RU2801735C1 - Control method and control system for production plants - Google Patents

Abstract

FIELD: production control systems.

SUBSTANCE: control method and a control system for multiple production plants of the same type. The method for managing production plants located at multiple production sites comprises the step of preparing information data for aggregating the operating data of each of the production units for each of the production sites, the stage of accumulating operating data, the stage of analysing the operating conditions of each of the production units, the stage of displaying information about current conditions operating each of the production plants on a display device located in a control section other than the production sites, a production plant control step.

EFFECT: accurate detection of operational abnormalities of production plants, without using a large amount of labour force and quick response to operational abnormalities.

4 cl, 3 dwg

Description

The field of technology to which the invention belongs

The present invention relates to a control method and a control system for production plants for controlling multiple production plants of the same type located in multiple production sites.

The field of technology to which the invention belongs

In the manufacturing industry, where a large number of large-scale production plants are required to manufacture products, consolidating production plants in one place will require a huge area and entail the risk of a complete shutdown in the event of a major accident such as an earthquake, so production sites are often distributed throughout the country and sometimes abroad. In this case, a production plan is drawn up at each production site, and the production plant is operated in accordance with the production plan.

In the iron and steel industry, steel mills are often located in multiple locations. Among steel plants, in integrated steel mills in which production units are combined for upstream processes and downstream processes, production units such as a blast furnace for producing an iron source, a converter for adjusting the components of molten iron produced in a blast furnace, a continuous casting plant for solidifying molten iron into slabs and a hot rolling plant for drawing slabs into thinner sheets are often installed together in steel mills. Until now, the determination of abnormality in the operation of these production units was carried out for each production unit at each production site, and at each production unit, the abnormality was eliminated individually.

As a rule, the production plant is operated by operators of the operation, checking the operating data. When it is determined that the operation is abnormal, the operating conditions are changed in order to prevent the development of the abnormality, thereby stabilizing the operation of the plant. The technology for detecting abnormality in operation of blast furnaces is described, for example, in patent literature 1. In particular, the value of the Q statistic, which is the second and subsequent components obtained by analyzing the principal components of data from pressure gauges installed in the shaft of the blast furnace vessel, increases with change in pressure in the mine in violation of the ventilation of the blast furnace. According to the technology described in Patent Literature 1, a threshold value is set for the Q statistic under normal conditions collected in advance, and when the Q statistic exceeds the threshold value, it is determined that an operating abnormality has occurred in the blast furnace.

List of cited literature

Patent Literature

Patent Literature 1: Japanese Patent Application Laid-Open No. 2017-128805

The essence of the invention

Technical task

In recent years, the number of operations operators has been declining against the background of a reduction in the workforce due to business optimization and a reduction in the local working population. In fact, when an abnormality occurs in the operation, it is necessary to temporarily stop the operation, and then immediately take the necessary actions to eliminate the malfunction. However, with fewer operators, most of the work associated with the operation falls to a few operators. In addition, when operating in this manner with fewer field operators, operational expertise in the event of anomalies does not transfer smoothly, and continuing to work without the awareness of an abnormal situation in production plants can lead to even more trouble, resulting in long downtime and high costs. for recovery. Moreover, this situation complicates the production of products and leads to a decrease in sales of products. In light of the foregoing, abnormality detection technologies are being developed for various processes, which have been described above. However, it is important that these anomaly detection technologies work reliably and are reflected in operation in order to prevent serious installation problems.

The present invention has been made in view of the above problem, and an object of the present invention is to provide a control method and a control system for manufacturing plants that accurately detects operational abnormalities of manufacturing plants without requiring a large amount of manpower and allowing prompt response to operational abnormalities.

The solution of the problem

The control method of the present invention for manufacturing plants for controlling multiple same-type manufacturing plants located in multiple manufacturing sites includes: preparing information data for aggregating operating data of each of the manufacturing plants for each of the manufacturing sites; a data accumulation step for accumulating operational data aggregated in the information data preparation step in a computer located in the data accumulation section; a data analysis step for analyzing the operating conditions of each of the production units using the operating data accumulated in the data accumulation step; a data display step for displaying information about the operating conditions of each of the production plants analyzed in the data analysis step on a display unit located in each of the production sites; and a plant control step for accessing the information displayed in the data display step and controlling the production plant located in the second production site from the first production site.

The step of displaying data may include the step of displaying information about the current operating conditions of each of the production plants on a display device located in a control area different from the production areas, and the step of controlling the plant may include a step of controlling a production plant located in a second production area. , from the control site, not from the first production site.

The production plant control method may include, instead of the production plant control step, an information presentation step for presenting, from the first production site or the control site, to the operation operator of the production plant located in the second production site, the operation control information of the production plant.

The production plant control method may include a step of determining control conditions for determining whether the operation is abnormal in each of the production plants based on the analysis results of the data analysis step, and the production plant control step or the information presentation step may be performed when it is determined that at least one of the production units has an operational abnormality.

The step of determining the operating conditions may include the step of determining whether the operation is abnormal using the performance indicator, operating data and at least one part of the operating conditions metadata converted to the same scale.

The control system according to the present invention for production plants for controlling multiple production plants of the same type located in multiple production sites includes: an information data preparation unit configured to aggregate operating data of each of the production plants for each of the production sites; a data accumulation unit configured to accumulate operational data aggregated by the information data preparation unit; a data analysis unit configured to analyze the operating conditions of each of the production units using operating data accumulated by the data accumulation unit; a display unit located at each of the production sites and configured to display information about the operating conditions of each of the production units analyzed by the data analysis unit; and a plant control unit configured to access the information displayed on the display unit and control the production plant located in the second production section from the first production section.

Useful effects of the invention

The control method and control system for production plants according to the present invention can accurately detect operating abnormalities in production plants without requiring a large amount of manpower and allowing quick response to operating abnormalities.

Brief description of the drawings

Fig. 1 is a block diagram illustrating the configuration of a control system for manufacturing plants according to an embodiment of the present invention.

Fig. 2 is a block diagram illustrating the configuration of the modification of the control system for production plants shown in FIG. 1.

Fig. 3 is a flowchart illustrating a plant control process according to an embodiment of the present invention.

Detailed description of the invention

Steel mills mainly include upstream process plants that produce slabs that serve as raw material for products, downstream process plants that produce end products, and power plants that supply electricity, fuel gas, water, etc. .P. Upstream process plants include production plants such as a raw material warehouse for storing iron ore raw materials and other materials, a coke oven for producing coke from coal, a sinter plant for fine iron ore sintering, a blast furnace for producing liquid iron by burning these raw materials. materials at high temperatures, an iron carrier for transporting molten iron and a railway vehicle for transporting molten iron, a pretreatment plant for adjusting molten iron components, a converter for producing liquid steel by adjusting the amount of carbon in molten iron, a secondary refining furnace, and a continuous casting machine for solidification molten steel into finished slabs.

The downstream process plants include manufacturing plants such as a hot rolling plant for producing thin steel strips by heating slabs, a cold rolling plant for producing thin elongated steel sheets with desired strength by repeatedly heating and cooling steel strips, a cold rolling plant for cold-rolled steel sheet coating, a pipe manufacturing plant by bending hot-rolled steel sheets and welding the ends of the sheets together, and a molding plant to produce a line of products called building material molds.

Power plants include production plants such as blast furnace and coke gas purification plant, purified gas power generation plant, gas pipelines for transporting gas to the power generation plant, and water pipes for transporting water for cooling various furnaces. and heated items.

Since there are so many manufacturing plants, and each of them requires a large installation area, integrated steel mills, including upstream processing plants and downstream processing plants, require large areas. The volume of production of steel products in one steel mill does not satisfy the demand, so it is necessary to produce steel products in several steel mills. In such a case, steel mills are sometimes located in the vicinity of customers who require a large volume of steel products to facilitate the delivery of products to customers of steel products, or steel products are produced in steel mills located in multiple production sites to avoid shutdowns due to large-scale accidents.

Each production plant in a steel mill is equipped with a number of sensors that measure the current parameters of the production plant and sensors that measure the condition of the steel product being produced. In daily work, operation operators set the necessary conditions for the production of steel products based on the data measured by these sensors, and perform the work. The following description is an example of the operation of a blast furnace and upstream process units.

Generally, in blast furnace operation, the opening degree of the top hopper is controlled so that raw materials such as iron ore, sintered ore, and coke are loaded from the top of the furnace at an appropriate loading rate, and the inclination and rotation speed of the loading device is set to achieve the intended load distribution. . Raw materials temporarily stored in the upper hopper at the top of the furnace are weighed in succession to control the loading volume. The surface profile of the loaded raw materials is measured by a microwave profilometer, and the correct distribution of the loaded raw materials is determined. Hot air is blown into the blast furnace from a plurality of tuyeres arranged circumferentially in the bottom of the furnace. The blown hot air enters the top of the furnace while exchanging heat with the descending raw materials. When the feed raw materials are distributed to their destination, the data from the many pressure gauges installed in the furnace body indicate almost the same behavior. The temperature and components of the gas reaching the top of the furnace are measured by a probe installed at the top of the furnace to confirm that the gas flow in the furnace is appropriate and the raw materials react chemically to form molten iron.

Set the flow rate, temperature, humidity, enriched oxygen content, etc. hot air injected from the lances and the discharge pressure is measured. In addition, coal dust is blown from the lance installed in the lance, and the blowing amount of coal dust is set. The tuyeres are equipped with cameras to monitor the state of the furnace for successive image capture in the tuyeres. The temperature of the tuyeres is also measured. In addition, a series of thermometers to control the temperature of the furnace body are attached for sequential temperature measurement. In addition, the quantity and temperature of the cooling water supplied to the cooling panel to cool the furnace body, as well as the temperature of the cooling panel, are measured. The resulting liquid iron is discharged to the outside through a tap-hole at the bottom of the furnace along with the slag generated at the same time in the furnace. The start time of release from the tap hole and the end time of release are also recorded. Liquid pig iron and slag coming out of the taphole are separated in a metal chute according to the difference in specific gravity, and liquid pig iron is poured into a pig iron carrier.

The slag enters the slag treatment plant, is cooled by water or air, and placed in a storage area. Immediately after the molten iron is discharged from the blast furnace, the temperature is measured and the amount of silicon contained in the molten iron is also measured. In addition, the weight of the molten iron entering the iron carrier is measured by a load cell. Slag basicity and slag consumption are also measured. In the case of installations for the production of the final product, production conditions such as the operating speed of each production line, the temperature of the oven, the heating time of the product and the mass of the hardening coating, the product defects and the shape obtained by the sensors, and any other product performance evaluation data are considered part of the set. working data. The product in a blast furnace can be considered as molten pig iron or slag.

These operating conditions, sensor data, and product information are aggregated in a control computer (called process computer) that can be controlled by an operating operator, and recorded with a single time stamp (with time information). In the process computer of the blast furnace, coefficients such as the permeability coefficient and the gas utilization coefficient, which indicate the state of the blast furnace, are calculated based on the recorded data and recorded along with sensor data associated with each plant, molten iron and slag measurement data, and operating conditions. Blast furnaces in remote areas operate in essentially the same way and produce similar sets of operating data.

Next, with reference to the drawings, a control system for production plants according to an embodiment of the present invention will be described in which the present invention is applied to a process for controlling a plurality of blast furnaces located in a plurality of production sites. In the present embodiment, the present invention is applied to a process for controlling a plurality of blast furnaces located at multiple production sites. However, the range of application of the present invention is not limited to the present embodiment, and the present invention is applicable to a wide range of general processes for controlling and operating numerous similar production plants located in multiple production sites.

Configuration

First, with reference to FIG. 1 and FIG. 2, the configuration of a control system for production plants according to an embodiment of the present invention will be described. In FIG. 1 is a block diagram illustrating the configuration of a control system for manufacturing plants according to an embodiment of the present invention. In FIG. 2 is a block diagram illustrating the configuration of a modification of the control system for production plants shown in FIG. 1.

As shown in FIG. 1, a control system 1 for production plants as an embodiment of the present invention is a system for controlling a plurality of blast furnaces (from blast furnace A to blast furnace X) located in multiple production sites. The system includes 2 blast furnace sensors (from blast furnace sensor A to blast furnace sensor X) installed on individual blast furnaces, and the measurement data of each indicate the state of the blast furnace and the molten iron produced in the blast furnace, process computers 3 for individual blast furnaces of production sites, each of which is electrically connected to the corresponding sensor 2 of the blast furnace, edge server computers 4 for individual production sites, each of which is electrically connected to the corresponding process computer 3, and a global data server computer 5 connected to the edge server computers computers 4 through the telecommunications network.

In this document, the technology computers 3, the edge server computers 4, and the global data server computer 5 are configured with well-known information processing devices. The global data server computer 5 is located at any production site or at a location other than the production sites. However, the location of the global data server computer 5 (data accumulation area) is not physically limited to one location, and may be divided into multiple locations. As shown in FIG. 2, devices such as a dedicated sensor data processing computer 6a, a programmable logic controller (PLC) 6b, and a distributed control system (DCS) 6c may be connected to the blast furnace sensors 2, and the blast furnace sensors 2 may be electrically connected to the boundary server computer 4 through these devices.

The production plant control system 1 in this configuration performs the plant monitoring process described below, and thus accurately detects blast furnace operating abnormalities without requiring a large amount of manpower, and enables prompt response to blast furnace operating abnormalities. The operation of the plant control system 1 in executing the plant control process will be described below with reference to the flowchart shown in FIG. 3.

Installation control process

In FIG. 3 is a flowchart illustrating an installation control process according to an embodiment of the present invention. The plant control process shown in FIG. 3 starts at the point in time when operation of the blast furnace is started and the plant control process proceeds to step S1.

In the process in step S1, the process computer 3 aggregates the work data sets and transmits the aggregated data sets to the edge server computer 4. The edge server computer 4 stores the work data sets transmitted from the process computer 3 in a storage device such as a hard disk, an optical disk, or USB drive. As used herein, the term "operational data set" refers to operational blast furnace data and metadata acquired at the same time. Operating data means data measured by the blast furnace sensor 2 and/or various coefficients calculated based on the data measured by the blast furnace sensor 2. Metadata means data about the operating conditions of the blast furnace at the time of receipt of the operating data, setting values, operating conditions and operating data other than the operating data of interest. Preferably, the number of elements, element names, and units of operational data and metadata are standardized so that they are identical across multiple manufacturing sites.

Technological computer 3 is heavily loaded, as it performs model calculations, controls various devices, and also collects and processes data. In addition, if the production data sets are transmitted to the global data server computer 5 via the telecommunications network, a waiting time may occur depending on the overload condition of the telecommunications network, which may affect the operation of the blast furnace. Thus, in the present embodiment, the edge server computer 4 is installed at a close distance in the telecommunications network, and the process computer 3 is connected to the edge server computer 4 via the telecommunications network. The work data sets are temporarily stored on the edge server computer 4, and the global data server computer 5 accumulates the work data sets stored on the edge server computer 4 via a telecommunications network as described below. When the control system 1 for production plants is configured as shown in FIG. 2, the edge server computer 4 also stores data transferred from devices such as the dedicated sensor data processing computer 6a, PLC 6b and DCS 6c. Then, the process in step S1 ends and the installation control process proceeds to the process in step S2.

In the process in step S2, the global data server computer 5 reads an electronic file containing operational data sets stored on each edge server computer 4 via the telecommunications network at predetermined time intervals (for example, every minute when changes are monitored in each sensor in the blast furnace ). The edge server computer 4 may transmit the electronic file containing the operational data sets to the global data server computer 5 at predetermined time intervals. Then, the process in step S2 ends and the installation control process proceeds to the process in step S3.

In the process in step S3, the global data server computer 5 stores the operating data sets of each blast furnace read in the process in step S2 in a storage data recorder such as a hard disk, an optical disk, or a USB memory. In this case, the global data server computer 5 synchronizes the operating data sets of each blast furnace in time so that the operating data sets of each blast furnace can be compared with each other at the same time. When the names and units of data items in the operational data sets differ, the global data server computer 5 converts them to predefined names and units to standardize the names and units. Then, the process in step S3 ends, and the installation control process proceeds to the process in step S4.

In the process in step S4, the global data server computer 5 analyzes the operating state of the blast furnace by analyzing the operating data sets of the blast furnace stored in the storage data recorder. In particular, the operating state of each blast furnace can be known from the permeability coefficient calculated from pressure data, the Q statistics calculated from pressure data, the gas utilization rate calculated from gas analysis data, and the dispersion in the perimeter direction of the furnace statistics Q, calculated from the brightness of image data in a plurality of lances in a blast furnace. The operating condition of a blast furnace can also be determined from the permeability obtained by dividing the difference between the pressure in the shaft and the pressure at the top of the furnace by the volume of the furnace. The gas utilization ratio, indicating the ratio of CO to CO ₂ components obtained by analyzing the gas components obtained from the top of the furnace, is another indication of the operating condition of the blast furnace. Based on the bottom of the blast furnace, it is possible to calculate the thermal coefficient of the furnace reflecting the thermal state in the blast furnace based on the calculation of the heat balance in the bottom of the blast furnace, and use it to determine the state of reactions in the blast furnace before reaching the temperature of liquid iron. The operating data sets include various other data from which the operating conditions of the blast furnace can be known, and the operating conditions of the blast furnace can also be determined from the moving average of such data, considered as time series data, the transition of the standard deviation from the average calculated in advance, and adding and subtracting multiple data.

The global data server computer 5 calculates a coefficient indicative of the operating conditions of each blast furnace by performing the above-described analysis of the operating data sets of each blast furnace. Preferably, after completion of the analysis, the global data server computer 5 stores a coefficient indicating the operating conditions of each blast furnace along with the operating data sets in the accumulated data recorder. This process in step S4 may be performed by the edge server computer 4, and the edge server computer 4 may include a coefficient indicative of the operating conditions of the blast furnace in the operating data sets. Then, the process in step S4 ends, and the installation control process proceeds to the process in step S5.

In the process, in step S5, the global data server computer 5 displays a factor indicating the operating conditions of each blast furnace and sets of operating data on a data display unit such as a liquid crystal display located in each blast furnace, for example, so that temporary changes. Preferably, the global data server computer 5 simultaneously displays the coefficient indicative of the operating conditions of each blast furnace and the operating data sets, so that the coefficient indicative of the operating conditions of each blast furnace and the operating data sets can be compared. In addition, the period for displaying data can be set as desired so that long-term trend and short-term trend can be referred to, thereby making it easier to understand the change in operating conditions of each blast furnace. This process makes it easier to determine whether the operating conditions of the blast furnace are satisfactory, and also facilitates the assessment of how the operating conditions of the blast furnace will change in the future. Then, the process in step S5 ends, and the installation control process proceeds to the process in step S6.

In the process, in step S6, the global data server computer 5 determines whether the operating conditions of each blast furnace are satisfactory, for example, based on a factor indicating the operating conditions of each blast furnace. In doing so, the global data server computer 5 can also determine where in the blast furnace and what type of anomaly occurs. Particularly in a blast furnace, high permeability in the furnace is important for continuous stable operation. However, when the gas flow in the furnace is disturbed for any reason, abnormalities such as gas stuck, slippage, and gas canalization may occur. "Hanging" is a phenomenon in which the raw materials do not descend normally and the descending stops. When the "hang" is removed, slippage occurs, or hot gas blown from the tuyeres at the bottom of the furnace is suddenly channeled into the top of the furnace for some reason. A number of shaft pressure gauges are installed in the blast furnace body. The Q statistic, obtained by analyzing the principal components of data from shaft gauges, is an indication of the degree of deviation from the distribution that pressure data can have under normal operating conditions. Thus, an operational abnormality can be determined by setting a threshold value for the Q statistic. When it is determined that the operation is abnormal, the direction in which the pressure perturbation occurred can be known by referring to all mine pressure data. In this document, the principal component analysis, which is a prerequisite for calculating the Q statistic, models the data from data normalized to the mean and standard deviation of the data. Thus, the determination of the Q statistic is not unique to each facility, and the determination of abnormality can be made on a scale common to facilities.

The permeability coefficient is an indicator for determining the state of ventilation in a blast furnace. The permeability coefficient is calculated by dividing the difference between the hot air injection pressure into the tuyeres and the pressure at the top of the furnace by the volume of the furnace. A threshold value is also set for the permeability coefficient so that operational anomalies can be identified. Also, instead of the injection pressure at the tuyeres, the measurement values of the pressure sensors at the top of the furnace, the middle part, and the bottom of the shaft can be used to analyze various areas, including the top, the intermediate part. a part and a lower part, which can be used to determine in which part an abnormal ventilation has occurred.

However, production plants in other production areas often have different characteristics. In such a case, if the operating conditions are judged using the same coefficient, it may be difficult to immediately determine whether an abnormality exists. Thus, coefficients obtained from production plants of the same type with different characteristics are evaluated on the same scale, whereby the possibility of erroneous determination can be reduced. In the case of blast furnaces, blast furnaces vary in volume and shape, so pressure gauges and thermometers also vary in location and quantity. Thus, the resulting ratio and fluctuation range may vary depending on the production plant.

The coefficient obtained from the operating data for determining the operating conditions of the plant is determined as the first coefficient, and the average value of the first coefficient obtained for the period in which the plant was supposed to be in a normal production state in the past is calculated. The first coefficient obtained in succession is divided by this average value to obtain a second coefficient, by which the conditions of plants of different production plants of the same type can be easily compared and abnormalities can be instantly identified. The period of data used to calculate the average can range from one month to several months, depending on the type of production plant. If the temperature of the atmosphere has some influence, taking into account the factors of seasonal fluctuations, it is possible to calculate the average value of the first coefficient for about a year. Such conversion to the same scale can be done by statistical normalization, that is, by dividing the value obtained by subtracting the mean from the operating data by the standard deviation. In the present description, such a process is called the conversion of indicators and/or operating data to the same scale. The permeability coefficient and the values monitored by pressure gauges and other gauges can be converted to the same scale.

It is preferable that the temperature of the molten iron exiting the blast furnace is constant, but if an abnormality occurs during operation, the temperature of the molten iron may drop excessively, thereby preventing the discharge of the molten iron and slag from the tap hole. This problem is called kiln cooling failure, which takes a long time to recover and delays production, resulting in a significant drop in production. In order to prevent such accidents, operational anomalies can be determined by a factor called the thermal coefficient TQ of the furnace. The thermal coefficient TQ of the furnace can be calculated from the equation (1) below. Then, a threshold value for the thermal coefficient TQ of the furnace is set in the same manner as for the permeability coefficient, and when the thermal coefficient TQ of the furnace falls below the threshold value, it is determined that the operation is abnormal.

TQ = (Q1 + Q2) - (Q3 + Q4 + Q5 + Q6) + (Q7 - Q8 - Q9) ... (1)

In this document, Q1 is the sensible heat of the air stream, Q2 is the heat of combustion of carbon on the lance, Q3 is the heat of decomposition of the moisture in the air stream, Q4 is the heat of dissolution reaction, Q5 is heat loss, Q6 is the heat of decomposition of pulverized coal (PC), Q7– the heat carried by the coke and the molten product, Q8 is the heat carried by the produced gas, and Q9 is the heat carried by the coke. Since the furnace thermal coefficient is also affected by design differences in the furnace shell of a blast furnace, it is possible to use the furnace thermal coefficient converted to the same scale so that the overall furnace thermal coefficient for installations can be compared in the same way as the permeability coefficient.

The global data server computer 5 prepares a plurality of such outlier algorithms and successively evaluates the operational data sets received from the same blast furnace. The global data server computer 5 can extract the period during which operating conditions have changed in the past, similar to changes over a certain period of time under current operating conditions, by comparing the metadata obtained during the current operation with past metadata, including metadata obtained at other production sites, and can determine whether the operation of the production plant is abnormal based on the operating conditions of the production plant in the extracted period. Examples of the metadata include operating conditions, including charge distribution, air flow, enriched oxygen content, pulverized coal flow, coke ratio, moisture content of the air stream, air stream temperature and air stream pressure, and manipulation conditions. Among these pieces of metadata, those with different scales for different institutions are stored with data converted to the same scale so that they can be used to diagnose abnormalities even for different settings.

The determination of operating conditions may include human intervention. In such a case, the input device is prepared to allow people to input the operational abnormality information when it is determined that the operation is abnormal. Thus, when it is determined that the operation is abnormal, a label indicating abnormality can be recorded along with the timestamp when synchronized with the time series data. In this case, different labels can be used to record different abnormal conditions. When the load on the global data server computer 5 in the abnormality determination process is large, the abnormality determination computer directly connected to the global data server computer 5 can be prepared so that the abnormality determination computer can execute the abnormality determination process. In addition, the global data server computer 5 can build a training model trained using as training data the operating data sets obtained when the operating conditions of the blast furnace are normal in the process in step S6, where the operating data set is an input value, and a blast furnace operating condition determination value is an output value. The global data server computer 5 can determine the operating conditions of the blast furnace based on the determination value of the operating conditions of the blast furnace outputted by the learning model to input the current operating data set.

A machine learning model for diagnosing anomalies can be built using a human-entered anomaly label and data obtained by converting the performance index (permeability coefficient, kiln thermal coefficient, etc.), operating data, and at least one piece of operating conditions metadata on each installation on a single scale and can be used to determine the operating conditions of a blast furnace. Operating data unique to each individual plant is standardized by a factor indicating work and then converted to the same scale, whereby a model can be formed with a huge amount of data using data from all plants. Thus, a machine learning model for abnormality detection can be generated with all data, even if the abnormality occurrence rate is low in each individual installation. Then, the process in step S6 ends, and the installation control process proceeds to the process in step S7.

In the process, in step S7, the global data server computer 5 provides the operating operator of each blast furnace with information about the blast furnace with an operating abnormality, the type of abnormality that has occurred, and the location of the abnormality. For example, anomalies detected by Q statistics and permeability are presumably related to the properties of raw materials, and anomalies detected by the kiln thermal coefficient are presumably caused by impaired slag drainage. As an example of an abnormality reporting means to an operating operator, a screen showing information about the abnormality can be prepared on a dedicated personal computer, tablet terminal or the like connected to one abnormality, then the abnormality can be displayed on the screen. Alternatively, an email notification can be sent to a mobile phone or smartphone belonging to the operating operator. The global data server computer 5 may notify the operating operator of each blast furnace to determine whether the operating conditions are being maintained or are deteriorating. Then, the process in step S7 ends, and the installation control process proceeds to the process in step S8.

In the process at step S8, the operator of the blast furnace in which no operational abnormality is detected communicates with the operator of the blast furnace in which the operational abnormality is detected, and the operator of the blast furnace in which the operational abnormality is detected transfers the right to operate the blast furnace furnace to a blast furnace operator that does not have an operational abnormality. When blast furnaces are operated at multiple sites, some sites are likely to have highly trained operators. A highly qualified operation operator assesses the situation based on various information and contacts the operation operator of the blast furnace that has an operating abnormality. If the level of urgency is low, the operator of the blast furnace operation found to have an operating abnormality follows the advice of a highly qualified control operator to operate the blast furnace. On the other hand, if the urgency level is high, the right to operate the blast furnace in the production area with the abnormality is transferred to a highly skilled operator, and the highly qualified operator operates the blast furnace and takes measures to eliminate the abnormality.

The above explanation assumes that a highly skilled operations operator is present at another production site. However, the command and control center (for example, the central office of a steel mill) can be set up separately from the production sites that have blast furnaces, and various information about blast furnaces can be displayed in the command and control center. Then a highly skilled operation operator, located in the command and control center, can eliminate abnormalities in the manner described above. When reviewing the displayed information and determining that some action needs to be taken, the highly skilled operations operator does not have to wait for a determination that an operational abnormality has occurred in the blast furnace and can contact the operation operator to take action, including the process of preventing abnormalities. Then, the process ends at step S8, and the installation control process ends.

Although the embodiment to which the invention made by the inventors is applied has been described above, the present invention is not intended to be limited to the description and drawings that form part of the disclosure according to the present embodiment. In other words, all other embodiments, examples, and control methods made by those skilled in the art based on the present embodiment are within the scope of the present invention.

Industrial Applicability

The present invention provides a control method and a control system for manufacturing plants that accurately detects operating abnormalities of manufacturing plants without requiring a large amount of manpower and allowing a quick response to operating abnormalities.

List of reference positions

1 - control system for production plants

2 – blast furnace sensor

3 - technological computer

4 - edge server

5 – global data server.

Claims (18)

1. A method for controlling production plants, designed to control numerous production plants of the same type located at numerous production sites, containing: step of preparing information data for aggregating the operating data of each of the production units for each of the production sites; a data accumulation step for accumulating operational data aggregated in the information data preparation step in a computer located in the data accumulation section; a data analysis step for analyzing the operating conditions of each of the production units using the operating data accumulated in the data accumulation step; a data display step for displaying information about the operating conditions of each of the production plants analyzed in the data analysis step on a display unit located in each of the production sections, the data display step including a step for displaying information about the current operating conditions of each of the production plants on a display device located in a control area other than production areas; And a production plant control step for accessing information displayed in the data display and control step of the production plant located in the second production station from the first production station, in which, when the data display step includes the information display step, the step of controlling the production plant includes the step of operating the production plant located in the second production site from the control site and not from the first production site, or the control method comprises, instead of the production plant control step, the step of presenting the production plant control guide information from the first production site or the control section to the production plant control operator located in the second production site. 2. The production plant control method of claim 1, further comprising an operating condition determination step of determining whether operation is abnormal in each of the production units based on the analysis result of the data analysis step, wherein the production plant control step or the information presentation step is performed when it is determined that at least one of the production plants has an operating abnormality. 3. The production plant control method according to claim 2, wherein the step of determining operating conditions includes the step of determining whether the operation is abnormal using the performance indicator, operating data, and at least one part of operating condition metadata converted into one volume. same scale. 4. A production plant control system designed to control numerous production units of the same type located at numerous production sites, comprising: an information data preparation unit configured to aggregate operating data of each of the production units for each of the production sites; a data accumulation unit configured to accumulate operational data aggregated by the information data preparation unit; a data analysis unit configured to analyze the operating conditions of each of the production units using operating data accumulated by the data accumulation unit; a display unit located in each of the production sites and configured to display information about the operating conditions of each of the production plants analyzed by the data analysis unit; And a plant control unit configured to access information displayed on the display unit and control the production plant located in the second production area from the first production area.

Images