Quality prediction model generating device, quality prediction device, quality prediction model generating method, and quality prediction method

Abstract

To provide a quality prediction model generating device, a quality prediction device, a quality prediction model generating method, and a quality prediction method, that can predict a product quality with high accuracy.SOLUTION: A quality prediction model generating device comprises: performance data storage means for storing in a database performance data obtained by linking an operating condition when manufacturing a product with a quality of the manufactured product; quality prediction model construction means that uses the performance data to construct a quality prediction model which predicts the quality using the operating condition as input data and learns a structure of the input data as a feature pattern; prediction accuracy evaluation means for evaluating prediction accuracy of the quality prediction model using evaluation data; and prediction accuracy optimization means for optimizing prediction accuracy of the quality prediction model by changing the structure of the input data based on a result of the prediction accuracy evaluation obtained by the prediction accuracy evaluation means.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a quality prediction model generating device, a quality prediction device, a quality prediction model generating method, and a quality prediction method. [Background technology]

[0002] Patent Document 1 discloses a method for predicting the quality of steel products, which predicts the presence or absence of defects in a slab by using a neural network model trained from past production performance data and quality performance data of the slab. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] JP 2020-157333 A Summary of the Invention [Problem to be solved by the invention]

[0004] However, in the method of Patent Document 1, a feature discrimination layer is formed without depending on the input structure of the neural network. Therefore, it is not possible to add input structure-related ideas that would emphasize characteristic relationships that may be latent in multiple manufacturing performance data. In addition, in the method of Patent Document 1, the improvement of the model accuracy depends only on adjusting the hyperparameters of the neural network, so there is a possibility that sufficient prediction accuracy will not be achieved.

[0005] The present invention has been made in consideration of the above, and aims to provide a quality prediction model generating device, a quality prediction device, a quality prediction model generating method, and a quality prediction method that are capable of predicting product quality with high accuracy. [Means for solving the problem]

[0006] In order to solve the above-mentioned problems and achieve the object, the quality prediction model generation device of the present invention comprises: performance data storage means for storing performance data linking operating conditions for manufacturing a product with the quality of the manufactured product in a database; quality prediction model construction means for constructing, using the performance data, a quality prediction model that predicts the quality using the operating conditions as input data, the quality prediction model learning the structure of the input data as a feature pattern; prediction accuracy evaluation means for evaluating the prediction accuracy of the quality prediction model using evaluation data; and prediction accuracy optimization means for optimizing the prediction accuracy of the quality prediction model by changing the structure of the input data based on the evaluation result of the prediction accuracy obtained by the prediction accuracy evaluation means.

[0007] Furthermore, in the quality prediction model generating device of the present invention, in the above invention, the performance data storage means creates performance data linking the operating conditions and the quality of the product for each specific position of the product when the quality of the product is evaluated for each specific position of the product, and creates performance data linking the operating conditions for each manufacturing process and the quality of the product for each specific position of the product when the quality of the product is evaluated for each specific position of the product and the product is manufactured through a plurality of manufacturing processes.

[0008] In addition, in the quality prediction model generating device of the present invention, in the above invention, the structure of the input data that is changed by the prediction accuracy optimization means is the order of the input data that corresponds to the order of the operating conditions.

[0009] In addition, in the quality prediction model generating device according to the present invention, in the above invention, the quality prediction model is a neural network having a function of extracting features of the input data.

[0010] In order to solve the above-mentioned problems and achieve the above-mentioned objectives, the quality prediction device of the present invention includes a quality prediction means for predicting the quality of a manufactured product based on the quality prediction model generated by the above-mentioned quality prediction model generation device.

[0011] In order to solve the above-mentioned problems and achieve the object, a quality prediction model generating method according to the present invention is a quality prediction model generating method executed by an apparatus constructed by a computer, the method including: a performance data storing step in which a performance data storing means included in the computer stores performance data linking operation conditions for manufacturing a product with the quality of the manufactured product in a database; a quality prediction model constructing step in which a quality prediction model constructing means included in the computer constructs a quality prediction model using the performance data to predict the quality using the operation conditions as input data, the quality prediction model learning a structure of the input data as a feature pattern; a prediction accuracy evaluating step in which a prediction accuracy evaluating means included in the computer evaluates the prediction accuracy of the quality prediction model using evaluation data; and a prediction accuracy optimizing step in which a prediction accuracy optimizing means included in the computer optimizes the prediction accuracy of the quality prediction model by changing the structure of the input data based on the evaluation result of the prediction accuracy obtained in the prediction accuracy evaluating step.

[0012] In order to solve the above-mentioned problems and achieve the above-mentioned objectives, a quality prediction method according to the present invention is a quality prediction method executed by an apparatus constructed by a computer, and includes a quality prediction step of predicting the quality of a manufactured product based on a quality prediction model generated by the above-mentioned quality prediction model generation method. Effect of the Invention

[0013] In the quality prediction model generating device, quality prediction device, quality prediction model generating method, and quality prediction method according to the present invention, when constructing a quality prediction model that learns the structure of input data as a feature pattern, the structure of the input data is changed to optimize prediction accuracy. This makes it possible to generate a quality prediction model that can predict product quality with high accuracy, and to predict product quality with high accuracy using the quality prediction model. [Brief description of the drawings]

[0014] [Figure 1] FIG. 1 is a diagram showing a schematic configuration of an information processing device which realizes a quality prediction model generating device and a quality prediction device according to first and second embodiments of the present invention. [Diagram 2] FIG. 2 is a diagram showing performance data created by linking operation condition data and quality data in the quality prediction model generating device and the quality prediction device according to the first embodiment of the present invention. [Diagram 3] FIG. 3 is a diagram for explaining a procedure for searching for the order of operation condition sequences in input data in the quality prediction model generating device and the quality prediction device according to the first embodiment of the present invention. [Figure 4] FIG. 4 is a flowchart showing the steps of a quality prediction model generating method executed by the quality prediction model generating device according to the first embodiment of the present invention. [Diagram 5] FIG. 5 is a diagram showing performance data created by linking operation condition data with quality data in the quality prediction model generating device and the quality prediction device according to the second embodiment of the present invention. [Figure 6] FIG. 6 is a diagram for explaining a procedure for searching for the order of operation condition sequences in input data in the quality prediction model generating device and the quality prediction device according to the second embodiment of the present invention. [Figure 7] FIG. 7 is a flowchart showing the steps of a quality prediction model generating method executed by the quality prediction model generating device according to the second embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0015] A quality prediction model generating device (hereinafter referred to as a "model generating device"), a quality prediction device, a quality prediction model generating method (hereinafter referred to as a "model generating method"), and a quality prediction method according to the embodiments of the present invention will be described with reference to the drawings. Note that the components in the following embodiments include those that are replaceable and easy for a person skilled in the art, or those that are substantially the same.

[0016] (Embodiment 1) A model generating device, a quality predicting device, a model generating method, and a quality predicting method according to a first embodiment of the present invention will be described with reference to FIGS. 1 to 4. FIG.

[0017] [Device configuration] 1 shows an example of the configuration of an information processing device 1 for realizing a quality prediction device according to this embodiment. The information processing device 1 includes an input unit 10, a storage unit 20, a calculation unit 30, and a display unit 40. The model generation device according to this embodiment is realized by the components of the information processing device 1 excluding the quality prediction unit 33 of the calculation unit 30.

[0018] The input unit 10 is an input means for the calculation unit 30, and is realized by an input device such as a keyboard, a mouse pointer, or a numeric keypad.

[0019] The storage unit 20 is composed of recording media such as an EPROM (Erasable Programmable ROM), a hard disk drive (HDD), and removable media. Examples of removable media include a universal serial bus (USB) memory, a compact disc (CD), a digital versatile disc (DVD), and a Blu-ray (registered trademark) disc (BD). The storage unit 20 can store an operating system (OS), various programs, various tables, various databases, and the like. An operation database (DB) 21 is stored in the storage unit 20.

[0020] The operation DB 21 stores searchable performance data of products manufactured in the past. In this embodiment, the "product" may be a steel product manufactured through a plurality of processes. For example, in the case of a steel product, the "plurality of processes" may include one or more of the following processes: iron-making process, steel-making process, hot rolling process, cold rolling process, annealing process, and surface treatment process.

[0021] The "performance data" includes, for example, the operating conditions in multiple processes when a product was manufactured in the past (hereinafter referred to as "operating condition data") and the quality of the product inspected after the final process (hereinafter referred to as "quality data"). The operating condition data includes, for example, manufacturing condition data and sensor data in each process.

[0022] For example, in the case of steel products, examples of operational condition data include data on product thickness, product width, product length, steel components, casting speed, etc. in the steelmaking process, data on product thickness, product width, product length, product temperature, rolling speed, etc. in the hot rolling process, data on product thickness, product width, product length, rolling speed, etc. in the cold rolling process, data on product thickness, product width, product length, temperature in the surface treatment furnace, surface treatment speed, etc. in the surface treatment process. In addition, examples of quality data include data on the results of surface defect inspection at the end of the surface treatment process.

[0023] The calculation unit 30 is realized by a processor such as a CPU (Central Processing Unit) and a memory (main storage unit) such as a RAM (Random Access Memory) or a ROM (Read Only Memory).

[0024] The calculation unit 30 loads a program into the working area of ​​the main storage unit, executes it, and controls each component unit through the execution of the program to realize a function that meets a predetermined purpose. Through the execution of the program, the calculation unit 30 functions as a data linking unit 31, a model construction unit 32, and a quality prediction unit 33. Note that, although FIG. 1 shows an example in which the functions of each unit are realized by one computer, the specific method of realizing the functions of each unit is not particularly limited, and for example, the functions of each unit may be realized by multiple computers.

[0025] The data linking unit 31 links the operation condition data with the quality data. First, the data linking unit 31 collects, from the operation DB 21, operation condition data measured by a plurality of measuring devices during a manufacturing process involving multiple processes, and quality data of a product actually inspected after the final process.

[0026] Here, the operation condition data X during the manufacture of a certain product has, for example, the following elements:

[0027] X = [x1,x2,…,x m ,…,x M ](1≦m≦M)

[0028] However, the above "x m " indicates one actual value of one manufacturing facility or manufacturing device in process k (1≦k≦K) of the K processes when manufacturing a product through K processes from the first process to the Kth process. In other words, "x m " indicates the mth (1≦m≦M) actual value out of a total of M types of actual values ​​in all processes.

[0029] Furthermore, quality data Y of a certain product may have, for example, the following elements:

[0030] Y=[y]

[0031] Here, the above "y" indicates the quality inspection result of the product after the final process, that is, the process K out of the K processes, when manufacturing a product through K processes from the first process to the Kth process. The quality of the product may be evaluated by multiple different quality evaluations. That is, y is used for each quality category. i It is also possible to consider multiple outputs as

[0032] The data linking unit 31 treats the product (final product) manufactured in the final process as a single unit, links the operating condition data in each process with the quality data of the product inspected after the final process, and creates performance data in a tabular format, for example, as shown in Figure 2.

[0033] In the example of the performance data shown in Figure 2, data for N products are arranged in the row direction, with quality data Y and operating condition data X. m (1≦m≦M) are arranged in the column direction. The performance data shown in the figure is input data when constructing a quality prediction model described later. The data linking unit 31 stores the performance data that has been linked in this way, for example, in the operation DB 21. In this way, the data linking unit 31 functions as a performance data storage means that stores in the operation DB 21 performance data that links the operation condition data when manufacturing a product with the quality data of the manufactured product.

[0034] The model construction unit 32 constructs a quality prediction model (mathematical model) that predicts the quality of a product for any operating condition, using the performance data created by the data linking unit 31. This quality prediction model is a model that predicts the quality of a product using operating condition data as input data, and is a model that learns the structure of the input data as a feature pattern. More specifically, the quality prediction model is composed of an input part, a feature discrimination part, a prediction judgment part, and an output part.

[0035] In the input section, as shown in FIG. 2, input data having a vector structure with a plurality of operation conditions (operation condition string) as elements is input.

[0036] In the feature discrimination section, model learning is performed to capture the pattern of the input data structure features and discriminate the features. This forms a feature discrimination layer that reacts to a specific feature pattern and transmits information. This feature discrimination layer is formed in multiple layers, and transmits information from one layer to the next layer in sequence. Also, information is transmitted from the layer connected to the input part that transmits information to the layer connected to the prediction and judgment part that receives information.

[0037] The prediction and assessment part outputs quality assessment and prediction results for the input data, based on the complex patterns of multiple operating conditions that affect quality.

[0038] Specifically, in this embodiment, a quality prediction model is constructed using a Convolutional Neural Network (hereinafter referred to as "CNN"), which is a type of deep learning algorithm that has a function of extracting features of input data. This CNN is a machine learning algorithm that mimics a network structure made of nerve cells (neurons) called the neural network of the human brain and its learning mechanism. In addition, the CNN is composed of an input layer, a convolution layer, a pooling layer, a fully connected layer, and an output layer.

[0039] Input data is input to the input layer. In the convolution layer, a filter called a kernel is applied to each partial region of the input data to convolve the data and extract features to construct a feature map (also called an "element"). There may be multiple filters, in which case there will also be multiple feature maps constructed. In addition, in CNN training, the filter parameters are automatically learned.

[0040] In the pooling layer, the size of each feature map obtained by convolution is reduced. The convolution layer and the pooling layer may be configured with the same number of layers alternately. In the fully connected layer, classification is performed to classify the combination of feature maps into a specific prediction result. For example, in the case of steel products, the above-mentioned "specific prediction result" may be the presence or absence of quality defects such as surface defects.

[0041] By applying CNN, it is possible to extract a characteristic pattern from the input data, and obtain a classification that is identified with high accuracy based on the pattern as output data. In this embodiment, complex patterns of multiple operating conditions that affect quality are identified by a CNN (one-dimensional CNN) that outputs a classification probability of the occurrence or absence of quality defects for a one-dimensional input of each operating condition. In this case, a threshold value for the classification probability is set so that when the classification probability of the occurrence of quality defects is less than an arbitrary value p (0≦p≦1), the data is classified as "no quality defect," and when the classification probability is equal to or greater than an arbitrary value p, the data is classified as "quality defect."

[0042] The model construction unit 32 performs CNN learning using operation conditions in multiple processes when manufacturing a product as input data and defect occurrence probability scores for each quality category of the manufactured product as output data. Note that the "defect occurrence probability score for each quality category" refers to the classification probability regarding the presence or absence of multiple types of defects, such as "the probability of occurrence of defect A is 70%," "the probability of occurrence of defect B is 30%," etc.

[0043] The model construction unit 32 may perform CNN learning in each process by using the operation conditions of parts corresponding to multiple parts in the product manufactured in the final process as input data and the defect occurrence probability scores of multiple parts in the product manufactured in the final process as output data. This makes it possible to trace the operation conditions corresponding to a specific part of the final product back through each process and build a quality prediction model.

[0044] The data created by the data linking unit 31, i.e., the vector-structured input data input when constructing a quality prediction model, can be considered to contain, as latent information, information regarding the relationship between operating conditions that are in close proximity in the data structure of the input data.

[0045] Here, consider a case where the order of elements of a plurality of operation condition sequences, i.e., input data (input vector), has a relationship with the quality of the product. In this case, a situation may arise where the order of the operation condition sequences makes it easier for the feature discrimination layer, which reacts to a specific pattern and transmits information, to discriminate the pattern of information on the relationship between operation condition data.

[0046] Therefore, the model construction unit 32 evaluates the prediction accuracy of the quality prediction model using the evaluation data. Then, based on the evaluation result of the prediction accuracy obtained at that time, the structure of the input data is changed to optimize the prediction accuracy of the quality prediction model. Here, the above-mentioned "input data structure" specifically refers to the arrangement order of the input data corresponding to the order of the operating conditions.

[0047] The model construction unit 32 rearranges the sequence of operation conditions in the input data, for example, as shown in FIG. 3, and searches for the order of the sequence of operation conditions that will maximize the prediction accuracy. For example, in the first search, the model construction unit 32 rearranges the sequence of operation conditions in the input data, for example, as shown in FIG. m …X M The model construction unit 32 performs learning of the CNN using input data arranged in the order of "operation condition data X2...X1...X" and evaluates the prediction accuracy (see (a) of the same figure). m …X M The model construction unit 32 performs learning of the CNN using input data consisting of the order of "operation condition data X M …X1…X m …X M-1 The CNN is trained using input data consisting of the sequence of "," and the prediction accuracy is evaluated (see (c) in the same figure).

[0048] In this way, the model construction unit 32 evaluates the prediction accuracy for each order while rearranging the order of the operation condition columns in the input data. When rearranging the order of the operation condition columns, the model construction unit 32 rearranges the data in the column direction of the performance data shown in Fig. 2, that is, rearranges the operation condition data in column units (process units). Note that, for ease of explanation, in Fig. 3, some of the symbols indicating the operation condition data or quality data are shown with symbols different from those in Fig. 2.

[0049] In this embodiment, a model that outputs classification of input as a prediction result is handled. Therefore, specifically, "Accuracy" is used as an index for evaluating the prediction accuracy in the model construction unit 32. That is, the model construction unit 32 searches for the arrangement order of the operation condition sequence in the input data that maximizes the value of the evaluation formula shown in the following formula (1).

[0050]

number

[0051] In the above formula (1), TP (True Positive) is the total number of prediction results in which the quality prediction model predicts "Positive" and the correct answer is "Positive". TN (True Negative) is the total number of prediction results in which the quality prediction model predicts "Negative" and the correct answer is "Negative". FP (False Positive) is the total number of prediction results in which the quality prediction model predicts "Positive" but the correct answer is "Negative". FN (False Negative) is the total number of prediction results in which the quality prediction model predicts "Negative" but the correct answer is "Positive".

[0052] In addition, the above indexes can be evaluated each time the sequence of operating conditions in the input data is rearranged using a method called hold-out validation or a method called cross-validation.

[0053] In hold-out validation, the performance data used for learning is divided into learning data for learning a quality prediction model and evaluation data for evaluating the quality prediction model. An index is then calculated based on the results for the evaluation data. In cross-validation, the performance data used for learning is divided into K groups, and in the kth round (1≦k≦K), the kth group is used as the evaluation data, and all the other groups are used as the learning data to repeatedly perform learning and evaluation. The indexes calculated each time are then averaged to calculate the final index.

[0054] Since the number of candidates for the input data structure, i.e., the candidates for the order of the operation condition string in the input data, is equal to the factorial of the total number of items in the operation condition data, it is not realistic to calculate the value of the evaluation function for all input data structures due to the problem of combinatorial explosion. As a method for solving these optimization problems, for example, metaheuristics, genetic algorithms, mathematical programming, etc. are known, and any of them may be used.

[0055] In addition, the order of the operation conditions in the input data obtained here does not need to be the optimal solution, and a condition may be set for the calculation time, and the best solution obtained within that time may be used. Alternatively, all solutions obtained within a preset calculation time may be saved and then all of them may be output at the end.

[0056] In this way, the model construction unit 32 functions as a means for constructing a quality prediction model, and also functions as a prediction accuracy evaluation means for evaluating the prediction accuracy of the quality prediction model, and a prediction accuracy optimization means for optimizing the prediction accuracy of the quality prediction model based on the evaluation of the prediction accuracy.

[0057] The quality prediction unit 33 outputs a quality prediction result by inputting vector-structured input data consisting of a plurality of operation condition strings of a product whose quality is to be predicted, into the quality prediction model constructed by the model construction unit 32. "Output of quality prediction result" specifically means calculation of a defect occurrence probability score of the target product.

[0058] Then, the quality prediction unit 33 displays the defect occurrence probability score, which is the quality prediction result, on the display unit 40. In this case, the quality prediction unit 33 may display a circle mark in the case of a prediction result of no quality defect, and a cross mark in the case of a prediction result of the presence of a quality defect, on the display unit 40 where information on the target product can be viewed, or may display a statement such as "The target product is predicted to have a quality defect" only in the case of a prediction result of the presence of a quality defect.

[0059] The display unit 40 is realized by a display device such as an LCD display, a CRT display, etc. Based on a display signal input from the calculation unit 30, the display unit 40 displays, for example, the quality prediction result (defect occurrence probability score) by the quality prediction unit 33 in the form of characters, figures, etc.

[0060] [Model generation method] A model generation method by the model generation device according to the first embodiment will be described with reference to Fig. 4. The quality prediction method includes a data collection step (step S1), a data linking step (step S2), and a model construction step (step S3).

[0061] In the data collection step, the data linking unit 31 collects operation condition data and quality data for multiple processes from the operation DB 21 (step S1). Then, in the data linking step, the data linking unit 31 creates performance data by linking the operation condition data with the quality data (step S2). Then, in the model construction step, the model construction unit 32 constructs a quality prediction model in which the operation conditions for multiple processes when manufacturing a product are used as input data and the defect occurrence probability score for each quality category of the manufactured product is used as output data (step S3).

[0062] In the model construction step, a prediction accuracy evaluation step and a prediction accuracy optimization step are performed when constructing a quality prediction model. In the prediction accuracy evaluation step, the model construction unit 32 evaluates the prediction accuracy of the quality prediction model using evaluation data. In the prediction accuracy optimization step, the model construction unit 32 optimizes the prediction accuracy of the quality prediction model by changing the structure of the input data (specifically, the order of the input data corresponding to the order of the operating conditions) based on the evaluation result of the prediction accuracy obtained in the prediction accuracy evaluation step.

[0063] [Quality prediction method] In the quality prediction method by the quality prediction device according to the first embodiment, the quality prediction unit 33 performs a quality prediction step of predicting the quality of a manufactured product based on the quality prediction model generated by the above-mentioned model generation method. After this quality prediction step, the quality prediction unit 33 may perform a prediction result display step of displaying the quality prediction result on the display unit 40.

[0064] (Embodiment 2) A model generating device, a quality predicting device, a model generating method, and a quality predicting method according to a second embodiment of the present invention will be described with reference to FIG. 1 and FIGS. 5 to 7.

[0065] [Device configuration] The configurations of the model generating device and the quality predicting device according to this embodiment are the same as those of the first embodiment (see FIG. 1). Therefore, in the following, specific processing contents of the data linking unit 31, the model constructing unit 32, and the quality predicting unit 33 will be described, and descriptions of other configurations will be omitted.

[0066] The data linking unit 31 links the operation condition data with the quality data. First, the data linking unit 31 collects, from the operation DB 21, operation condition data measured by a plurality of measuring devices during a manufacturing process involving multiple processes, and quality data of a product actually inspected after the final process.

[0067] Here, the operation condition data X during the manufacture of a certain product has elements such as those shown in the following formula (2), for example.

[0068]

number

[0069] However, "x" in the above formula (2) mn " indicates the m-th (1≦m≦M) value among the first to M-th values ​​divided for each predetermined unit length (for example, 1 m) as manufacturing condition information for a specific position (area) in a specific direction of the product (for example, the length direction of the product). mn " indicates the nth (1≦n≦N) value of a set of N types of setting values ​​or output values ​​of a number of manufacturing facilities or manufacturing devices that exist throughout the multiple processes for manufacturing a product, arranged from the first to the Nth. Or, "x mn " indicates the nth (1≦n≦N) value out of a set of N types of measurement values ​​obtained by measuring the operational state, arranged from the first to the Nth.

[0070] Furthermore, quality data Y of a certain product may have, for example, the following elements:

[0071] Y=[y]

[0072] Here, the above "y" indicates the quality inspection result of the product after the final process, that is, the process K out of the K processes, when manufacturing a product through K processes from the first process to the Kth process. The quality of the product may be evaluated by multiple different quality evaluations. That is, y is used for each quality category. i It is also possible to consider multiple outputs as a single output. Furthermore, the quality data of the final product is obtained by tallying up the quality inspected in any length unit in a specific direction of the product for each predetermined length (e.g. 100 m) and giving a judgment result based on the tallying up results. For example, if there is one or more parts detected as defective quality within the quality of the predetermined length, the judgment result for that predetermined length is "defective quality present," and conversely, if there is no part detected as defective quality within the quality of the predetermined length, the judgment result for that predetermined length is "no defective quality."

[0073] The data linking unit 31 creates input data to be input when creating a quality prediction model, which will be described later. This input data is performance data (two-dimensional data) in which one or more pieces of operating condition data in each process are linked to quality data of the final product inspected after the final process for each specific position in a specific direction of the final product after the final process, and is composed of multiple two-dimensional matrices, for example, as shown in FIG.

[0074] The two-dimensional data shown in Fig. 5 is composed of data D1, D2, and D3, which are two-dimensional matrices. The columns of the data D1, D2, and D3 are arranged with quality data Y and operation condition data for each process. The rows of the data D1, D2, and D3 are arranged with data related to a specific position (specific area) in a specific direction of the product.

[0075] The above-mentioned "data relating to a specific position in a specific direction of a product" may be, for example, a number or coordinates indicating a position when a final product is divided in a specific direction (for example, the length direction). Two-dimensional data such as the example shown in FIG. 5 is input data when constructing a quality prediction model, which will be described later. Furthermore, the data linking unit 31 stores the thus created performance data (two-dimensional data) in, for example, the operation DB 21. In this way, the data linking unit 31 functions as performance data storage means that stores in the operation DB 21 performance data that links operation condition data when manufacturing a product with quality data of the manufactured product.

[0076] The data D1, D2, and D3 shown in FIG. 5 may be data relating to the same product (e.g., a coil) or may be data relating to different products. When the data D1, D2, and D3 are data relating to the same product, the data D1 is, for example, data in which the operation condition data for each process up to the product position "1 to 100" is arranged in the row direction. The data D2 is data in which the operation condition data for each process up to the product position "101 to 200" is arranged in the row direction. The data D3 is data in which the operation condition data for each process up to the product position "201 to 300" is arranged in the row direction.

[0077] The operation condition data in the length direction of the final product after the final process, which is used by the data linking unit 31 when creating the two-dimensional data, is collected for each predetermined unit length, and M in the above formula (2) is set to a predetermined length. For example, if the "unit length" is set to 1m and the "predetermined length" is set to 100m, then "M=100" is obtained, and two-dimensional data is created in a matrix format for every 100m in the length direction of the final product, in other words, for every 100 data rows. In this case, for example, the length corresponding to A in FIG. 5 is 1m, and the length corresponding to B is 100m.

[0078] The values ​​in the "unit length" are treated as the average value of the observed data. Also, some of the items may include items that take the maximum value or items that take the minimum value. The "unit length" is characterized in that the unit length of the steel plate manufactured in the final process has a minimum unit limit according to the data collection period in the length direction of the quality data required to construct a quality prediction model. Also, the "unit length" is preferably a unit longer than the data collection period corresponding to the unit length of the final process for the operation condition data in the length direction obtained in the intermediate process, and is set individually according to the data collection period of the target of prediction.

[0079] Moreover, the "predetermined length" is preferably a small value in order to specify the position in the coil as precisely as possible. On the other hand, in order to achieve high quality prediction performance, a minimum length that expresses the characteristics of the fluctuations in the operating condition data is necessary. Therefore, the "predetermined length" has an appropriate value according to the requirements for position specification and the characteristics of the input data, and is set individually according to the target and purpose of the prediction.

[0080] In this way, the data linking unit 31 links the operating condition data and quality data in each process for each predetermined unit length (e.g., 1 m) in the longitudinal direction of the final product, for the quality of each specified length (e.g., 100 m), to create two-dimensional data.

[0081] That is, when the quality of a product is evaluated for each specific position of the product, the data linking unit 31 creates performance data linking the operating condition data and the quality data for each specific position of the product. When the quality of a product is evaluated for each specific position of the product and the product is manufactured through multiple manufacturing processes, the data linking unit 31 creates performance data linking the operating condition data and the quality data for each manufacturing process for each specific position of the product.

[0082] The model construction unit 32 constructs a quality prediction model (mathematical model) that predicts the quality of a product for any operating condition, using the performance data created by the data linking unit 31. This quality prediction model is a model that predicts the quality of a product using operating condition data as input data, and is a model that learns the structure of the input data as a feature pattern. More specifically, the quality prediction model is composed of an input part, a feature discrimination part, a prediction judgment part, and an output part.

[0083] In the input section, as shown in FIG. 5, input data (two-dimensional data) having a two-dimensional matrix structure with the position coordinates of the product in a specific direction (e.g., the length direction of the product) in the row direction and each operating condition in the column direction is input.

[0084] In the feature discrimination section, model learning is performed to capture patterns of two-dimensional features of input data and discriminate features. This forms a feature discrimination layer that reacts to a specific two-dimensional pattern and transmits information. This feature discrimination layer is formed in multiple layers, and transmits information from one layer to the next layer in sequence. Also, information is transmitted from the layer connected to the input part that transmits information to the layer connected to the prediction and judgment part that receives information.

[0085] The prediction and assessment part outputs quality assessment prediction results for the input of two-dimensional data, based on the complex patterns of multiple operating conditions that affect quality.

[0086] In this embodiment, a quality prediction model is constructed using CNN, which is a type of deep learning algorithm that has a function of extracting features of input data, as in embodiment 1. CNN is composed of an input layer, a convolution layer, a pooling layer, a fully connected layer, and an output layer, and the processing content in each layer is the same as in embodiment 1.

[0087] By applying CNN, it is possible to extract a characteristic pattern from the input data, and obtain a classification that is identified with high accuracy based on the pattern as output data. In this embodiment, complex patterns of multiple operating conditions that affect quality are identified by a CNN (two-dimensional CNN) that outputs a classification probability of the presence or absence of quality defects for a two-dimensional input of each operating condition. In this case, a threshold value for the classification probability is set so that when the classification probability of the occurrence of quality defects is less than an arbitrary value p (0≦p≦1), the data is classified as "no quality defect," and when the classification probability is equal to or greater than an arbitrary value p, the data is classified as "quality defect."

[0088] The model construction unit 32 performs CNN learning using operation conditions in multiple processes when manufacturing a product as input data and defect occurrence probability scores for each quality category of the manufactured product as output data. Note that the "defect occurrence probability score for each quality category" refers to the classification probability regarding the presence or absence of multiple types of defects, such as "the probability of occurrence of defect A is 70%," "the probability of occurrence of defect B is 30%," etc.

[0089] The data created by the data linking unit 31, i.e., the two-dimensional matrix data input when constructing a quality prediction model, can be considered to contain, as latent information, information on fluctuations in operational conditions for a specific direction of the product and information on the relationship between adjacent operational conditions in the data structure of the input data. Therefore, by making the input data two-dimensional and increasing the amount of information on operational information for a specific direction of the product, it becomes easier to grasp patterns of operational fluctuations related to the quality of the product.

[0090] Here, consider a case where a sequence of multiple operation conditions, i.e., the order of each element of the input data (two-dimensional data), has a relationship with the quality of the product. In this case, a situation may arise where the sequence of the operation conditions makes it easier for the feature discrimination layer, which reacts to a specific two-dimensional pattern and transmits information, to discriminate the pattern of information on the relationship between the operation condition data.

[0091] Therefore, the model construction unit 32 evaluates the prediction accuracy of the quality prediction model using the evaluation data. Then, based on the evaluation result of the prediction accuracy obtained at that time, the structure of the input data is changed to optimize the prediction accuracy of the quality prediction model. Here, the above-mentioned "input data structure" specifically refers to the arrangement order of the input data corresponding to the order of the operating conditions.

[0092] The model constructing unit 32 rearranges the sequence of operation conditions in the input data, for example as shown in FIG. 6, and searches for the order of the sequence of operation conditions that maximizes the prediction accuracy.

[0093] For example, in the first search, the model construction unit 32 may 11~M1 …x 12~M2 …x 1n~Mn …x 1N~MN The CNN is trained using input data (two-dimensional data) consisting of the order of "operation condition data x ", and the prediction accuracy is evaluated (see (a) of the same figure). 12~M2 …x 11~M1 …x 1n~Mn …x 1N~MN The CNN is trained using input data (two-dimensional data) consisting of the order of "operation condition data x 1N~MN …x 11~M1 …x 1n~Mn …x 1N-1~MN-1 The CNN is trained using input data (two-dimensional data) consisting of a sequence of "," and the prediction accuracy is evaluated (see (c) in the same figure).

[0094] For example, in (b) of Fig. 6, for data D1 (front side of the page), the order of the operation condition column in the first column from the left and the operation condition column in the second column from the left are swapped with respect to (a) of the same figure, and a similar swap is performed for data D2 (rear side of the page). That is, for data D2, the order of the operation condition column in the first column from the left and the operation condition column in the second column from the left are swapped as with data D1. When swapping the order of the input data, the model construction unit 32 also performs a similar swap for each data of the two-dimensional data constituting the input data.

[0095] In this way, the model construction unit 32 evaluates the prediction accuracy for each order while rearranging the order of the operation condition columns in the input data. When rearranging the order of the operation condition columns, the model construction unit 32 rearranges the operation condition data in the column direction of the performance data shown in Fig. 5, that is, in column units (process units).

[0096] In this embodiment, a model that outputs classification of input as a prediction result is handled. Therefore, specifically, "Accuracy" is used as an index for evaluating the prediction accuracy in the model construction unit 32. That is, the model construction unit 32 searches for the arrangement order of the operation condition sequence in the input data that maximizes the value of the evaluation formula shown in the following formula (1).

[0097]

number

[0098] In the above formula (1), TP (True Positive) is the total number of prediction results in which the quality prediction model predicts "Positive" and the correct answer is "Positive". TN (True Negative) is the total number of prediction results in which the quality prediction model predicts "Negative" and the correct answer is "Negative". FP (False Positive) is the total number of prediction results in which the quality prediction model predicts "Positive" but the correct answer is "Negative". FN (False Negative) is the total number of prediction results in which the quality prediction model predicts "Negative" but the correct answer is "Positive".

[0099] In addition, the above indexes can be evaluated each time the sequence of operating conditions in the input data is rearranged using a method called hold-out validation or a method called cross-validation.

[0100] In hold-out validation, the performance data used for learning is divided into learning data for learning a quality prediction model and evaluation data for evaluating the quality prediction model. An index is then calculated based on the results for the evaluation data. In cross-validation, the performance data used for learning is divided into K groups, and in the kth round (1≦k≦K), the kth group is used as the evaluation data, and all the other groups are used as the learning data to repeatedly perform learning and evaluation. The indexes calculated each time are then averaged to calculate the final index.

[0101] Since the number of candidates for the input data structure, i.e., the candidates for the order of the operation condition string in the input data, is equal to the factorial of the total number of items in the operation condition data, it is not realistic to calculate the value of the evaluation function for all input data structures due to the problem of combinatorial explosion. As a method for solving these optimization problems, for example, metaheuristics, genetic algorithms, mathematical programming, etc. are known, and any of them may be used.

[0102] In addition, the order of the operation conditions in the input data obtained here does not need to be the optimal solution, and a condition may be set for the calculation time, and the best solution obtained within that time may be used. Alternatively, all solutions obtained within a preset calculation time may be saved and then all of them may be output at the end.

[0103] In this way, the model construction unit 32 functions as a means for constructing a quality prediction model, and also functions as a prediction accuracy evaluation means for evaluating the prediction accuracy of the quality prediction model, and a prediction accuracy optimization means for optimizing the prediction accuracy of the quality prediction model based on the evaluation of the prediction accuracy.

[0104] The quality prediction unit 33 outputs a quality prediction result by inputting two-dimensional input data consisting of a plurality of operation condition strings for each specific position in a specific direction of a product whose quality is to be predicted, into the quality prediction model constructed by the model construction unit 32. "Output of quality prediction result" specifically means calculation of a defect occurrence probability score of the target product.

[0105] Then, the quality prediction unit 33 displays the defect occurrence probability score, which is the quality prediction result, on the display unit 40. In this case, the quality prediction unit 33 may display a circle mark in the case of a prediction result of no quality defect, and a cross mark in the case of a prediction result of the presence of a quality defect, on the display unit 40 where information on the target product can be viewed, or may display a statement such as "The target product is predicted to have a quality defect" only in the case of a prediction result of the presence of a quality defect.

[0106] [Model generation method] A model generation method by the model generation device according to the second embodiment will be described with reference to Fig. 7. The quality prediction method includes a data collection step (step S11), a data linking step (step S12), and a model construction step (step S13).

[0107] In the data collection step, the data linking unit 31 collects operation condition data and quality data for multiple processes from the operation DB 21 (step S11). Then, in the data linking step, the data linking unit 31 creates performance data by linking the operation condition data and the quality data as two-dimensional data (step S12). Then, in the model construction step, the model construction unit 32 constructs a quality prediction model in which the operation conditions for multiple processes when manufacturing a product are used as input data and the defect occurrence probability score for each quality category of the manufactured product is used as output data (step S13).

[0108] In the model construction step, a prediction accuracy evaluation step and a prediction accuracy optimization step are performed when constructing a quality prediction model. In the prediction accuracy evaluation step, the model construction unit 32 evaluates the prediction accuracy of the quality prediction model using evaluation data. In the prediction accuracy optimization step, the model construction unit 32 optimizes the prediction accuracy of the quality prediction model by changing the structure of the input data (specifically, the order of the input data corresponding to the order of the operating conditions) based on the evaluation result of the prediction accuracy obtained in the prediction accuracy evaluation step.

[0109] [Quality prediction method] In the quality prediction method by the quality prediction device according to the second embodiment, the quality prediction unit 33 performs a quality prediction step of predicting the quality of a manufactured product based on the quality prediction model generated by the above-mentioned model generation method. After this quality prediction step, the quality prediction unit 33 may perform a prediction result display step of displaying the quality prediction result on the display unit 40.

[0110] In the quality prediction model generating device, quality prediction device, quality prediction model generating method, and quality prediction method according to the above-described embodiments, when constructing a quality prediction model that learns the structure of input data as a feature pattern, the structure of the input data is changed to optimize prediction accuracy. This makes it possible to generate a quality prediction model that can predict product quality with high accuracy, and to predict product quality with high accuracy using the quality prediction model.

[0111] The quality prediction model generating device, the quality prediction device, the quality prediction model generating method, and the quality prediction method according to the present invention have been specifically described above using the mode for carrying out the invention and examples, but the gist of the present invention is not limited to these descriptions and must be broadly interpreted based on the claims. Furthermore, it goes without saying that various changes, modifications, etc. based on these descriptions are also included in the gist of the present invention. [Explanation of symbols]

[0112] 1. Information processing device 10 Input section 20 Memory section 21 Operation DB 30 Arithmetic section 31 Data Linking Department 32 Model Construction Department 33 Quality Prediction Department 40 Display section

Claims

1. A performance data storage means for storing performance data in a database, the performance data correlating the operating conditions when manufacturing a product with the quality of the manufactured product; a quality prediction model construction means for constructing a quality prediction model for predicting the quality using the operational conditions as input data, the quality prediction model learning a structure of the input data as a feature pattern, by using the performance data; a prediction accuracy evaluation means for evaluating the prediction accuracy of the quality prediction model using evaluation data; a prediction accuracy optimization means for optimizing the prediction accuracy of the quality prediction model by changing a structure of the input data based on the evaluation result of the prediction accuracy obtained by the prediction accuracy evaluation means; Equipped with The performance data storage means includes: When the quality of the product is evaluated for each specific position of the product, performance data is created that links the operating conditions and the quality of the product for each specific position of the product; When the quality of the product is evaluated for each specific position of the product and the product is manufactured through a plurality of manufacturing processes, performance data is created for each specific position of the product, linking the operating conditions for each manufacturing process with the quality of the product. A device for generating quality prediction models.

2. The structure of the input data changed by the prediction accuracy optimization means is an arrangement order of the input data corresponding to an order of the operation conditions. The quality prediction model generating device according to claim 1 .

3. The quality prediction model is a neural network having a function of extracting features of the input data. A quality prediction model generating device according to claim 1 or 2.

4. A quality prediction device comprising: a quality prediction means for predicting the quality of a manufactured product based on a quality prediction model generated by the quality prediction model generation device according to any one of claims 1 to 3.

5. A method for generating a quality prediction model executed by a computer-implemented device, comprising: a performance data storage step in which performance data storage means included in the computer stores performance data in a database linking operation conditions when a product is manufactured with the quality of the manufactured product; a quality prediction model construction step in which a quality prediction model construction means included in the computer constructs, using the performance data, a quality prediction model that predicts the quality using the operation conditions as input data, the quality prediction model learning a structure of the input data as a feature pattern; a prediction accuracy evaluation step in which a prediction accuracy evaluation means included in the computer evaluates the prediction accuracy of the quality prediction model using evaluation data; a prediction accuracy optimization step in which a prediction accuracy optimization means included in the computer optimizes the prediction accuracy of the quality prediction model by changing a structure of the input data based on the evaluation result of the prediction accuracy obtained in the prediction accuracy evaluation step; Including, The performance data storage step includes: When the quality of the product is evaluated for each specific position of the product, performance data is created that links the operating conditions and the quality of the product for each specific position of the product; When the quality of the product is evaluated for each specific position of the product and the product is manufactured through a plurality of manufacturing processes, performance data is created for each specific position of the product, linking the operating conditions for each manufacturing process with the quality of the product. How to generate quality prediction models.

6. A quality prediction method implemented by a computer-implemented apparatus, comprising: A quality prediction method comprising a quality prediction step of predicting quality of a manufactured product based on the quality prediction model generated by the quality prediction model generating method according to claim 5.

Images