JP6662222B2 - Abnormal state diagnostic method and abnormal state diagnostic apparatus for manufacturing process - Google Patents

Description

  The present invention relates to a method and apparatus for diagnosing abnormal conditions in a manufacturing process such as a steel process.

  Methods for diagnosing abnormal conditions in the manufacturing process include a model-based approach and a database approach. The model-based approach is an approach in which a model expressing a physical or chemical phenomenon in a manufacturing process by mathematical formulas is constructed, and an abnormal state of the manufacturing process is diagnosed using the constructed model. On the other hand, the database approach is an approach in which a statistical analysis model is constructed from operation data obtained in the production process, and an abnormal state of the production process is diagnosed using the constructed model.

  In a manufacturing process such as a steel process, products of various types and sizes are manufactured in one manufacturing line, and thus there are countless operation patterns. Further, in a manufacturing process such as a blast furnace, a natural process such as iron ore or coke is used as a raw material, and thus the manufacturing process varies greatly. For this reason, when diagnosing an abnormal state of a manufacturing process such as a steel process, there is a limit in an approach using only the model-based approach.

  As a database approach, there is a diagnostic method to determine the similarity with the current operation data by converting the operation data at the time of past abnormal occurrence into a database, and conversely, the difference from the current operation data by converting the normal operation data into a database There is a diagnostic method for determining However, in a manufacturing process such as a steel process, in addition to the large number of equipment used for manufacturing, especially when there are many aging facilities like Japan, troubles unprecedented in the past often occur. . For this reason, there is a limit in predicting an abnormal state in the former diagnostic method based on past trouble cases.

  On the other hand, the latter diagnostic method is described in Patent Documents 1 and 2. Specifically, Patent Literatures 1 and 2 disclose a method of predicting or detecting an abnormal state of a manufacturing process based on prediction by a model created using normal operation data.

WO 2013/011745 Japanese Patent No. 492265

  However, the methods described in Patent Literatures 1 and 2 merely predict or detect an abnormal state of the manufacturing process, and cannot estimate the cause of the abnormal state that has occurred in the manufacturing process. For this reason, it has been expected to provide a technique capable of estimating the cause of the abnormal state that has occurred in the manufacturing process.

  The present invention has been made in view of the above problems, and an object of the present invention is to provide a manufacturing process abnormal state diagnosis method and an abnormal state diagnosis device capable of estimating a cause of an abnormal state occurring in a manufacturing process. is there.

  The method for diagnosing an abnormal state of a manufacturing process according to the present invention predicts a state of the manufacturing process and / or a product being manufactured in the manufacturing process using actual values of a plurality of types of variables in the manufacturing process obtained during normal operation. A plurality of sub-models to be created, and diagnosing an abnormal state of the manufacturing process using a deviation index from a normal state of the manufacturing process calculated for each of the sub-models from a prediction error of the sub-model. In the state diagnosis method, a reliability and an attribute are defined for each of the sub-models, and when there is one or more sub-models whose magnitude of the deviation index is a predetermined value or more, the reliability and the attributes of the sub-model are used. The method includes a step of estimating a cause of an abnormal state that has occurred in the manufacturing process.

  The method for diagnosing an abnormal state of a manufacturing process according to the present invention, in the above invention, calculates a modified deviation index by weighting the deviation index of the sub-model according to the degree of reliability of the sub-model, The method includes a step of creating a deviation ranking in which the sub-models are arranged in the order of the index, and estimating a cause of an abnormal state occurring in a manufacturing process based on the deviation ranking.

  In the method of diagnosing an abnormal state of a manufacturing process according to the present invention, in the above-mentioned invention, the sub-model includes, as a first attribute, a relation between an actual value of a state of a material before processing in the manufacturing process and equipment included in the manufacturing process. Related to the set values, related to the actual value of the state of the equipment included in the manufacturing process, related to the actual value of the operation amount of the operator in the equipment included in the manufacturing process, related to the actual value of the state of the intermediate product being processed in the manufacturing process, and It is characterized in that the actual value relation of the state of the product after processing in the manufacturing process is defined, and as a second attribute, the category of the equipment included in the manufacturing process is defined.

  The method for diagnosing an abnormal state of a manufacturing process according to the present invention is characterized in that, in the above invention, when the manufacturing process is a steel process including a plurality of rolling mills, It is characterized in that a relation, a driving-side relation of the rolling mill, and a both-side relation of the operating side and the driving side of the rolling mill are defined.

  An apparatus for diagnosing an abnormal state of a manufacturing process according to the present invention predicts the state of the manufacturing process and / or the product being manufactured in the manufacturing process using actual values of a plurality of types of variables in the manufacturing process obtained during normal operation. A plurality of sub-models to be created, and diagnosing an abnormal state of the manufacturing process using a deviation index from a normal state of the manufacturing process calculated for each of the sub-models from a prediction error of the sub-model. In the state diagnosis device, a reliability and an attribute are defined for each of the sub-models, and when there is one or more sub-models whose magnitude of the deviation index is a predetermined value or more, the reliability and the attributes of the sub-model are used. It is characterized by including a cause estimating unit for estimating the cause of the abnormal state occurring in the manufacturing process.

  According to the method and the apparatus for diagnosing an abnormal state of a manufacturing process according to the present invention, since the reliability and the attribute are defined for each sub-model, the reliability of the sub-model having a deviation index of a predetermined value or more is considered. Based on the degree and the attribute, the cause of the abnormal state that has occurred in the manufacturing process can be estimated.

FIG. 1 is a block diagram showing a configuration of an abnormal state diagnosis apparatus for a manufacturing process according to an embodiment of the present invention. FIG. 2 is a conceptual diagram illustrating a configuration of a sub model stored in the sub model DB. FIG. 3 is a diagram illustrating an example of the scenario inference list stored in the scenario inference list DB. FIG. 4 is a diagram illustrating an example of a scenario inference list stored in the scenario inference list DB. FIG. 5 is a flowchart showing the flow of the abnormal state diagnosis processing of the manufacturing process according to one embodiment of the present invention. FIG. 6 is a diagram illustrating an example of the deviation ranking of the sub model.

  Hereinafter, an apparatus for diagnosing an abnormal state of a manufacturing process according to an embodiment of the present invention will be described with reference to the drawings.

〔Device configuration〕
First, a configuration of an apparatus for diagnosing an abnormal state of a manufacturing process according to an embodiment of the present invention will be described with reference to FIG.

  FIG. 1 is a block diagram showing a configuration of an abnormal state diagnosis apparatus for a manufacturing process according to an embodiment of the present invention. As shown in FIG. 1, an abnormal state diagnostic apparatus 1 for a manufacturing process according to an embodiment of the present invention is an apparatus for diagnosing an abnormal state of a manufacturing process such as a steel process, and includes an input unit 11, an output unit 12, and an external unit. The apparatus 13 includes a device 13, a storage unit 14, and a control unit 15 as main components.

  The input unit 11 is a device that receives, via an information / control system network, actual operation data to be diagnosed, which performs prediction and cause estimation based on a sub-model described later. The input unit 11 inputs the received actual operation data of the diagnosis target to the control unit 15.

  The output unit 12 includes an output device such as a display device or a printing device, and outputs various types of processing information of the control unit 15.

  The external device 13 is connected to the control unit 15 in a form capable of communicating information via an electric communication line. The external device 13 includes an operation database (operation DB) 13a. The operation DB 13a can read, through the telecommunication line, time-series data of a plurality of types of variables acquired during normal operation among time-series data of a plurality of types of variables acquired during past operations of the manufacturing process. Stored in the form.

  The storage unit 14 is configured by a storage device such as a hard disk device, and is connected to the control unit 15. The storage unit 14 stores a sub model database (sub model DB) 14a and a scenario inference list database (scenario inference list DB) 14b. The sub model DB 14a stores, as a sub model, a mathematical expression for calculating a time-series predicted value indicating the state of a manufacturing process and / or a product being manufactured in the manufacturing process. The scenario inference list DB 14b stores a scenario inference list used when estimating the cause of the abnormal state that has occurred in the manufacturing process. The details of the sub model and the scenario inference list will be described later.

  The control unit 15 is configured by an arithmetic processing device such as a CPU (Central Processing Unit), and controls the operation of the entire abnormal state diagnosis device 1 of the manufacturing process. The control unit 15 functions as the deviation index calculating unit 15a and the cause estimating unit 15b when the arithmetic processing device executes a computer program. The functions of these units will be described later.

[Sub model configuration]
Next, the configuration of the sub model stored in the sub model DB 14a will be described with reference to FIG.

  FIG. 2 is a conceptual diagram showing a configuration of a sub model stored in the sub model DB 14a. In the present invention, the sub model refers to a mathematical expression for calculating a time-series predicted value indicating a state of a manufacturing process and / or a product being manufactured in the manufacturing process. By constructing a plurality of types of sub-models, early detection of an abnormal state and estimation of the cause are easier than constructing one model in the entire manufacturing process. In FIG. 2, a plurality of sub-models are defined for all blocks, but the number is arbitrary and may or may not be different for each block. In the manufacturing process, various models have been built to produce products with the desired quality and dimensions. Such an existing model may be used as a sub model. If there are not enough submodels, a new submodel can be added by statistical processing. For example, a regression equation can be obtained using a plurality of variables other than the self obtained during normal operation of the manufacturing process, and used as a sub model. In addition, a reliability (a value that increases as the prediction error decreases) is assigned to each sub-model according to the prediction error of the sub-model during a predetermined evaluation period. If the calculated reliability is low, it is desirable to review the configuration of the sub model.

  Each sub model defines an attribute in the manufacturing process. Specifically, in the present embodiment, as shown in FIG. 2, each sub model has, as a first attribute, an actual value of a state (for example, temperature, thickness, shape, etc.) of a material before processing in a manufacturing process. (Material performance before manufacturing) related, Equipment set value included in manufacturing process (Equipment setting) related, Actual value obtained from equipment included in manufacturing process (Equipment actual) related, Equipment included in manufacturing process Related to the actual value of the operation amount of the operator (actual result of manual intervention), related to the actual value of the state of the intermediate product being processed in the manufacturing process (status result of the intermediate product), and actual value of the state of the product after processing in the manufacturing process (Product performance after manufacturing) is defined in advance. In the present embodiment, six attributes are defined as the first attribute. However, it is not always necessary to define all of these six attributes, and at least two or more of the six attributes are defined. Is also good. However, in order to perform more detailed cause estimation, it is desirable to define as many as possible.

  In each submodel, as a second attribute, equipment divisions, specifically, N (N = 1 to n) equipments arranged in order from the upstream side to the downstream side of the manufacturing process. Are defined. Further, when the manufacturing process is a steel process including a plurality of rolling mills, as a second attribute, the operating side of the rolling mill, the driving side of the rolling mill, and the operating side and the driving side of the rolling mill. Two-sided associations are included. In addition, the operation side of the rolling mill means a width direction end side of the rolling mill in which the motor for driving the rolling mill is not installed. Further, the driving side of the rolling mill means a width direction end side of the rolling mill in which the motor for driving the rolling mill is installed.

[Structure of scenario inference list]
Next, the configuration of the scenario inference list stored in the scenario inference list DB 14b will be described with reference to FIGS.

  FIGS. 3 and 4 are diagrams illustrating an example of the scenario inference list stored in the scenario inference list DB 14b. In the present invention, a scenario inference list refers to electronic data that defines rules used when estimating the cause of an abnormal state that has occurred in a manufacturing process, and is described by, for example, an if-then statement. Specifically, the scenario inference list shown in FIG. 3 is a scenario inference list in a case where the category of the equipment to which the main factor of the abnormal state of the manufacturing process belongs and the category of the equipment to which the subfactor of the abnormal state of the manufacturing process belongs are the same. In addition, the scenario inference list shown in FIG. 4 is a scenario inference list in a case where the category of the equipment to which the main factor of the abnormal state of the manufacturing process belongs and the category of the equipment to which the sub-factor of the abnormal state belongs are different. For example, in the scenario inference list shown in FIG. 3, in a certain facility i, when the main factor of the abnormal state of the manufacturing process is “facility result” and the sub-factor of the abnormal state of the manufacturing process is “manual intervention result”, An equipment abnormality has occurred in the equipment i, and an operator has operated the equipment i in order to compensate for the equipment abnormality, but an error has occurred. " In the scenario inference list shown in FIG. 4, the main factor (upstream factor) of the abnormal state of the manufacturing process in the facility i is “equipment performance”, and the secondary factor (downstream factor) of the abnormal state of the manufacturing process in the facility k. ) Is “manual intervention record”, it means that “the equipment abnormality has occurred in the equipment i and the operator has operated the equipment k to compensate for the equipment abnormality but an error has occurred” as the cause of the abnormal state of the manufacturing process. presume.

[Abnormal state diagnosis processing]
Next, a method of diagnosing an abnormal state of a manufacturing process by the abnormal state diagnosing apparatus 1 of a manufacturing process according to an embodiment of the present invention will be described with reference to FIG.

  FIG. 5 is a flowchart showing the flow of the abnormal state diagnosis processing of the manufacturing process according to one embodiment of the present invention. The flowchart illustrated in FIG. 5 starts when the input unit 11 inputs the actual operation data to the control unit 15, and the abnormal state diagnosis process proceeds to the process of step S1.

  In the processing of step S1, the deviation index calculating unit 15a reads data of a plurality of types of variables acquired from the manufacturing process at the processing target time from the operation DB 13a. Thereby, the process of step S1 is completed, and the abnormal state diagnosis process proceeds to the process of step S2.

  In the process of step S2, the deviation index calculation unit 15a uses the data of the plurality of types of variables read in the process of step S1 to change the manufacturing state of the manufacturing process at the processing target time to the manufacturing state of the manufacturing process during normal operation. Is calculated for each sub model as a deviation index. Specifically, first, the deviation index calculating unit 15a reads the data of the sub-model from the sub-model DB 14a, and substitutes the data of the actual value of the variable into the corresponding sub-model to thereby calculate the predicted value at the processing target time for each variable. Is calculated. Next, the deviation index calculation unit 15a normalizes the data of the actual value and the predicted value of a plurality of types of variables in order to normalize the difference between the absolute amount and the unit between the variables. Next, the deviation index calculation unit 15a calculates a difference value between the normalized predicted value of the variable at the processing target time and the normalized actual value as a deviation index for each sub model. Then, the deviation index calculation unit 15b calculates the reliability of the sub model (for example, 1 / (normalized prediction error of the latest section + 1). The current normalized prediction error (deviation index) = (current prediction error−model creation) The deviation index of the sub-model is weighted according to the magnitude of the prediction error average value at the time) / the prediction error standard deviation at the time of model creation)) to calculate the modified deviation index. The tables shown in FIGS. 3 and 4 and the calculation formulas of the deviation index and the reliability are merely examples, and the contents of the scenario and the calculation formula may be appropriately changed according to the manufacturing process. Thereby, the process of step S2 is completed, and the abnormal state diagnosis process proceeds to the process of step S3.

  In the process of step S3, the deviation index calculation unit 15a extracts a sub model in which the size of the corrected deviation index calculated in the process of step S2 is equal to or larger than a predetermined value (cut line), and extracts the extracted sub model. A deviation ranking arranged in the order of the magnitude of the modified deviation index is created (see FIG. 6). Thus, the process of step S3 is completed, and the abnormal state diagnosis process proceeds to a process of step S4.

  In the process of step S4, the cause estimating unit 15b reads the scenario inference list from the scenario inference list DB 14b. Then, the cause estimating unit 15b estimates the cause of the abnormal state using the deviation ranking and the scenario inference list created in the process of step S3. For example, when the sub-models having the first and second places in the deviation ranking are set as the main factor and the sub-factor of the abnormal state, respectively, the cause estimating unit 15b sets the first place in the deviation ranking shown in FIG. "Equipment performance" is estimated to be the main factor of the abnormal state, and "Equipment performance" of "Rolling mill No. 6", which is the second place in the deviation ranking shown in FIG. 6, is a secondary factor of the abnormal state. presume. In this case, the classification of the equipment corresponding to the main factor of the abnormal state is different from the classification of the equipment corresponding to the sub-factor of the abnormal state, so the cause estimating unit 15b refers to the scenario inference list shown in FIG. In the scenario where the main factor (upstream factor) is the actual performance of the equipment and the secondary factor (downstream factor) is the actual performance of the equipment, the “fault of rolling mill No. 5 or No. 6” is a scenario. Estimate as the cause of the abnormal state.

  Similarly, for example, a submodel having a deviation ranking of 1st is regarded as a main factor of an abnormal state, and a submodel having a deviation ranking of 2nd in the same equipment category as the equipment category to which the main factor belongs is a subfactor of an abnormal state. In this case, the cause estimating unit 15b estimates that the “equipment performance” of “rolling mill No. 5”, which is the first place in the deviation ranking shown in FIG. 6, is the main factor of the abnormal state, and the deviation shown in FIG. It is estimated that the “manual intervention record” of “rolling mill No. 5”, which is the fourth place in the ranking, is a secondary factor of the abnormal state. In this case, the category of the facility to which the main factor in the abnormal state belongs is the same as the category of the facility to which the sub-factor in the abnormal state belongs. Therefore, the cause estimating unit 15b refers to the scenario inference list shown in FIG. The scenario in which the main factor is the equipment performance and the sub-factor is the manual intervention performance is as follows: "A facility abnormality of the rolling mill No. 5 has occurred, and the operator has performed the rolling mill No. 5 in order to compensate for the facility abnormality. Is operated, but a mistake has occurred "as the cause of the abnormal state.

  Similarly, when the first and second submodels belonging to the category of the largest equipment in the deviation ranking are set as the main factor and the sub-factor of the abnormal state, respectively, the cause estimating unit 15b sets the deviation ranking shown in FIG. It is estimated that “Equipment performance” of “Rolling mill No. 6” which is the second place is the main factor of the abnormal state, and “Equipment of the“ Rolling mill No. 6 ”which is the third place in the deviation ranking shown in FIG. It is estimated that “setting” is a sub-factor of the abnormal state. In this case, the category of the facility to which the main factor in the abnormal state belongs is the same as the category of the facility to which the sub-factor in the abnormal state belongs. Therefore, the cause estimating unit 15b refers to the scenario inference list shown in FIG. In this case, a scenario in which the main factor is the equipment performance and the sub-factor is the equipment setting, which is a scenario when the equipment is abnormal, is estimated as the cause of the abnormal state. Thereby, the processing of step S4 is completed, and a series of abnormal state diagnosis processing ends.

  As is apparent from the above description, in the abnormal state diagnosis processing according to the embodiment of the present invention, when there is one or more sub-models in which the magnitude of the deviation index is equal to or larger than a predetermined value, Since the cause of the abnormal state generated in the manufacturing process is estimated from the reliability and the attribute of the sub model, the cause of the abnormal state generated in the manufacturing process can be estimated.

  As described above, the embodiments to which the invention made by the present inventor is applied have been described. However, the present invention is not limited by the description and the drawings that constitute a part of the disclosure of the present invention according to the present embodiment. That is, other embodiments, examples, operation techniques, and the like performed by those skilled in the art based on this embodiment are all included in the scope of the present invention.

DESCRIPTION OF SYMBOLS 1 Abnormal state diagnostic device of manufacturing process 11 Input unit 12 Output unit 13 External device 13a Operation database (operation DB)
14 storage unit 14a sub model database (sub model DB)
14b Scenario Inference List Database (Scenario Inference List DB)
15 control unit 15a departure index calculating unit 15b cause estimating unit

Claims (3)

A plurality of types of predicting a part of the state of the manufacturing process and / or a part of the state of a product being manufactured in the manufacturing process using actual values of a plurality of types of variables in the manufacturing process obtained during normal operation. Creating a sub-model, and diagnosing an abnormal state of the manufacturing process using a deviation index from a normal state of the manufacturing process calculated for each sub-model from a current prediction error of the sub-model; A diagnostic method,
For each of the sub-models, the reliability that is a value that increases as the prediction error in a predetermined period decreases, the relationship between the actual value of the material state before processing in the manufacturing process, and the set value of equipment included in the manufacturing process Related, related to the actual value of the state of the equipment included in the manufacturing process, related to the actual value of the operation amount of the operator in the equipment included in the manufacturing process, related to the actual value of the state of the intermediate product being processed in the manufacturing process, and in the manufacturing process Define a first attribute that includes at least two or more of the actual values related to the state of the product after processing, and a second attribute that indicates a category of equipment included in the manufacturing process,
A modified deviation index is calculated by weighting the deviation index of the sub-model according to the degree of reliability of the sub-model, and a deviation ranking in which the sub-models are arranged in the order of the magnitude of the modified deviation index is created. Using a scenario inference list that defines the relationship between the first and second attributes and the cause of the abnormal state of the manufacturing process, the first and second attributes defined in the submodel according to the deviation ranking A method of diagnosing an abnormal state of a manufacturing process, comprising a step of estimating a cause of an abnormal state occurring in the manufacturing process from an eye attribute.   When the manufacturing process is a steel process including a plurality of rolling mills, the sub-model includes, as a second attribute, an operating side of the rolling mill, a driving side of the rolling mill, and an operating side and driving of the rolling mill. 2. The method for diagnosing an abnormal state of a manufacturing process according to claim 1, wherein a relationship between the two sides is defined. A plurality of types of predicting a part of the state of the manufacturing process and / or a part of the state of a product being manufactured in the manufacturing process using actual values of a plurality of types of variables in the manufacturing process obtained during normal operation. Creating a sub-model, and diagnosing an abnormal state of the manufacturing process using a deviation index from a normal state of the manufacturing process calculated for each sub-model from a current prediction error of the sub-model; A diagnostic device,
For each of the sub-models, the reliability that is a value that increases as the prediction error in a predetermined period decreases, the relationship between the actual value of the material state before processing in the manufacturing process, and the set value of equipment included in the manufacturing process Related, related to the actual value of the state of the equipment included in the manufacturing process, related to the actual value of the operation amount of the operator in the equipment included in the manufacturing process, related to the actual value of the state of the intermediate product being processed in the manufacturing process, and in the manufacturing process Define a first attribute that includes at least two or more of the actual values related to the state of the product after processing, and a second attribute that indicates a category of equipment included in the manufacturing process,
A modified deviation index is calculated by weighting the deviation index of the sub-model according to the degree of reliability of the sub-model, and a deviation ranking in which the sub-models are arranged in the order of the magnitude of the modified deviation index is created. Using a scenario inference list that defines the relationship between the first and second attributes and the cause of the abnormal state of the manufacturing process, the first and second attributes defined in the submodel according to the deviation ranking An abnormal state diagnosis apparatus for a manufacturing process, comprising: a cause estimating unit for estimating a cause of an abnormal state occurring in a manufacturing process from an eye attribute.

Images