in the mind (propagation of knowledge) understands the information obtained by searching and processing the lowest level data, Knowledge is gained or internalized to gain wisdom, which is expressed and fed back to knowledge.

By streamlining the process (area A) to display (visualize) various data so that it can be understood as human knowledge, it is possible to more smoothly propagate knowledge from data to human wisdom. .

JP 2007-52629 A

The conventional database search method has the following problems in the operation of confirming the state of plant operation.
(1) The operator must specify a search target from a huge number of tags based on his / her own experience (difficulty in the process of “information → knowledge” in FIG. 25).

(2) The operator must manually search operation history, historical data, and the like to acquire information (difficulty in the process of “data → information” in FIG. 25).

(3) The acquired information is the state of trend graphs and history lists arranged in chronological order for each tag. In some cases, statistical calculations such as KPI (Key Performance Indicator) are performed to determine the plant status. It is necessary to perform processing such as performing correlation or correlation with related tags (difficulty in the process of “information → knowledge” in FIG. 25).

(4) When comparing past process operation results, it is necessary to manually search for a comparison target by going back in time in the historical database (difficulty in the process of “information → knowledge” in FIG. 25). .

(5) At present, 4M necessary for process management is mainly managed by human hands. Computers only store information as a database, and no mechanism is provided to support smooth propagation from data to human wisdom.

  The present invention has been made to solve the above-described problems, and collects plant operation data accumulated in a database or the like so that a quick and accurate situation can be grasped, and also supports a display / analysis. The purpose is to realize an operation data management system.

In order to achieve such an object, the configuration of the present invention is as follows.
(1) A state transition diagram creation unit that divides continuous plant operation data in units of lots in batch processing and generates a state transition diagram that defines a state transition of a process in the separation;
A data collection server that acquires the defined state transition diagram and automatically collects plant operation data of the process according to the definition;
A collected data holding unit for holding the collected plant operation data;
A data display unit for editing the stored plant operation data and displaying the data for analysis ;
A plant operation data management system comprising:

( 2 ) The plant operation data management system according to (1 ), wherein the data display unit displays a list of finished lots.

( 3 ) The plant operation data management system according to ( 1) or ( 2) , wherein the data display unit displays all large processes of all lots.

( 4 ) The plant operation data management system according to any one of ( 1) to ( 3) , wherein the data display unit displays details of one process.

( 5 ) Any one of ( 1) to ( 4) , wherein the data display unit displays comparison data for each lot based on past plant operation data accumulated in the collected data holding unit. The plant operation data management system according to the item .

According to the present invention, the following effects can be expected.
(1) The definition of data collection necessary for visualization is only to write a state transition diagram, and it is possible to make a purpose (separation of lots, etc.) for time series data.

(2) At the end event, historical data (visualization information) from the start to the end of the break can be automatically collected online and displayed so that it can be analyzed.

(3) Along with automatic collection of visualization information, by describing calculation logic in the definition of the state transition diagram, statistical calculation such as KPI calculation can be executed simultaneously with data collection.

(4) By accumulating past visualization information in units such as lots, it is possible to perform comparative analysis with past visualization information in units such as lots.

(5) Since information collection is performed via a general-purpose OPC server, it is not necessary to modify the logic of an existing control device that controls the plant equipment.

(6) A change instruction can be issued to an existing control device. That is, by issuing a write command to the OPC server, it is possible not only to display visualization information but also to monitor and control the process.

By displaying the visualization information on the data display unit, the following effects can be expected.
(1) Since it can be confirmed by displaying a list of end lots, it is easy to recognize past operation lots. In the finished lot list display, classification such as abnormality or normal is automatically performed, thereby facilitating the subsequent analysis.

(2) Plant operation data and 4M elements can be displayed by an easy and intuitive operation of dragging from a list display of end lots. By performing process management and display / analysis of the finished lot, it is possible to prevent the occurrence of quality defects thereafter.

(3) By comparing and displaying the best process and the analysis target process, the user can easily perform comparative analysis. Analysis can be performed by dragging the display of the hierarchical process with a mouse operation. In addition, when a difference from the best process is found in a large process, the small process included in the focused large process can be easily displayed, which facilitates analysis.

(4) Since the display according to the display target is performed in the process display, the user can intuitively recognize all the factors, and the analysis becomes easy. Devices such as pumps simultaneously perform device status, device start, and device stop information in addition to process data and manual operation timing, so that necessary information can be viewed immediately.

(5) By making the target of process management not only the end lot but an active lot, process management can be performed during operation. Thereby, it is possible to prevent the occurrence of quality defects.

  Hereinafter, the present invention will be described in detail with reference to the drawings. FIG. 1 is a functional block diagram showing a basic configuration of a plant operation data management system to which the present invention is applied. In the present invention, various operation data of the plant are retrieved and processed to become information, and knowledge is obtained by human understanding (visualization), and finally, to support “propagation of knowledge” that becomes wisdom. , Provide a mechanism from data collection to knowledge creation.

  The plant operation data management system of the present invention includes a process analyzer 100 and a data collection server 200. The data collection server 200 collects real-time plant operation data from the plant equipment 400 via the control device 300 and accesses various databases 501 of the data server 500 to collect plant operation result data.

  The process analyzer 100 includes a state transition diagram generation unit 101, a collected data holding unit 102, and a data display unit 103. The data collection server 200 includes an automatic data collection unit 201 and a data collection logic generation unit 202.

  According to the present invention, data collection for realizing visualization of processes can be executed only by defining a state transition diagram by a builder provided in the state transition diagram generation unit 101. The state transition diagram generation unit 101 delimits continuous plant operation data within a predetermined range, generates a state transition diagram that defines the state transition of the process within the delimiter, and passes it to the data collection server 200.

  The data collection server 200 acquires a defined state transition diagram from the process analyzer 100, constructs a logic for automatically collecting plant operation data of each process in accordance with the definition contents, and builds a logic in the collection logic generation unit 202 to automatically collect the data. Pass to 201.

  The automatic data collection unit 201 accesses the control device 300 and the data server 500 according to the collection logic, collects the plant operation data of each process, and returns it to the process analyzer 100. The collected plant operation data is stored in the collected data holding unit 102, edited, and displayed on the data display unit 103 so that it can be analyzed.

  FIG. 2 is a functional block diagram illustrating an embodiment of the data display unit 103. The data display unit 103 includes data editing means 103 a for inputting plant operation data to be displayed from the collected data holding unit 102. In this embodiment, plant operation data in units of lots in batch processing is an editing target.

  The edited plant operation data is transferred to the end lot list display means 103b, the all lot display means 103c, the process display means 103d, the comparison display means 103e, and the 4M element display means 103f, and is displayed so that it can be analyzed in various display forms. . A specific display form will be described later with reference to FIGS.

  FIG. 3 is a user interface screen example of a state transition diagram generation unit, a data display unit, and a data collection logic generation unit of a plant operation data management system to which the present invention is applied. 101a is a state transition diagram screen, 103b is a process display screen, and 202a is a collection logic screen.

  FIG. 4 is an image diagram illustrating details of the process definition in the state transition diagram. Focusing on one reactor MT (mixing tank), the case where the period from the raw material input to the completion of the product is defined as one segment (referred to as one lot) is illustrated. One lot consists of the following steps (1) to (3). In this example, details of each process of three lots are defined according to the type of product.

(1) Preparation process: Process until the raw material is charged into the reactor (2) Reaction process: Process until the charged raw material is reacted in the tank to produce a product (3) Transfer process: Finished product is taken out The process until it is transferred to the container

  FIG. 5 is an image diagram for explaining the concept of state transition and process. FIG. 5A is a state transition diagram of the MT described in FIG. A state transition diagram is a diagram that expresses how a state changes due to various events. Generally, the state is represented by a circle, and the state transition is represented by an arrow.

  In this example, manufacturing of one lot is completed at the process transition of the charging process SW = 1 → the reaction process SW = 2 → the transfer process SW = 3, and the next process transition at the transfer process SW = 3 → the charging process SW = 1. The state transition to the lot preparation process SW = 1 is shown.

  FIG. 5B is a description example in which the process of the MT state transition diagram of FIG. 5A is defined by a builder provided in the process analyzer 100.

  FIG. 6 is an image diagram for explaining the details of the process definition in the MT state transition diagram, and focuses on the preparation process SW = 1. First, a process start condition and a process end condition are defined, and further details of the process are defined.

  In this example, a KPI calculation is defined as a definition of a tag for data collection and a statistical calculation. When a KPI operation is defined in a process, trend data of KPI is also collected along with data collection of process visualization. In the tag definition, a necessary tag is selected from the tag list of the plant finder provided in the control system and dropped with the mouse to complete the definition.

  FIG. 7 is an image diagram for explaining delimitation for continuous time series trends. A feature of the present invention is that the plant operation data is targeted. By defining a state transition diagram, a continuous process is separated from the beginning of the lot (including abnormally terminated conditions).

  As a result, purpose information such as what is moving under what conditions is added to the graph and event history that are only in time series. That is, plant operation data can be transformed from time-series display to purpose-specific display.

FIG. 8 is an example of a user interface screen of the state transition diagram generation unit, the data display unit, and the data collection logic generation unit when a change command function for the control device is defined in the process analyzer. Data collection server is indicated by the data collection and monitoring server 200 '. The data collection logic generation unit is indicated by a data collection / monitoring logic generation unit 202 ′. The collection logic screen is indicated by a collection logic / monitoring screen 202a ′.

  Since the OPC server (not shown), which is the information provider, has a writing function in addition to the data reading function, the process is visualized and compared with past processes. Thus, it is possible to monitor the state of the current process, and in some cases, send a change instruction to the logic of the control device 300.

  When this mechanism is used, it is possible to incorporate process monitoring logic into a process in which logic for monitoring a process is not incorporated in the control device 300 without changing the logic of the control device. That is, the function can be extended from the visualization of the process to the process monitoring.

  The data collection server 200 automatically collects visualization information online according to the data collection logic generated based on the state transition diagram acquired from the process analyzer 100. Normally, historical data (visualization information) from the start of a break to the end of a break is automatically collected at a break end event such as a lot.

  If an abnormal event is issued from the start to the end of the separation, plant operation data is automatically collected at that timing.

The process analyzer 100 accumulates the past operation history in the collected data holding unit 102 , and can compare the past visualization information with the past visualization information. Furthermore, by defining statistical logic such as KPI calculation for the plant operation data to be collected, it can be calculated and acquired when the data collection server 200 collects data.

  Hereinafter, various display forms of the plant operation data visualized by the data display unit 103 will be described with reference to FIGS. 9 to 22. FIG. 9 is a process display / analysis screen for an end lot by the end lot list display means 103b.

This screen configuration consists of two screens, a lot display section (a) and a process display section (b), and operates as follows.
(1) In the lot display section (a), the lot name of the analysis target lot and the completed lots are all displayed in a tree format.

(2) In the process display unit (b), the process of the end lot and the information related to the process are displayed in a graph, and the lot display unit (c) and the end lot result display unit (d) are configured.

The data display unit 103 has the following functions.
(1) End lot list display (2) All lot display (3) Process display (4) Comparison display (5) 4M element display The configuration and operation of each function will be described below with reference to FIGS.

(1) End lot list display: see FIG. 9 The contents of the end lot result are displayed in a tree form in the lot display section (a). The user selects a tree node to be displayed / analyzed and drags it to the process display part (b) to display the contents.

  When it is desired to display and analyze the contents of a large process, the node tree is selected and dragged to (b) for display. When displaying / analyzing a small process, the node tree is selected and dragged to (b) for display.

  To display and analyze the contents of the small process during the large process display analysis, the display contents are changed by operating the mouse in the small process column displayed in the area (d). In the area (e), not only a single end lot but also a plurality of different end lots are displayed.

  Regions (f) and (g) are examples of parallel display of a plurality of end lots. If (f) is the end lot result of lot 1, (g) is the end lot result of lot 2. The parallel display of multiple lots facilitates comparative analysis between lots.

(2) All lot display: see FIG. 10 In the all lot display, all large processes of all lots are displayed. This display example is a time chart of a large process for all lots for which operation has been completed, and displays the period from the start to the end of each lot and the timing for the next process. This screen is displayed when a lot name node is selected and dragged in the area (a) of FIG.

(3) Process display: See FIGS. 11, 12, and 13. In the large process display of FIG. 11, all large processes of the analysis target lot are displayed. This display example represents a large process from the start to the end of the lot. Here, the elapsed time from the start of each lot to the start of each large process, the period from the start to the end of the large process, and the timing to the next process are displayed. This screen is displayed when a lot completion node is selected and dragged in the area (a) of FIG.

FIG. 12 is a display example of one process. The explanation of (a) to (f) in the figure is shown below.
(A) Process name, process start time, process end time, period, transition condition display to process (b) Elapsed time display from process transition timing to actual process start (c) Actual process start to end period (D) Display period of (c) in character string (e) Process start timing (f) Process end timing

  FIG. 13 is a display example of a small process consisting of 4M. This screen is displayed when a node is selected and dragged in the area (a) of FIG. Here, it is a display when a small process node is selected and dragged. This screen can also be displayed by double-clicking on the large process display one level above.

(4) Comparison display: see FIG. 13 In FIG. 13, by displaying the best process and the process to be analyzed simultaneously, the processes of two lots can be compared and analyzed.

(5) 4M element display: see FIGS. 14 to 22 There are a plurality of 4M element display methods as shown in FIGS.
FIG. 14 is a display example of a timer. This is used for display when driving is continued after a certain period of time. The explanation of (a) to (f) in the figure is shown below.
(A) Display of timer name and timer set time (b) Character string display of elapsed time from process transition timing to timer start (c) Timer start timing (d) Elapsed period display from timer start timing to time up (e ) (D) Character string display in actual period (f) Timer end timing

  FIG. 15 is a display example of pump start and stop timing. (A) shows the pump name, (b) shows the pump start timing, and (c) shows the pump stop. This screen is displayed together with FIGS.

  FIG. 16 is a display example of a pump status signal. (A) displays the pump name and display contents, and (b) and (c) display the pump status in a graph. This screen is displayed together with FIGS. 15 and 17.

FIG. 17 is a display example of a pump command signal. (A) displays the pump name and display contents.
(B) A start pulse graph and (c) a stop pulse graph are displayed. This screen is displayed together with FIGS.

FIG. 18 is a display example of manual operation timing. This is used for display when a manual operation is performed. The explanation of (a) to (f) in the figure is shown below.
(A) Name of manual operation, transition condition to manual operation, manual operation start time, end time, display of actual elapsed time (b) Elapsed period from manual operation transition timing to actual manual operation (C) Actual manual operation timing (d) Elapsed period from manual operation timing to operation completion (e) Character string display of actual period of (d) (f) Manual operation completion timing

FIG. 19 is a display example of determination timing when the subsequent operation is continued based on various conditions. The explanation of (a) to (c) in the figure is shown below.
(A) Display of judgment name (b) Timing of displayed judgment (c) When the mouse cursor is placed at a point where the timing is plotted, the judgment condition is displayed.

FIG. 20 is a display example of raw material switching timing. The explanation of (a) to (c) in the figure is shown below.
(A) Raw material name display (b) Displayed raw material switching timing (c) Raw material details (origin or lot) displayed by placing the mouse cursor on the band

FIG. 21 is a display example of raw material input timing and raw material input order. (A) ~ in the figure
A description of (c) is given below.
(A) Title display (b) Raw material input timing and raw material name are displayed in the graph (c) When the mouse is placed on the timing display section in the graph, the input raw material name and input amount are displayed.

FIG. 22 shows an example in which each person in charge is represented by a band for each color and the person in charge name is displayed in the band. The explanation of (a) to (c) in the figure is shown below.
(A) Title display (b) Person A → Person B change timing (c) Person B → Person C change timing

  The plant operation data to be collected in the present invention includes all information related to plant operation such as operation history information, trouble information, maintenance information, equipment management information, etc., in addition to process data represented by physical quantities.

  In the embodiment, the plant operation data to be collected has been described with respect to lot unit data in batch processing. However, continuous process data can be collected. In this case, the delimitation of data is based on the establishment of specific conditions (temperature, pressure) of physical quantities as delimitation start and end conditions.

  In the embodiment, the target of process management has been described as the end lot, but it is possible to perform process management during operation by setting an active lot as a target. Thereby, it is possible to prevent the occurrence of quality defects.

It is a functional block diagram which shows the basic composition of the plant operation data management system to which this invention is applied. It is a functional block diagram which shows embodiment of a data display part. It is an example of a user interface screen of a state transition diagram generation unit, a data display unit, and a data collection logic generation unit of a plant operation data management system to which the present invention is applied. It is an image figure explaining the detail of the process definition in a state transition diagram. It is an image figure explaining the concept of a state transition and a process. It is an image figure explaining the detail of the process definition in a state transition diagram. It is an image figure explaining the division | segmentation with respect to a continuous time series trend. It is an example of a user interface screen of a state transition diagram generation unit, a data display unit, and a data collection logic generation unit when a change command function for a control device is defined. It is a process display / analysis screen. This is the entire lot display screen. It is a large process display screen. It is a process zone display image. It is a small process display screen. It is a timer display screen. It is a start / stop display screen. It is a status display screen. It is a command signal screen. It is a manual operation screen. It is a judgment screen. It is a raw material screen. It is a raw material input order and timing screen. This is the person in charge screen. It is an image figure of the conventional plant operation data search. It is an example of a display screen of searched plant operation data. It is an image figure of knowledge making from data to wisdom.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Process analyzer 101 State transition diagram production | generation part 102 Collected data holding part 103 Data display part 200 Data collection server 201 Data automatic collection part 202 Data collection logic production | generation part 300 Control apparatus 400 Plant equipment 500 Data server 501 Various databases

Claims (5)

A state transition diagram creation unit that generates continuous state operation diagram that defines the state transition of the process in the division, and the continuous plant operation data is divided in lot units in batch processing,
A data collection server that acquires the defined state transition diagram and automatically collects plant operation data of the process according to the definition;
A collected data holding unit for holding the collected plant operation data;
A data display unit for editing the stored plant operation data and displaying the data for analysis ;
A plant operation data management system comprising: The plant operation data management system according to claim 1, wherein the data display unit displays a list of completed lots. The plant operation data management system according to claim 1, wherein the data display unit displays all large processes of all lots. The said data display part displays the detail of one process, The plant operation data management system as described in any one of the Claims 1 thru | or 3 characterized by the above-mentioned. The said data display part displays the comparison data for every lot based on the past plant operation data accumulate | stored in the said collection data holding part, The plant as described in any one of Claim 1 thru | or 4 characterized by the above-mentioned. Operation data management system.