CN109996615B - Abnormality diagnosis method and apparatus for rolling equipment - Google Patents
Abnormality diagnosis method and apparatus for rolling equipment Download PDFInfo
- Publication number
- CN109996615B CN109996615B CN201680091146.9A CN201680091146A CN109996615B CN 109996615 B CN109996615 B CN 109996615B CN 201680091146 A CN201680091146 A CN 201680091146A CN 109996615 B CN109996615 B CN 109996615B
- Authority
- CN
- China
- Prior art keywords
- data
- rolling
- determination
- result
- normal range
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B21—MECHANICAL METAL-WORKING WITHOUT ESSENTIALLY REMOVING MATERIAL; PUNCHING METAL
- B21B—ROLLING OF METAL
- B21B38/00—Methods or devices for measuring, detecting or monitoring specially adapted for metal-rolling mills, e.g. position detection, inspection of the product
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B23/00—Testing or monitoring of control systems or parts thereof
- G05B23/02—Electric testing or monitoring
Abstract
The invention provides a method and a device for diagnosing the abnormality of a rolling facility. The abnormality diagnosis device for a rolling mill of the present invention is connected to a data collection device that collects and records process data of the rolling mill in time series, and diagnoses an abnormality of the rolling mill based on the data recorded by the data collection device. The abnormality diagnosis device includes at least a data extraction unit, a determination unit, and a determination result evaluation unit. The data extraction unit extracts data corresponding to the same rolled product from the data recorded by the data collection device. The judgment unit judges whether or not the extracted data is within a normal range defined by the normal data group stored in the 1 st database. The determination result evaluation unit evaluates the determination result of the determination unit based on the rolling result of the rolled product corresponding to the extracted data. Specifically, when the determination result of the determination unit does not match the rolling result, the determination result evaluation unit changes the determination criterion defining the normal range.
Description
Technical Field
The present invention relates to a method and an apparatus for diagnosing an abnormality in a rolling mill for rolling a metal material to produce a rolled product.
Background
In recent years, the specifications required of rolled products by customers have become more and more strict. In particular, it is important to satisfy the specifications required by customers by making mechanical properties such as strength and ductility fall within allowable ranges in addition to the dimensions and shapes of rolled products. However, in hot rolling, there is no direct control of mechanical properties. Since the mechanical properties are closely related to the temperature history during rolling, the mechanical properties are currently indirectly managed by temperature information during rolling.
The dimensional shape and temperature of the rolled product are controlled by automated equipment to maintain the required accuracy. However, in actual circumstances, the dimensional shape and temperature of the rolled product are greatly affected by the maintenance state of the equipment. In particular, if the facility maintenance staff is short of the staff, the quality of the rolled product may be adversely affected by a delay in maintenance. Therefore, it is desirable to construct a structure for automatically determining an abnormal state by monitoring process data during rolling. For example, the following is proposed in japanese patent No. 5158018: the feature quantity of the apparatus is detected from the time-series data of the apparatus, and an abnormal state is judged based on whether or not the feature quantity is similar compared with the past abnormal phenomenon.
In the method described in the above publication, it is necessary to accumulate the past abnormal phenomena in advance for accurate diagnosis. However, there is a limit to collecting the abnormal phenomenon and its feature amount in advance. When taking a hot rolling line as an example, the number of signals from the equipment and sensors reaches several tens of thousands. On the other hand, the frequency of occurrence of the abnormal phenomenon is not so high. Therefore, a lot of labor is required to collect the abnormal phenomenon and the feature amount thereof. Furthermore, since it is necessary to define an abnormal phenomenon in order to collect the abnormal phenomenon and its feature value, it is impossible to cope with an unknown abnormal phenomenon.
Documents of the prior art
Patent document
Patent document 1: japanese patent No. 5158018
Disclosure of Invention
Problems to be solved by the invention
The present invention has been made in view of the above-described problems, and provides a method and an apparatus for diagnosing an abnormality of a rolling apparatus, which do not require the accumulation of past abnormal phenomenon data for highly accurate abnormality diagnosis, and which can cope with unknown abnormalities.
Means for solving the problems
The abnormality diagnosis method for a rolling mill according to the present invention is a method for collecting and recording process data of a rolling mill in time series by a data collection device, and diagnosing an abnormality of the rolling mill based on the data recorded by the data collection device, and includes at least the following three steps.
The 1 st step is a step of extracting data corresponding to the same rolled product from the data recorded by the data collecting device. The 2 nd step is a step of determining whether or not the extracted data is within a normal range defined by the normal data group stored in the 1 st database. Then, in step 3, the judgment result in step 2 is evaluated based on the rolling result of the rolled product corresponding to the extracted data, and when the judgment result does not match the rolling result, the judgment criterion in the predetermined normal range is changed.
By providing the 1 st step and the 2 nd step among the three steps, even if data of past abnormal phenomena are not accumulated, it is possible to diagnose an abnormality by providing normal data, that is, data obtained when a good rolling result is obtained. Collecting normal data is easier and does not require labor than collecting abnormal data. In addition, if abnormality diagnosis is performed based on normal data, undefined unknown abnormality can also be dealt with. Further, since the 3 rd step is provided, the judgment criterion for whether or not the rolling mill is within the normal range is updated so as to match the actual rolling result, and therefore the accuracy of the abnormality diagnosis using the normal data can be improved.
The abnormality diagnosis method for a rolling mill according to the present invention may further include the following 4 th step or 5 th step. The 4 th step is a step of registering the extracted data in the 1 st database when it is determined that the extracted data is within the normal range and the rolling result of the rolled product is good. The step 5 is a step of registering the extracted data in the 2 nd database in which the abnormal data is accumulated, when it is determined that the extracted data is not within the normal range and the rolling result of the rolled product is not good. When step 4 is provided, the amount of accumulation of normal data used for setting the determination criterion can be increased, and the accuracy of the determination criterion can be improved. In the case of the 5 th step, the data relating to the abnormality occurring in the rolling mill can be accumulated in the 2 nd database including the undefined unknown abnormality.
In step 3, when it is determined that the extracted data is within the normal range but the rolling result of the rolled product is not good, the determination criterion may be changed in a strict direction. In step 3, when it is determined that the extracted data is not within the normal range but the rolling result of the rolled product is good, the determination criterion may be changed in a relaxed direction. Alternatively, the method may further include step 6: if the extracted data is determined not to be within the normal range but the rolling result of the rolled product is good, an alarm is output to a display device.
The data extracted in step 1 may be data having a plurality of related dimensions as components. In this case, in step 2, it may be determined whether or not the extracted data is within a normal range based on a distance between the extracted data and a normal data group in a space having a plurality of scales as axes. Further, the distance between the normal data group and the extracted data may be calculated by a multidimensional scaling method. In step 2, the distance between the normal data group and the extracted data may be corrected by using a dynamic time warping algorithm.
Further, according to the present invention, there are also provided a program for causing a computer to execute the processing of each step in the above-described abnormality diagnosis method for a rolling mill, and a storage medium storing the program.
The abnormality diagnosis device for a rolling mill according to the present invention is connected to a data collection device that collects and records process data of the rolling mill in time series, and diagnoses an abnormality of the rolling mill based on the data recorded by the data collection device, and is configured as follows in detail.
That is, the abnormality diagnosis device for a rolling mill according to the present invention includes a data extraction unit, a determination unit, and a determination result evaluation unit. The data extraction unit is configured to extract data corresponding to the same rolled product from the data recorded by the data collection device. The determination unit is configured to determine whether or not the data extracted by the data extraction unit is within a normal range defined by the normal data group stored in the 1 st database. The determination result evaluation unit is configured to evaluate the determination result of the determination unit based on the rolling result of the rolled product corresponding to the data extracted by the data extraction unit. More specifically, the determination result evaluation unit is configured to change the determination criterion for defining the normal range when the determination result of the determination unit does not match the rolling result.
According to the above configuration, particularly by providing the data extracting unit and the determining unit, even if the data of the past abnormal phenomenon is not accumulated, the abnormality diagnosis can be performed as long as the normal data, that is, the data obtained when the good rolling result is obtained is present. Collecting normal data is easier and does not require labor than collecting abnormal data. In addition, if abnormality diagnosis is performed based on normal data, undefined unknown abnormality can also be dealt with. Further, by providing the determination result evaluation unit, the determination criterion as to whether or not the rolling pass is within the normal range is updated so as to match the actual rolling result, and therefore the accuracy of the abnormality diagnosis using the normal data can be improved.
The computer constituting the abnormality diagnosis device may execute each process of the data extraction unit, the determination unit, and the determination result evaluation unit. That is, the abnormality diagnosis device may be configured by a computer including a processor and a memory storing a program, and the program may be configured such that the computer operates as the data extraction unit, the determination unit, and the determination result evaluation unit when the program read out from the memory is executed by the processor.
The determination result evaluation unit may be configured to register the extracted data in the 1 st database when the determination unit determines that the extracted data is within the normal range and that the rolling result of the rolled product is good. With this configuration, the amount of accumulation of normal data used for setting the determination criterion can be increased, and the accuracy of the determination criterion can be improved.
Further, the determination result evaluation unit may be configured to register the extracted data in the 2 nd database in which the abnormal data is accumulated, when the determination unit determines that the extracted data is not within the normal range and the rolling result of the rolled product is not good. With this configuration, it is possible to accumulate data relating to an abnormality occurring in the rolling facility, including undefined unknown abnormalities, in the 2 nd database.
The determination result evaluation unit may be configured to change the determination criterion in a strict direction when the rolling result of the rolled product is not good although the extracted data is determined to be within the normal range by the determination unit. The determination result evaluation unit may be configured to change the determination criterion in a relaxed direction when the rolling result of the rolled product is good although the extracted data is determined not to be within the normal range by the determination unit. Alternatively, the determination result evaluation unit may be configured to output an alarm to the display device when the rolling result of the rolled product is good although the extracted data is determined by the determination unit not to be within the normal range.
The data extraction unit may be configured to extract data having a plurality of related dimensions as components. In this case, the determination unit may be configured to determine whether or not the extracted data is within a normal range based on a distance between a normal data group in a space having a plurality of scales as axes and the extracted data. Further, the distance between the normal data group and the extracted data may be calculated by a multidimensional scaling method. The determination unit may be configured to correct the distance between the normal data group and the extracted data by using a dynamic time warping algorithm.
ADVANTAGEOUS EFFECTS OF INVENTION
According to the present invention, even if data of a past abnormal phenomenon is not accumulated, it is possible to diagnose an abnormality as long as normal data, that is, data obtained when a good rolling result is obtained. Collecting normal data is easier and does not require labor than collecting abnormal data. In addition, by performing abnormality diagnosis based on normal data, undefined unknown abnormalities can be dealt with. Further, according to the present invention, since the criterion for determining whether or not the rolling mill is within the normal range is updated so as to match the actual rolling result, the accuracy of the abnormality diagnosis using the normal data can be improved.
Drawings
Fig. 1 is a diagram showing a configuration example of a rolling mill to which an abnormality diagnostic device according to an embodiment of the present invention is applied.
Fig. 2 is a block diagram showing an example of a hardware configuration of the abnormality diagnostic device according to the embodiment of the present invention.
Fig. 3 is a functional block diagram showing a part of functions of the abnormality diagnostic device according to the embodiment of the present invention.
Fig. 4 is a diagram showing a range of guaranteed values relating to the quality of a rolled product.
Fig. 5 is a block diagram showing a flow of score calculation processing used for abnormality determination.
Fig. 6 is a diagram showing an example of time-series data and an example of calculation of a score based on the time-series data.
Fig. 7 is a diagram showing an example of a method for determining an abnormality from the score.
Fig. 8 is a diagram showing a method of determining an abnormality in the abnormality diagnostic device according to the embodiment of the present invention.
Fig. 9 is a diagram illustrating a change in the determination criterion.
Fig. 10 is a diagram illustrating a change in the determination criterion.
Fig. 11 is a table showing the correspondence between the determination result, the rolling result, and the selected process.
Fig. 12 is a flowchart showing a procedure of abnormality diagnosis performed by the abnormality diagnosis apparatus according to the embodiment of the present invention.
Detailed Description
Embodiments of the present invention will be described with reference to the accompanying drawings. However, the embodiments described below are illustrative of apparatuses and methods for embodying the technical ideas of the present invention, and are not intended to limit the structures, arrangements, and processing sequences of the constituent elements to the following unless otherwise specifically indicated. The present invention is not limited to the embodiments described below, and various modifications can be made without departing from the scope of the present invention.
< example of Rolling Mill >
Fig. 1 is a diagram showing a configuration example of a rolling mill to which an abnormality diagnostic device according to an embodiment of the present invention is applied. Here, a hot rolling line 20 for producing steel sheets, which is widely used, is exemplified as a typical example of a rolling facility. The arrows in the figure indicate the direction of flow of the rolled material. The hot rolling line 20 is a line for producing a rolled product from a rolling stock (hereinafter, referred to as a "slab").
The Hot rolling line 20 includes a plurality of devices such as a heating furnace 2, an HSB (Hot scalebeater) 3, a roughing edger 4, a roughing horizontal mill 5, an edge heater 8, an FSB (finish scalebeater) 11, an F1 edger 12, a finishing mill 13, a run-out table 16, and a lower coiler 19, which are arranged along the flow direction of a material to be rolled. These devices are connected by a conveyance table, not shown, and are driven by a motor and a hydraulic device, respectively. The hot rolling line 20 includes a plurality of measuring devices such as a rough-outlet thermometer 6, a thermal imaging device 7, a precise-inlet thermometer 10, a multipurpose measuring instrument 14, a precise-outlet thermometer 15, a thermal imaging device 17, and a coiler-inlet thermometer 18. Hereinafter, the outline of the equipment and the measuring device constituting the hot rolling line 20 will be described.
The heating furnace 2 is a furnace for heating a slab. The HSB3 is used to remove an oxide film formed on the surface of the slab by heating in the heating furnace 2. The oxide film is removed mainly by spraying high-pressure water from a nozzle to the slab.
The roughing edger 4 is a device for pressing down a material to be rolled (including a state in the middle of completion of a slab as a product, which will be described below) in the width direction mainly in order to ensure width accuracy. The roughing edger 4 is configured to bring the rolls into contact with the material to be rolled in the width direction when viewed from above the production line. The roughing horizontal rolling mill 5 is constituted by one or more stands. In order to shorten the line length, and since the rough rolling requires a plurality of passes, the rough horizontal rolling mill 5 is often constituted by a reversing mill. The rough horizontal rolling mill 5 is provided with a device called a descaler for removing an oxide film on the surface of the material 1 to be rolled, which is a semi-finished product, by spraying high-pressure water. Since rolling is performed at high temperature, an oxide film is easily formed, and it is necessary to appropriately use such an oxide film removing apparatus. The roughing-side thermometer 6 is disposed on the outlet side of the roughing horizontal rolling mill 5, and measures the surface temperature of the material to be rolled, which is a semi-finished product during rolling.
The edge heater 8 is a device for raising the temperature of the end portion in the width direction of the material to be rolled by induction heating using electromagnetic force. Thermal imaging devices 7 are disposed on the entrance side and the exit side of the edge heater 8, respectively. A finish side thermometer 10 is disposed further downstream in the direction of flow of the rolled material and on the entry side of the finishing mill 13. The entry side temperature of the finishing mill 13 is closely related to the prediction of the deformation resistance of the material. Therefore, the temperature just before the machining needs to be measured by the precision thermometer 10. Alternatively, it is necessary to obtain a highly accurate predicted temperature value taking into account the conveyance time from the rough-side thermometer 6 to the finishing mill 13 using the actual measurement value of the rough-side thermometer 6.
The FSB11 is a device for removing an oxide film on the surface of a material to be rolled. The FSB11 is used to remove scale generated by the distance and time required for the material to be rolled to reach the entry side of the finishing mill 13 after the completion of rough rolling, and to improve the surface properties of the rolled product after finish rolling. The edger 12 is provided on the entry side of the finishing mill 13 and ensures the dimensional accuracy in the width direction of the finish-rolled product. The edger 12 contributes to a temperature rise at the end in the width direction of the rolled material by plastic working by rolling.
The finishing mill 13 is a tandem mill in which a plurality of rolling mills called stands are arranged. In the finishing mill 13, rolling is performed to obtain a target product accuracy of a rolled product. A multi-purpose gauge 14 and a finishing side thermometer 15 are disposed on the outlet side of the finishing mill 13. The multi-purpose measuring instrument 14 has a system in which detectors of X-rays are arranged along the width direction. The multipurpose measuring instrument 14 is a composite measuring instrument capable of measuring a thickness distribution in the width direction, and therefore can measure a thickness, a protrusion, a width, and the like of a board by 1 instrument. The multi-purpose measuring instrument 14 has a thermal imaging device therein, and measures the temperature of the rolled product in the width direction by the thermal imaging device to correct the detection value by the X-ray. The finishing side thermometer 15 measures the surface temperature of the rolled material after rolling. The temperature of the rolled product is closely related to the formation of the metal structure and the material of the product, and therefore needs to be controlled to an appropriate temperature.
The run-out table 16 is a device for cooling a rolled product with cooling water in order to control the temperature of the rolled product. Further, a forced cooling device may be provided instead of or in addition to the normal run-out table cooling device. In the run-out table 16, in order to prevent a temperature decrease at the end in the width direction of the rolled product, an edge shield for preventing the cooling water from reaching the width direction may be applied. On the exit side of the run-out table 16 and the entry side of the lower coiler 19, a thermal imaging device 17 and a coiler entry side thermometer 18 are disposed. The coiler-entry-side thermometer 18 measures the surface temperature of the rolled product after rolling. The temperature of the rolled product is closely related to the formation of the metal structure and the material of the product, and therefore needs to be controlled to an appropriate temperature. The lower coiler 19 is a device for coiling the rolled product in order to convey it. The rolled product in the form of a roll wound by the lower coiler 19 is referred to as a coil.
A data collection device 22 is provided in the hot rolling line 20. The data collection device 22 collects, at all times or intermittently, set values and actual values for the respective facilities constituting the hot rolling line 20, measurement values of the measurement device, operation amounts for appropriately operating the facilities, and the like, and records the collected values in a recording device such as a hard disk. For example, for a rolled product wound by the lower coiler 19, a guaranteed value to a customer is determined for quality indexes such as a sheet thickness, a sheet width, a bulge, a temperature on the outlet side, a temperature on the inlet side of the coiler, and the like measured by the multipurpose measuring instrument 14, the temperature on the outlet side 15, the temperature on the inlet side of the coiler 18, and the like. The data collection device 22 collects and records process data including these quality indexes as components in time series. The data collection device 22 may be constituted by a single computer or may be constituted by a plurality of computers connected to a network.
< example of hardware configuration of abnormality diagnosis apparatus >
Fig. 2 is a block diagram showing an example of the hardware configuration of the abnormality diagnostic device 30 according to the present embodiment. The abnormality diagnosis device 30 is a computer including a cpu (central Processing unit)31, a rom (read Only memory)32, a ram (random access memory)33, an input/output interface 34, a system bus 35, a memory 36, an input device 37, a display device 38, and a communication device 39.
The CPU31 is a processing device that executes various arithmetic processes using programs, data, and the like stored in the ROM32 and the RAM 33. The ROM32 is a read-only storage device that stores basic programs for causing a computer to realize the respective functions, environment files, and the like. The RAM33 is a main storage device that stores programs executed by the CPU31 and data necessary for executing the programs, and is capable of high-speed reading and writing. The input/output interface 34 is a device that mediates connection between various hardware and the system bus 35. The system bus 35 is an information transfer path shared by the CPU31, the ROM32, the RAM33, and the input/output interface 34.
Further, hardware such as an input device 37, a display device 38, a memory 36, and a communication device 39 is connected to the input/output interface 34. The input device 37 is a device that processes an input from a user. The display device 38 is a device that displays information related to the diagnosis result and the abnormality diagnosis. The memory 36 is a large-capacity auxiliary storage device that stores programs and data, and is, for example, a hard disk device, a nonvolatile semiconductor memory, or the like. The communication device 39 is a device capable of performing data communication with the data collection device 22 by wire or wirelessly.
< function of abnormality diagnosis apparatus >
Fig. 3 is a functional block diagram showing a part of functions of the abnormality diagnostic device 30 according to the present embodiment. The abnormality diagnostic device 30 includes a data extraction unit 101, a determination unit 102, and a determination result evaluation unit 103. The functions of these functional units 101, 102, and 103, that is, the functions of the abnormality diagnostic device 30, are realized in a computer by the CPU31 (see fig. 2) of the abnormality diagnostic device 30 executing a program read from the ROM32 (see fig. 2). The abnormality diagnostic device 30 includes a normal pattern database 104 as a 1 st database and an abnormal pattern database 105 as a 2 nd database. The normal pattern database 104 and the abnormal pattern database 105 are constructed in the memory 36 (refer to fig. 2). The program for causing the computer to function as the abnormality diagnostic device 30 is supplied via a network or a computer-readable storage medium (for example, a CD-ROM, a DVD, a USB memory, or the like).
The data extraction unit 101 is configured to cut out data recorded by the data collection device 22 on a roll-to-roll basis. Data of each item is recorded in time series in the data collection device 22. The data extraction unit 101 extracts data corresponding to the same web from the data recorded by the data collection device 22 based on the time information of the data. The data extracted by the data extraction unit 101 is data having a plurality of related dimensions as components, such as the plate thickness and the load.
The determination unit 102 is configured to determine whether or not the data extracted by the data extraction unit 101 is within a normal range. Here, fig. 4 is a diagram showing data extracted by the data extraction unit 101 and margins of guaranteed values set for the data. For a rolled product, a margin of a guaranteed value is determined for each quality index. During rolling, each facility is controlled so that each quality index is within the tolerance. However, the quality index may deviate from the margin due to various reasons such as a failure of the apparatus due to insufficient maintenance and an inappropriate control gain. The normal range used by the determination section 102 in the determination is associated with the margin of the guaranteed value of the quality index.
The determination unit 102 is configured to calculate a score from the data extracted by the data extraction unit 101. Here, fig. 5 is a block diagram showing a flow of a score calculation process used for abnormality determination. First, a model of the distribution of data is learned from the data obtained so far. The model may be a probabilistic model or a statistical model. Then, the degree of abnormality of the data or the degree of abnormal change of the model is scored using the learned model. Fig. 6 is a diagram showing an example of time-series data extracted by the data extraction unit 101 and an example of calculating a score based on the time-series data.
As a method of determining an abnormality from the calculated score, it can be said that determination based on a threshold value is a general method. Fig. 7 is a diagram showing an example of a method for determining an abnormality from the score. In fig. 7, a straight line drawn with a dotted line represents a threshold value for the score. However, in this method, the accuracy of the abnormality determination depends on the method of determining the threshold value, and therefore it is difficult to say that the accuracy of the abnormality determination can be sufficiently secured in this method.
The determination unit 102 is configured to determine an abnormality not only by one scale but also by combining scales that enable subordinate determination. Fig. 8 is a diagram showing a method of determining an abnormality by the determination unit 102. Here, on the plane defined by scale 1 and scale 2, a point defined by a combination of the scores (representative values of the scores) of scale 1 and scale 2 is drawn. In fig. 8, the data group grouped into the "normal data group" is the data group accumulated in the normal pattern database 104. Only normal data that has been finally determined to be normal by the determination result evaluation unit 103 described below among the data extracted up to this time by the data extraction unit 101 is stored in the normal pattern database 104.
The determination unit 102 measures the distance between data on a plane defined by the scale 1 and the scale 2, specifically, the distance between the data extracted by the data extraction unit 101 and the normal data group, by a multidimensional scaling method, a k-nearest neighbor method, a density ratio estimation method, or the like. Further, the distance may be corrected by using a dynamic time warping algorithm. In fig. 8, a curve drawn by a broken line is a determination reference line defining a normal range, and is set with reference to the normal data group stored in the normal mode database 104. The determination unit 102 determines that the data closer to the normal data group than the determination reference line is within the normal range, and determines that the data farther from the normal data group than the determination reference line is outside the normal range. When the sheet thickness is used as the dimension 1, for example, it is preferable to use a load having a close relationship with the sheet thickness as the dimension 2. When an abnormality occurs in the sheet thickness, the possibility of the abnormality occurring also in a load having close relation with the sheet thickness is high, and therefore, by adopting the dimensions 1 and 2 as described above, it is possible to eliminate the noise of a single sensor and to more clearly determine the abnormality.
The determination result evaluation unit 103 is configured to evaluate the determination result of the determination unit 102. As a criterion for the evaluation, an actual rolling result of the rolled product corresponding to the data extracted by the data extraction unit 101, that is, the data to be judged as whether or not the data is within the normal range in the judgment unit 102 is used. Specifically, the rolling result is obtained by checking whether or not the thickness, width, temperature, shape, and the like of the steel sheet are within the quality control values after the rolling is completed. The inspection for obtaining the rolling result is preferably automatically performed by a dedicated device or the like and is acquired online via the communication device 39. However, the result of the examination by the person may be input via the input device 37.
Specifically, the determination result evaluation unit 103 compares the determination result of the determination unit 102 with the rolling result as the quality standard, and evaluates whether or not the determination result matches the rolling result. Then, a process corresponding to the evaluation result is performed. Here, fig. 11 is a table showing the correspondence between the determination result, the rolling result, and the selected process. Hereinafter, how to perform the processing based on the evaluation result will be described with reference to fig. 11.
First, the result of the determination in the normal range is matched with the good rolling result. In this case, the determination result evaluation unit 103 registers the data determined to be within the normal range by the determination unit 102 in the normal pattern database 104. In this way, the amount of accumulation of the normal data used for setting the determination criterion can be increased, and the accuracy of the determination criterion can be improved. The data registered here is data corresponding to the same rolled product (coil) extracted by the data extraction unit 101, and there is a difference in the pattern of deviation of the scale (quality index) with respect to the longitudinal direction of the coil among the data. Normal data having various patterns are accumulated in the normal pattern database 104.
On the other hand, the determination result that is within the normal range does not match the defective rolling result. In this case, the determination result evaluation unit 103 changes the determination criterion of the determination unit 102 in order to improve the accuracy of the next and subsequent determinations. Specifically, the determination criterion is changed in a strict direction so that the data determined to be within the normal range this time is determined to be outside the normal range. For example, in fig. 9, data a is plotted on a plane defined by scale 1 and scale 2. The data a is located on the same side as the normal data group with respect to the curve representing the judgment reference 1. Therefore, according to the judgment reference 1 shown in fig. 9, the data a is judged to be within the normal range. However, when the actual rolling result of the rolled product corresponding to the data a is poor, it cannot be said that the determination that the data a is within the normal range is accurate. Therefore, in this case, as shown in fig. 10, the determination criterion 1 is changed to the determination criterion 2. According to the judgment reference 2, the data a is judged to be out of the normal range, and becomes matched with the actual rolling result.
The judgment result that is not within the normal range matches the poor rolling result. In this case, the determination result evaluation unit 103 registers the data determined by the determination unit 102 to be outside the normal range in the abnormal pattern database 105. In this way, it is possible to accumulate data relating to an abnormality occurring in the hot rolling line 20 into the abnormality pattern database 105, including undefined unknown abnormalities. Further, by comparing the pattern of the newly registered data, specifically, the pattern of the deviation of the scale (quality index) with respect to the longitudinal direction of the web, with the pattern of the abnormal data group already registered in the abnormal pattern database 105, it is possible to determine whether or not an undefined unknown abnormality has occurred.
On the other hand, the judgment result outside the normal range does not match the good rolling result. In this case, the determination result evaluation unit 103 changes the determination criterion of the determination unit 102 in order to improve the accuracy of the next and subsequent determinations. Specifically, the determination criterion is changed in a loose direction so that the data determined to be outside the normal range this time is determined to be within the normal range. However, if the rolling result is good but the data is outside the normal range, there is a possibility that a potential failure may exist inside the apparatus. Therefore, in this case, an alarm screen may be output to the display device 38 so that the operator can investigate the failure of the equipment.
< sequence of abnormality diagnosis >
According to the abnormality diagnostic device 30 having the above-described functions, the abnormality diagnosis of the rolling facility is performed in the following procedure. The procedure of the abnormality diagnosis performed by the abnormality diagnosis device 30 will be described below with reference to the flowchart shown in fig. 12.
In step S1, the abnormality diagnostic device 30 extracts data corresponding to the same rolled product from the data recorded by the data collection device 22.
Next, in step S2, the abnormality diagnostic device 30 determines whether or not the data extracted in step S1 is within the normal range defined by the normal data group stored in the normal pattern database 104. Specifically, the abnormality diagnostic device 30 measures the distance between the normal data group and the data extracted in step S1 by a multidimensional scaling method or the like. In this case, the distance is preferably corrected by using a dynamic time warping algorithm (DTW). Then, it is determined whether or not the data extracted in step S1 is within the normal range based on the measured distance.
Next, the abnormality diagnostic device 30 evaluates the determination result in step S2 based on the rolling result of the rolled product corresponding to the data extracted in step S1. In detail, in the case where it is determined in step S2 that the data extracted in step S1 is within the normal range, the abnormality diagnostic device 30 determines in step S3 whether the affirmative determination in step S2 matches the actual rolling result.
If the affirmative determination in step S2 is matched with the actual rolling result, that is, if the actual rolling result is good, the abnormality diagnostic device 30 selects the process in step S4. In step S4, the abnormality diagnostic device 30 registers the data extracted in step S1 in the normal pattern database 104.
If the affirmative determination at step S2 does not match the actual rolling result, that is, if the actual rolling result is not good, the abnormality diagnostic device 30 selects the process at step S5. In step S5, the abnormality diagnostic device 30 changes the criterion of the determination made in step S2 in a strict direction, and registers the data extracted in step S1 in the abnormality pattern database 105.
In addition, in the case where it is determined in step S2 that the data extracted in step S1 is not within the normal range, the abnormality diagnostic device 30 determines in step S6 whether the negative determination result in step S2 matches the actual rolling result.
If the negative determination result in step S2 matches the actual rolling result, that is, if the actual rolling result is not good, the abnormality diagnostic device 30 selects the process in step S7. In step S7, the abnormality diagnostic device 30 registers the data extracted in step S1 in the abnormality pattern database 105.
If the negative determination result of step S2 does not match the actual rolling result, that is, if the actual rolling result is good, the abnormality diagnostic device 30 selects the process of step S8. In step S8, abnormality diagnostic device 30 changes the criterion of the determination made in step S2 in a direction in which the determination is relaxed. Alternatively, the abnormality diagnostic device 30 outputs an alarm to the display device 38.
The normal pattern database 104 and the abnormal pattern database 105 are used to determine the criteria for determining the abnormal coil and the normal coil based on the rolling condition that changes every day, based on the abnormality diagnosis performed in the above-described order. In this way, by combining the normal pattern database 104 and the abnormal pattern database 105, the accuracy of abnormality detection can be improved, and by comparing with the normal data group stored in the normal pattern database 104, it is possible to cope with unknown abnormality.
Description of the symbols
20: a hot rolling line (rolling facility); 22: a data collection device; 30: an abnormality diagnostic device; 36: a memory; 37: an input device; 38: a display device; 39: a communication device; 101: a data extraction unit; 102: a judgment section; 103: a judgment result evaluation unit; 104: a normal mode database; 105: an abnormal pattern database.
Claims (16)
1. A method for diagnosing an abnormality of a rolling mill, which collects and records process data of the rolling mill in time series by a data collection device, and diagnoses the abnormality of the rolling mill based on the data recorded by the data collection device, the method comprising:
a step 1 of extracting data corresponding to the same rolled product from the data recorded by the data collecting device;
a step 2 of judging whether or not the extracted data is within a normal range defined by a normal data group stored in a 1 st database; and
and a 3 rd step of evaluating the determination result in the 2 nd step based on the rolling result of the rolled product corresponding to the extracted data, and changing the determination criterion defining the normal range when the determination result does not match the rolling result.
2. The abnormality diagnostic method of a rolling facility according to claim 1,
further comprises the following step 4: when it is determined that the extracted data is within the normal range and the rolling result of the rolled product is good, the extracted data is registered in the 1 st database.
3. The abnormality diagnostic method of a rolling facility according to claim 1 or 2,
further comprises the following step 5: when it is determined that the extracted data is not within the normal range and the rolling result of the rolled product is not good, the extracted data is registered in the 2 nd database in which abnormal data is stored.
4. The abnormality diagnostic method of a rolling facility according to claim 1 or 2,
in the step 3, when it is determined that the extracted data is within the normal range but the rolling result of the rolled product is not good, the determination criterion is changed in a strict direction.
5. The abnormality diagnostic method of a rolling facility according to claim 1 or 2,
in the step 3, when it is determined that the extracted data is not within the normal range but the rolling result of the rolled product is good, the determination criterion is changed in a direction to be relaxed.
6. The abnormality diagnostic method of a rolling facility according to claim 1 or 2,
further comprises the following step 6: and outputting an alarm to a display device when the rolling result of the rolled product is good although the extracted data is determined not to be within the normal range.
7. The abnormality diagnostic method of a rolling facility according to claim 1 or 2,
the data extracted in the above-described step 1 is data having associated plural scales as components,
in the step 2, a distance between the normal data group and the extracted data in the space having the plurality of scales as axes is measured, and whether or not the extracted data is within the normal range is determined based on the distance.
8. The abnormality diagnostic method of a rolling facility according to claim 7,
in the step 2, the distance is corrected using a dynamic time warping algorithm.
9. An abnormality diagnosis device for a rolling mill, which is connected to a data collection device that collects and records process data of the rolling mill in time series, and diagnoses an abnormality of the rolling mill based on the data recorded by the data collection device, the abnormality diagnosis device comprising:
a data extraction unit configured to extract data corresponding to the same rolled product from the data recorded by the data collection device;
a determination unit configured to determine whether or not the extracted data is within a normal range defined by a normal data group stored in a 1 st database; and
a determination result evaluation unit configured to evaluate the determination result of the determination unit based on the rolling result of the rolled product corresponding to the extracted data,
the determination result evaluation unit is configured to change the determination criterion that defines the normal range when the determination result of the determination unit does not match the rolling result.
10. The abnormality diagnostic device of a rolling facility according to claim 9,
the determination result evaluation unit is configured to register the extracted data in the 1 st database when the determination unit determines that the extracted data is within the normal range and that the rolling result of the rolled product is good.
11. The abnormality diagnostic device of a rolling facility according to claim 9 or 10,
the determination result evaluation unit is configured to register the extracted data in a 2 nd database in which abnormal data is stored, when the determination unit determines that the extracted data is not within the normal range and that the rolling result of the rolled product is not good.
12. The abnormality diagnostic device of a rolling facility according to claim 9 or 10,
the determination result evaluation unit is configured to change the determination criterion in a strict direction when the rolling result of the rolled product is not good although the determination unit determines that the extracted data is within the normal range.
13. The abnormality diagnostic device of a rolling facility according to claim 9 or 10,
the determination result evaluation unit is configured to change the determination criterion in a relaxed direction when the rolling result of the rolled product is good although the determination unit determines that the extracted data is not within the normal range.
14. The abnormality diagnostic device of a rolling facility according to claim 9 or 10,
the determination result evaluation unit is configured to output an alarm to a display device when the rolling result of the rolled product is good although the determination unit determines that the extracted data is not within the normal range.
15. The abnormality diagnostic device of a rolling facility according to claim 9 or 10,
the data extraction unit is configured to extract data having a plurality of related dimensions as components,
the determination unit is configured to measure a distance between the normal data group and the extracted data in a space having the plurality of scales as axes, and determine whether the extracted data is within the normal range based on the distance.
16. The abnormality diagnostic device of a rolling facility according to claim 15,
the determination unit is configured to correct the distance using a dynamic time scaling algorithm.
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| PCT/JP2016/085211 WO2018096682A1 (en) | 2016-11-28 | 2016-11-28 | Method and device for diagnosing abnormality in rolling equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109996615A CN109996615A (en) | 2019-07-09 |
| CN109996615B true CN109996615B (en) | 2020-07-07 |
Family
ID=62194917
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201680091146.9A Active CN109996615B (en) | 2016-11-28 | 2016-11-28 | Abnormality diagnosis method and apparatus for rolling equipment |
Country Status (4)
| Country | Link |
|---|---|
| JP (1) | JP6791261B2 (en) |
| CN (1) | CN109996615B (en) |
| TW (1) | TWI649133B (en) |
| WO (1) | WO2018096682A1 (en) |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| AU2018426458B2 (en) * | 2018-06-08 | 2023-12-21 | Chiyoda Corporation | Assistance device, learning device, and plant operation condition setting assistance system |
| US11567482B2 (en) * | 2018-08-31 | 2023-01-31 | Toshiba Mitsubishi-Electric Industrial Systems Corporation | Manufacturing process monitoring apparatus |
| JP7214417B2 (en) * | 2018-09-20 | 2023-01-30 | 株式会社Ｓｃｒｅｅｎホールディングス | Data processing method and data processing program |
| JP2020095617A (en) * | 2018-12-14 | 2020-06-18 | コニカミノルタ株式会社 | Safety management support system and control program |
| KR20210017577A (en) * | 2019-08-09 | 2021-02-17 | 주식회사 엘지화학 | Quantitatively diagnosis method for manufacturing facility |
| WO2021145156A1 (en) * | 2020-01-14 | 2021-07-22 | Ｊｆｅスチール株式会社 | Abnormality diagnosis system and abnormality diagnosis method |
| TWI770536B (en) * | 2020-06-22 | 2022-07-11 | 中國鋼鐵股份有限公司 | Method and system for identifying causes of hot-rolled product defects |
| JP7031713B1 (en) * | 2020-10-22 | 2022-03-08 | Ｊｆｅスチール株式会社 | Abnormality diagnosis model construction method, abnormality diagnosis method, abnormality diagnosis model construction device and abnormality diagnosis device |
| JP7447779B2 (en) * | 2020-12-21 | 2024-03-12 | 東芝三菱電機産業システム株式会社 | Shape control system for rolled materials |
| TWI808407B (en) * | 2021-01-21 | 2023-07-11 | 日商東芝三菱電機產業系統股份有限公司 | Roll management device |
| CN116324654A (en) * | 2021-07-13 | 2023-06-23 | 东芝三菱电机产业系统株式会社 | Abnormality detection device |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3397726B2 (en) * | 1999-07-14 | 2003-04-21 | 株式会社日立製作所 | Rolling mill abnormality diagnosis method and apparatus |
| CN1609736A (en) * | 2003-10-22 | 2005-04-27 | 株式会社日立制作所 | Controlling device, system utilizing the same controller and controlling method |
| CN1267215C (en) * | 2002-11-20 | 2006-08-02 | Posco株式会社 | Fault diagnosis apparatus and method for hot fine rolling band steel |
| JP2011050990A (en) * | 2009-09-02 | 2011-03-17 | Jfe Steel Corp | On-line diagnostic method by tracking sensor on hot rolling line |
| CN102441579A (en) * | 2010-10-13 | 2012-05-09 | 上海宝钢工业检测公司 | Online monitoring method for running state of hot continuous rolling mill |
| CN103920717A (en) * | 2013-01-10 | 2014-07-16 | 东芝三菱电机产业系统株式会社 | Set value calculating device and set value calculating method |
| WO2015177870A1 (en) * | 2014-05-20 | 2015-11-26 | 東芝三菱電機産業システム株式会社 | Manufacturing equipment diagnosis support system |
| CN105738136A (en) * | 2016-01-27 | 2016-07-06 | 安徽容知日新信息技术有限公司 | Equipment abnormity detection method and device |
-
2016
- 2016-11-28 CN CN201680091146.9A patent/CN109996615B/en active Active
- 2016-11-28 WO PCT/JP2016/085211 patent/WO2018096682A1/en active Application Filing
- 2016-11-28 JP JP2018552375A patent/JP6791261B2/en active Active
-
2017
- 2017-02-17 TW TW106105278A patent/TWI649133B/en active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3397726B2 (en) * | 1999-07-14 | 2003-04-21 | 株式会社日立製作所 | Rolling mill abnormality diagnosis method and apparatus |
| CN1267215C (en) * | 2002-11-20 | 2006-08-02 | Posco株式会社 | Fault diagnosis apparatus and method for hot fine rolling band steel |
| CN1609736A (en) * | 2003-10-22 | 2005-04-27 | 株式会社日立制作所 | Controlling device, system utilizing the same controller and controlling method |
| JP2011050990A (en) * | 2009-09-02 | 2011-03-17 | Jfe Steel Corp | On-line diagnostic method by tracking sensor on hot rolling line |
| CN102441579A (en) * | 2010-10-13 | 2012-05-09 | 上海宝钢工业检测公司 | Online monitoring method for running state of hot continuous rolling mill |
| CN103920717A (en) * | 2013-01-10 | 2014-07-16 | 东芝三菱电机产业系统株式会社 | Set value calculating device and set value calculating method |
| WO2015177870A1 (en) * | 2014-05-20 | 2015-11-26 | 東芝三菱電機産業システム株式会社 | Manufacturing equipment diagnosis support system |
| CN105738136A (en) * | 2016-01-27 | 2016-07-06 | 安徽容知日新信息技术有限公司 | Equipment abnormity detection method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2018096682A1 (en) | 2018-05-31 |
| TW201819061A (en) | 2018-06-01 |
| JP6791261B2 (en) | 2020-11-25 |
| CN109996615A (en) | 2019-07-09 |
| TWI649133B (en) | 2019-02-01 |
| JPWO2018096682A1 (en) | 2019-10-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109996615B (en) | Abnormality diagnosis method and apparatus for rolling equipment | |
| CN107949813B (en) | Manufacturing equipment diagnosis support device and manufacturing equipment diagnosis support method | |
| KR101906029B1 (en) | Manufacturing equipment diagnosis support system | |
| JP7044175B2 (en) | Abnormality judgment support device | |
| EP2286935B1 (en) | Steel plate quality assurance system and method | |
| JP7158569B2 (en) | Methods and electronic devices for monitoring the production of metal products, associated computer programs and equipment | |
| JP4106040B2 (en) | Abnormality diagnosis and abnormality avoidance method for steel sheet cooling control device | |
| KR102426172B1 (en) | Tightening Occurrence Prediction System | |
| Haapamäki et al. | Data Mining Methods in Hot Steel Rolling for Scale Defect Prediction. | |
| JP6760503B2 (en) | Manufacturing process monitoring device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |