WO2022158066A1 - 製造条件最適化装置、プログラムおよび製造条件最適化方法 - Google Patents
製造条件最適化装置、プログラムおよび製造条件最適化方法 Download PDFInfo
- Publication number
- WO2022158066A1 WO2022158066A1 PCT/JP2021/039329 JP2021039329W WO2022158066A1 WO 2022158066 A1 WO2022158066 A1 WO 2022158066A1 JP 2021039329 W JP2021039329 W JP 2021039329W WO 2022158066 A1 WO2022158066 A1 WO 2022158066A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- quality
- manufacturing
- product
- change
- amount
- Prior art date
Links
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 199
- 238000005457 optimization Methods 0.000 title claims abstract description 79
- 238000000034 method Methods 0.000 title claims description 33
- 238000007689 inspection Methods 0.000 claims abstract description 65
- 238000012545 processing Methods 0.000 claims abstract description 35
- 230000008859 change Effects 0.000 claims description 84
- 238000009826 distribution Methods 0.000 claims description 30
- 230000008569 process Effects 0.000 claims description 15
- 238000003062 neural network model Methods 0.000 claims description 4
- 230000002068 genetic effect Effects 0.000 claims description 3
- 238000002922 simulated annealing Methods 0.000 claims description 3
- 230000006870 function Effects 0.000 description 31
- 238000010586 diagram Methods 0.000 description 9
- 238000005259 measurement Methods 0.000 description 7
- 230000004048 modification Effects 0.000 description 7
- 238000012986 modification Methods 0.000 description 7
- 239000011159 matrix material Substances 0.000 description 5
- 238000004891 communication Methods 0.000 description 3
- 238000012417 linear regression Methods 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 230000001131 transforming effect Effects 0.000 description 3
- 230000007547 defect Effects 0.000 description 2
- 230000002950 deficient Effects 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000007477 logistic regression Methods 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012804 iterative process Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 238000000491 multivariate analysis Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- 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
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/418—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM]
- G05B19/41875—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM] characterised by quality surveillance of production
-
- 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
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/418—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM]
- G05B19/41865—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM] characterised by job scheduling, process planning, material flow
-
- 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
- G05B19/00—Programme-control systems
- G05B19/02—Programme-control systems electric
- G05B19/418—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM]
- G05B19/41885—Total factory control, i.e. centrally controlling a plurality of machines, e.g. direct or distributed numerical control [DNC], flexible manufacturing systems [FMS], integrated manufacturing systems [IMS] or computer integrated manufacturing [CIM] characterised by modeling, simulation of the manufacturing system
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/04—Manufacturing
-
- 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
- G05B2219/00—Program-control systems
- G05B2219/30—Nc systems
- G05B2219/31—From computer integrated manufacturing till monitoring
- G05B2219/31372—Mes manufacturing execution system
-
- 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
- G05B2219/00—Program-control systems
- G05B2219/30—Nc systems
- G05B2219/32—Operator till task planning
- G05B2219/32194—Quality prediction
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/02—Total factory control, e.g. smart factories, flexible manufacturing systems [FMS] or integrated manufacturing systems [IMS]
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Definitions
- the present invention relates to a manufacturing condition optimization device, a program, and a manufacturing condition optimization method for optimizing manufacturing conditions in manufacturing equipment.
- the plant operating condition optimization system described in Patent Document 1 includes operating state data acquisition means for acquiring operating state data indicating the operating state of the plant measured by a plurality of sensors, and sensors provided in the plant. an operation index data acquisition unit for acquiring operation index data for evaluating plant operation, which is obtained based on the operation state data acquired by the operation state data acquisition unit; and the operation state data acquisition unit.
- measurement data recording means for recording the measured data in a data recording unit by associating the obtained driving state data and the driving index data obtained by the driving index data acquisition means with each other based on a predetermined item as a set of measurement data; Then, based on the plurality of sets of measurement data recorded in the data recording unit, a predetermined multivariate analysis is performed using the driving state variables representing the driving state data side as explanatory variables and the driving index variables representing the driving index data side as objective variables.
- a regression model creating means for creating a regression model by performing and an inverse transforming means for estimating explanatory variables from the components by a method corresponding to the component transforming means, which is created by the regression model creating means
- Driving indicator variable optimization means for obtaining driving state variables for optimizing driving indicator variables based on a regression model, the purpose predicted by the predicting means while satisfying constraints on explanatory variables estimated by the inverse transforming means. It is characterized by obtaining explanatory variable values when optimizing an evaluation function relating to variables, and using the explanatory variable values as the optimum operating conditions.
- the product quality depends on the manufacturing conditions of the product. Even if the proper manufacturing conditions are set, the quality may change due to disturbances that are difficult to control completely, such as temperature and material viscosity, and the yield of the product may decrease. The same is true for disturbances that are not measured or monitored in the manufacturing environment. Therefore, it is required to adjust manufacturing conditions in real time during manufacturing of products.
- the operating condition optimization system described in Patent Literature 1 does not anticipate changes due to disturbances and adjust manufacturing conditions in real time. For this reason, optimization corresponding to uncontrollable changes in disturbance cannot be performed.
- the present invention has been made in view of such a background, and is a manufacturing condition optimization apparatus, program, and manufacturing method that enable improvement of the manufacturing index by adjusting the manufacturing conditions of the product even when there is a disturbance.
- An object of the present invention is to provide a condition optimization method.
- (2) further comprising a quality variation estimation unit for estimating a variation in the quality of the product manufactured by the manufacturing equipment from the variation in the manufacturing conditions, wherein the yield estimation unit is configured to: estimating the probability that the product will pass the inspection by the inspection equipment from the estimated amount of change in the quality of the product, and calculating the yield of the product having the quality that passes the inspection by the inspection equipment from the probability of passing the inspection;
- the production condition optimization device according to (1) for estimating
- a regression model generation unit is further provided, and the regression model generation unit generates the above-mentioned A regression model is generated for estimating the amount of change in the quality of the product from the amount of change in the manufacturing conditions, and the quality change amount estimating unit uses the regression model to estimate the amount of change in the quality of the product from the amount of change in the manufacturing conditions.
- the manufacturing condition optimization device according to (2) for estimating
- the optimization processing unit maximizes the yield using at least one of an optimization algorithm that does not use derivatives, a local search method, a simulated annealing method, a tabu search method, and a genetic algorithm.
- the production condition optimization apparatus according to (1) which calculates the amount of change in the production conditions.
- the yield estimating unit when estimating the probability that the product will pass the inspection by the inspection facility from the amount of change in quality, based on the quality of a predetermined number of products manufactured most recently,
- the production condition optimization device according to (2) which estimates the probability that the product will pass the inspection by the inspection equipment.
- the yield estimating unit obtains a probability distribution of the quality of the product from the amount of change in the quality of the product, and estimates the probability that the product passes the inspection by the inspection equipment based on the probability distribution ( 2) The production condition optimization device described in 2).
- a method for optimizing manufacturing conditions for a manufacturing condition optimizing apparatus comprising a step of estimating the yield of a product having a quality that passes an inspection by an inspection equipment when the manufacturing conditions for the product by the manufacturing equipment are changed. and calculating the amount of change in the manufacturing conditions that maximizes the yield.
- the present invention it is possible to provide a manufacturing condition optimization device, a program, and a manufacturing condition optimization method that enable improvement of the manufacturing index by adjusting the manufacturing conditions of products even when there is disturbance.
- FIG. 1 is an overall configuration diagram of a manufacturing system according to an embodiment; FIG. It is a figure for demonstrating the process outline
- 1 is a functional block diagram of a manufacturing condition optimization device according to this embodiment;
- FIG. 4 is a data configuration diagram of learning data according to the embodiment;
- FIG. 4 is a data configuration diagram of manufacturing control data according to the embodiment;
- FIG. 4 is a flowchart of regression model generation processing according to the present embodiment; 4 is a flowchart of objective function processing executed by an objective function processing unit according to the embodiment; 4 is a flowchart of optimization processing according to the embodiment;
- FIG. 1 is an overall configuration diagram of a manufacturing system 10 according to this embodiment.
- the manufacturing system 10 includes a manufacturing facility 410 , an inspection facility 420 , a manufacturing control device 430 and a manufacturing condition optimization device 100 .
- Manufacturing equipment 410 , inspection equipment 420 , manufacturing control device 430 , and manufacturing condition optimization device 100 can communicate with each other via network 499 .
- the manufacturing equipment 410 manufactures the products 460 under the manufacturing conditions set by the manufacturing control device 430 .
- Manufacturing facility 410 transmits individual manufacturing conditions for product 460 to manufacturing control device 430 .
- Product 460 is transported to inspection facility 420 for inspection.
- a product 470 that has passed the inspection is shipped as a non-defective product.
- Products 480 that fail inspection are defective and discarded.
- the inspection facility 420 measures one or more qualities (quality items, inspection items) of each product 460, and if all the measured values are within the respective criteria (between the upper limit and the lower limit), The product shall pass.
- the inspection facility 420 transmits the quality measurement value (simply referred to as quality or quality value) of each product 460 to the manufacturing control device 430 .
- the manufacturing control device 430 stores manufacturing control data 440 (see FIG. 5 described later) that stores individual manufacturing conditions and quality of the product 460 .
- the manufacturing condition optimization device 100 obtains manufacturing conditions that maximize the yield of the product 460 based on the manufacturing control data 440 . This makes it possible to improve the manufacturing index.
- FIG. 2 is a diagram for explaining the outline of processing of the manufacturing condition optimization device 100 according to this embodiment.
- a regression model 150 is generated from learning data 140 .
- the learning data 140 is a set of the amount of change in the manufacturing conditions in the manufacturing equipment 410 (see FIG. 1) and the amount of change in the quality of the product 460 in the inspection equipment 420 when the manufacturing conditions are changed by the amount of change. data.
- the regression model 150 is a model for estimating the amount of change in the quality of the product 460 (the amount of quality change) from the amount of change in the manufacturing conditions (the amount of change in the manufacturing conditions), and is, for example, a linear regression model.
- the manufacturing condition optimization apparatus 100 uses the amount of change in the manufacturing conditions as an input (variable) and uses a yield estimation function for estimating the yield of the product 460 as the objective function to solve the optimization problem of maximizing the yield.
- the yield estimation function uses the regression model 150 to calculate an estimated value of quality change from the input manufacturing condition change, obtains an estimated quality from the quality change, and estimates the yield.
- the solution to this optimization problem is the variation in manufacturing conditions that maximizes yield.
- the manufacturing condition optimization device 100 transmits this manufacturing condition change amount to the manufacturing management device 430 .
- the manufacturing control device 430 can maximize the yield of the product 460 by setting the manufacturing conditions of the manufacturing equipment 410 according to the amount of change in the manufacturing conditions.
- FIG. 3 is a functional block diagram of the manufacturing condition optimization device 100 according to this embodiment.
- Manufacturing condition optimization apparatus 100 includes control unit 110 , storage unit 130 , communication unit 170 , and input/output unit 180 .
- Communication unit 170 transmits and receives communication data to and from other devices including manufacturing control device 430 .
- User interface devices such as a display, a keyboard, and a mouse are connected to the input/output unit 180 .
- the storage unit 130 is composed of ROM (Read Only Memory), RAM (Random Access Memory), SSD (Solid State Drive), and the like.
- the storage unit 130 stores a program 131 , learning data 140 (see FIG. 4 described later), and a regression model 150 .
- the program 131 describes procedures for a regression model generation process (see FIG. 6 described later), an optimization process, and an objective function (yield estimation function) process (see FIG. 7 described later).
- FIG. 4 is a data configuration diagram of the learning data 140 according to this embodiment.
- the learning data 140 is, for example, tabular data, and one row (record) includes columns (attributes) of manufacturing condition variation 141 and quality variation 142 .
- the manufacturing condition change amount 141 indicates the amount of change when the manufacturing conditions in the manufacturing facility 410 are changed.
- the manufacturing conditions include one or more items, and the manufacturing condition variation 141 indicates the variation of the one or more items.
- the quality change amount 142 indicates the quality change amount of the product 460 in the inspection equipment 420 when the manufacturing conditions indicated by the manufacturing condition change amount 141 are changed.
- Quality variation 142 includes variation in one or more qualities (inspection measurements, quality values) for one or more products 460 .
- control unit 110 includes a CPU (Central Processing Unit), and includes a learning data generation unit 111, a regression model generation unit 112, an optimization processing unit 113, an objective function processing unit 114, and a quality variation estimation unit. It includes a unit 115 , a quality estimator 116 and a yield estimator 117 .
- the learning data generator 111 acquires manufacturing control data 440 (see FIG. 5 described later) stored in the manufacturing control device 430 and generates learning data 140 (see FIG. 4).
- FIG. 5 is a data configuration diagram of the manufacturing control data 440 according to this embodiment.
- the manufacturing control data 440 is, for example, tabular data, and one row (record) includes product identification information 441 (described as product ID (identifier) in FIG. 5), manufacturing date and time 442, manufacturing conditions 443, and quality 444. , and columns (attributes) of inspection results 445 .
- the product identification information 441 is information for identifying each product 460, such as a serial number.
- the date and time of manufacture 442 is the date and time when the product 460 was manufactured.
- the manufacturing conditions 443 are the manufacturing conditions in the manufacturing equipment 410 when the product 460 is manufactured.
- Quality 444 is a measure of inspection of product 460 by inspection facility 420 .
- the inspection result 445 indicates pass/fail (OK/NG), which is the inspection result of the product.
- the learning data generation unit 111 acquires the manufacturing control data 440 and divides the products into groups of products that are continuously manufactured and have the same manufacturing conditions 443 . Next, the learning data generator 111 obtains the amount of change in the manufacturing conditions 443 from the immediately preceding group for each group, and uses this as the amount of manufacturing condition change. Subsequently, for each group, the learning data generation unit 111 obtains the difference between the quality 444 of each product in the group and the average value of the quality 444 in the immediately preceding group, and uses it as a quality change amount. The learning data generation unit 111 adds a record to the learning data 140 (see FIG. 4), and sets the manufacturing condition variation 141 and the quality variation 142 as the obtained manufacturing condition variation and quality variation.
- the regression model generation unit 112 generates a regression model 150 based on the learning data 140.
- the learning data 140 is data indicating the amount of change in product quality (quality change amount 142) when the manufacturing conditions are changed by the manufacturing condition change amount 141.
- the regression model 150 is a model referred to when estimating (calculating) the quality variation from the manufacturing condition variation, and is, for example, a linear regression model.
- the regression model 150 is calculated as ⁇ shown by the following equation (1).
- X is a matrix indicating manufacturing conditions referred to when generating the learning data 140 .
- ⁇ X is a matrix indicating the amount of change in manufacturing conditions included in the learning data 140 .
- ⁇ X T denotes the transposed matrix of ⁇ X.
- ⁇ y is a vector indicating one quality item (inspection measurement value) of the product included in the learning data 140 . There are ⁇ y as many as the number of quality items (N, which will be described later).
- ⁇ is the estimate of the partial regression coefficient, and there are as many quality items as there are.
- the optimization processing unit 113 solves the optimization problem that maximizes the yield score represented by the equation (2) described later, using the yield estimation function described later as the objective function, and determines the change in manufacturing conditions that maximizes the yield. Calculate quantity.
- Methods for solving optimization problems include optimization algorithms that do not use derivatives, local search methods, simulated annealing methods, tabu search methods, and genetic algorithms.
- N is the number of qualities (quality items). i is the subscript of the quality item, and there are the 1st quality item to the Nth quality item.
- M is the number of products whose yield is to be estimated, for example, the number of products manufactured during a predetermined period in the past. The optimization problem in this embodiment is to obtain the amount of change in manufacturing conditions that maximizes the number of products that pass the inspection among the M products to be manufactured.
- j is the index of the M products to be manufactured, from the 1st product to the Mth product.
- l L,i is the lower bound of the i-th quality item and the lower bound of the acceptance range of the i-th quality item.
- Equation (3) Equation (3)
- N( ⁇ , ⁇ 2 ) in Equation (3) represents a normal distribution with mean ⁇ and variance ⁇ 2 .
- ⁇ y i is an estimated value of the average amount of change related to the i-th quality item, and is calculated by Equation (4) described later.
- y 0,i,j is the measured value of the i-th quality item in the j-th product among the M products manufactured most recently.
- SE( ⁇ y i ) is the standard error of the estimated value of the amount of change related to the i-th quality item, and is calculated by Equation (5) below.
- s i is the standard error of regression for the i-th quality item, and is calculated by Equation (6) below.
- Equation (4) is a vector indicating the amount of change in manufacturing conditions.
- ⁇ is an estimated value of the partial regression coefficient of the learned linear model, and is calculated by Equation (1) described above.
- M 0 is the number of products when ⁇ was calculated.
- X is a matrix indicating manufacturing conditions referred to when generating the learning data 140 .
- X T denotes the transposed matrix of X.
- ⁇ x T denotes the transposed vector of ⁇ x.
- K is the number of manufacturing conditions.
- RSS is the residual sum of squares.
- the objective function processing unit 114 estimates (calculates) the yield from the production condition change amount as the yield estimation function, which is the objective function. Specifically, the objective function processing unit 114 obtains the probability distribution of the quality of the product 460 shown in Equation (3) from the amount of change in the quality of the product 460, and based on this probability distribution, the product 460 is inspected by the inspection equipment 420. Estimate the probability of passing The objective function processing unit 114 estimates the yield using a quality variation estimation unit 115, a quality estimation unit 116, and a yield estimation unit 117, which will be described later.
- the quality change amount estimator 115 calculates the quality change amount of the product 460 manufactured by the manufacturing equipment 410 from the manufacturing condition change amount of the manufacturing equipment 410 using the regression model 150 . Specifically, the quality change amount estimator 115 calculates the estimated value ⁇ yi of the quality change amount from the manufacturing condition change amount ⁇ x using Equation (4). Note that ⁇ in Equation (4) is ⁇ corresponding to the i-th quality item among N ⁇ s.
- the quality estimator 116 obtains the distribution of yi,j , which is the predicted value of the i-th quality item in the j-th product. Specifically, the sum of the estimated value ⁇ y i of the quality change amount and the measured value y 0,i,j of the i-th quality item in the j-th product among the M products manufactured most recently is taken as the distribution Calculated as the average value of The variance of the distribution is obtained from the standard error SE( ⁇ y i ) of the estimated value ⁇ y i of the quality variation (see equation (5)) and the standard error of regression s i (see equation (6)). Let the obtained mean value and the normal distribution of the variance be the distribution of yi,j (see equation (3)).
- the yield estimating unit 117 obtains the probability P (l L,i ⁇ y i,j ⁇ l U,i ) that y i,j satisfies the inspection criteria from the distribution of y i ,j , yield score ( (2)) is calculated.
- objective function processing section 114 estimates the yield using quality variation estimation section 115 , quality estimation section 116 , and yield estimation section 117 .
- FIG. 6 is a flowchart of regression model generation processing according to this embodiment.
- the regression model generation process is executed at a predetermined timing, such as after a predetermined number of products have been inspected after a predetermined period or manufacturing conditions have been changed.
- the learning data generator 111 acquires the manufacturing control data 440 (see FIG. 5) from the manufacturing control device 430 (see FIG. 1).
- the learning data generator 111 generates learning data 140 (see FIG. 4 ) from the acquired manufacturing control data 440 .
- the regression model generator 112 generates the regression model 150 (see formula (1)) from the learning data 140 .
- FIG. 7 is a flowchart of objective function processing executed by the objective function processing unit 114 according to this embodiment.
- the objective function processing is called and executed at necessary timing when the optimization processing unit 113 solves the optimization problem (see step S31 in FIG. 8, which will be described later).
- step S21 the objective function processing unit 114 starts a process of repeating steps S22 to S26 for each of M products whose yield is estimated.
- step S22 the objective function processing unit 114 starts a process of repeating steps S23 to S25 for each of N qualities (quality items).
- step S ⁇ b>23 the quality variation estimation unit 115 uses the regression model 150 to calculate the quality variation from the manufacturing condition variation. Specifically, the quality change amount estimator 115 calculates the estimated value ⁇ yi of the quality change amount from the manufacturing condition change amount ⁇ x using Equation (4). In step S24, the quality estimation unit 116 obtains the distribution of y i,j (see equation (3)), which is the predicted value of quality (quality item).
- step S25 the yield estimation unit 117 obtains the probability P (l L,i ⁇ y i,j ⁇ l U,i ) that y i, j satisfies the inspection criteria from the distribution of y i,j .
- ⁇ i 1, ..., N P (l L, i ⁇ y i, j ⁇ l U, i )
- ⁇ i 1, ..., N log (P ( l L,i ⁇ y i,j ⁇ l U,i )).
- step S27 the yield estimation unit 117 calculates yield score (see formula (2)).
- FIG. 8 is a flowchart of optimization processing according to this embodiment.
- the optimization process is executed at a predetermined timing, such as a predetermined cycle or after manufacturing a predetermined number of products.
- the optimization processing unit 113 solves the optimization problem that maximizes the yield score (see formula (2)) using the yield estimation function as the objective function.
- the optimization processing unit 113 transmits the manufacturing condition change amount, which is the optimum solution, to the manufacturing management device 430 .
- the manufacturing control device 430 that has received the optimum solution instructs the manufacturing equipment 410 to change the manufacturing conditions by the manufacturing condition change amount.
- the manufacturing condition optimizing apparatus 100 obtains the manufacturing condition change amount (manufacturing condition change amount) that maximizes the yield based on the measured value (y 0,i,j ) of the quality of the most recently manufactured product. Specifically, the production condition optimization device 100 obtains the amount of change in the production conditions that is the amount of change in quality that maximizes the yield. For quality caused by unmeasured or uncontrollable disturbances, current disturbances that are considered to be the same or little different from the most recent disturbances in order to change the manufacturing conditions based on the quality of the most recently manufactured product. Manufacturing conditions can be set to improve quality depending on the situation. Ultimately, the manufacturing conditions are adjusted according to the uncontrollable disturbance state, and it is possible to maximize the yield of the product and improve the manufacturing index.
- Regression model 150 in the above embodiment is a linear regression model, but may be another model.
- a Gaussian process regression model may be used instead of the (generalized) linear model.
- it may be a machine learning model such as a neural network model.
- a machine learning model for predicting the amount of change in quality from the amount of change in manufacturing conditions may be generated and used in the yield estimation function (objective function).
- the regression model 150 is a model for estimating the amount of quality change from the manufacturing conditions and the amount of change from the manufacturing conditions.
- the manufacturing condition optimization device 100 uses the product quality distribution (N( ⁇ y i + y 0 , i, j , SE( ⁇ y i ) 2 +s i 2 )) to estimate the yield.
- the manufacturing condition optimization device 100 acquires the quality measurement value as the inspection result (see step S11 in FIG. 6). Instead of the measured value of quality (quality item, inspection item), it is also possible to acquire the inspection result represented by two values of pass or fail such as an external inspection or the number of defects in one product.
- the regression model 150 is a logistic regression model in the case of inspection results represented by binary values, and a Poisson regression model in the case of the number of defects. Both logistic regression models and Poisson regression models are included in generalized linear models.
- the objective function processing unit 114, the quality variation estimation unit 115, the quality estimation unit 116, and the yield estimation unit 117 are divided for convenience of explanation.
- a yield estimator for example, the yield estimating unit uses a regression model to estimate the yield of the product 470 having a quality that passes the inspection by the inspection facility 420 when the manufacturing conditions of the product 460 by the manufacturing facility 410 are changed. good too.
- the distribution of the quality predictor y i,j (see equation (3)) is a normal distribution, but it may be a binomial distribution or a Poisson distribution.
Landscapes
- Engineering & Computer Science (AREA)
- Manufacturing & Machinery (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Quality & Reliability (AREA)
- General Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- General Health & Medical Sciences (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- Primary Health Care (AREA)
- General Business, Economics & Management (AREA)
- Marketing (AREA)
- Theoretical Computer Science (AREA)
- Human Resources & Organizations (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- General Factory Administration (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
Description
特許文献1に記載のプラントの運転条件最適化システムは、複数のセンサにより測定されたプラントの運転状態を示す運転状態データを取得する運転状態データ取得手段と、プラントに設けられたセンサにより測定された、または前記運転状態データ取得手段により取得された運転状態データに基づいて求められた、プラントの運転を評価する運転指標データを取得する運転指標データ取得手段と、前記運転状態データ取得手段により取得された運転状態データと前記運転指標データ取得手段により求められた運転指標データとを所定の項目に基づき関連させた一組の計測データとし、該計測データをデータ記録部に記録する計測データ記録手段と、前記データ記録部に記録された複数組の計測データに基づき、運転状態データ側を表す運転状態変数を説明変数とし、運転指標データ側を表す運転指標変数を目的変数として所定の多変量解析を行い回帰モデルを作成する回帰モデル作成手段であって、該回帰モデルは、説明変数を相互に無相関でかつ元の説明変数より少ない数の成分へ変換する成分変換手段と、該成分変換手段により変換された成分から目的変数を予測する予測手段と、該成分変換手段に対応した方法で成分から説明変数を推定する逆変換手段とを有するものであり、前記回帰モデル作成手段により作成された回帰モデルに基づき運転指標変数を最適化する運転状態変数を求める運転指標変数最適化手段であって、前記逆変換手段により推定された説明変数に関する制約条件を満たしつつ前記予測手段により予測された目的変数に関する評価関数を最適化する際の説明変数値を求め、該説明変数値を最適な運転条件とすることを特徴とする。
以下に、本発明を実施するための形態(実施形態)における製造条件最適化装置を説明する。
図1は、本実施形態に係る製造システム10の全体構成図である。製造システム10は、製造設備410、検査設備420、製造管理装置430、および製造条件最適化装置100を含んで構成される。製造設備410、検査設備420、製造管理装置430、および製造条件最適化装置100は、ネットワーク499を介して相互に通信可能である。
製品460は、検査設備420に搬送されて、検査を受ける。検査に合格した製品470は、良品として出荷される。検査で不合格だった製品480は、不良品であり、廃棄される。
製造管理装置430は、製品460の個々の製造条件と品質とを格納する製造管理データ440(後記する図5参照)を記憶する。
製造条件最適化装置100は、製造管理データ440に基づいて製品460の収率が最大となるような製造条件を求める。これにより、製造指標の向上が可能となる。
図2は、本実施形態に係る製造条件最適化装置100の処理概要を説明するための図である。回帰モデル生成処理では、学習データ140から回帰モデル150が生成される。学習データ140は、製造設備410(図1参照)における製造条件の変化量と、当該変化量の製造条件の変更を行ったときの検査設備420における製品460の品質の変化量とがセットになったデータである。回帰モデル150は、製造条件の変化量(製造条件変化量)から製品460の品質の変化量(品質変化量)を推定するモデルであって、例えば、線形回帰モデルである。
図3は、本実施形態に係る製造条件最適化装置100の機能ブロック図である。製造条件最適化装置100は、制御部110、記憶部130、通信部170、および入出力部180を含んで構成される。通信部170は、製造管理装置430を含む他の装置との通信データを送受信する。入出力部180には、ディスプレイやキーボード、マウスなどのユーザインタフェース機器が接続される。
図4は、本実施形態に係る学習データ140のデータ構成図である。学習データ140は、例えば表形式のデータであって、1つの行(レコード)は、製造条件変化量141、および品質変化量142の列(属性)を含む。
製造条件変化量141は、製造設備410における製造条件を変更した際の変更量を示す。製造条件は1つ以上の項目を含み、製造条件変化量141は、この1つ以上の項目の変化量を示す。
品質変化量142は、製造条件変化量141に示される製造条件の変更を行ったときの検査設備420における製品460の品質変化量を示す。品質変化量142は、1つ以上の製品460に係る1つ以上の品質(検査の測定値、品質値)の変化量を含む。
図3に戻って、制御部110はCPU(Central Processing Unit)を含んで構成され、学習データ生成部111、回帰モデル生成部112、最適化処理部113、目的関数処理部114、品質変化量推定部115、品質推定部116、および収率推定部117を含む。
学習データ生成部111は、製造管理装置430が記憶する製造管理データ440(後記する図5参照)を取得して、学習データ140(図4参照)を生成する。
製品識別情報441は、個々の製品460を識別するための情報であって、例えばシリアル番号である。製造日時442は、製品460が製造された日時である。製造条件443は、製品460が製造されたときの製造設備410における製造条件である。品質444は、検査設備420による製品460の検査の測定値である。検査結果445は、製品の検査結果である合否(OK/NG)を示す。
ΔXは、学習データ140に含まれる製造条件変化量を示す行列である。
ΔXTは、ΔXの転置行列を示す。
Δyは、学習データ140に含まれる製品の1つの品質項目(検査の測定値)を示すベクトルである。Δyは、品質項目の数(後記するN)だけある。
βは、偏回帰係数の推定値であって、品質項目の数だけある。
iは、品質項目の添え字であって、1番目の品質項目からN番目の品質項目がある。
Mは、収率を推定する対象の製品の数であり、例えば過去の所定期間に製造した製品の数である。本実施形態における最適化問題とは、製造されるM個の製品のなかで検査に合格する製品の数を最大化する製造条件変化量を求めることである。
jは、製造されるM個の製品の添え字であって、1番目の製品からM番目の製品がある。
lL,iは、i番目の品質項目の下限値であり、i番目の品質項目の合格範囲における下限値である。
lU,iは、i番目の品質項目の上限値であり、i番目の品質項目の合格範囲における上限値である。
yi,jは、j番目の製品におけるi番目の品質項目の予測値である。yi,jは、以下の式(3)に示される確率分布に従う。
Δyiは、i番目の品質項目に係る変化量の平均の推定値であり、後記する式(4)で算出される。
y0,i,jは、直近に製造されたM個の製品のなかでj番目の製品におけるi番目の品質項目の実測値である。
SE(Δyi)は、i番目の品質項目に係る変化量の推定値の標準誤差であり、後記する式(5)で算出される。
siは、i番目の品質項目に係る回帰の標準誤差であり、後記する式(6)で算出される。
βは、学習された線形モデルの偏回帰係数の推定値であり、前述した式(1)で算出される。
Xは、学習データ140を生成する際に参照した製造条件を示す行列である。
XTは、Xの転置行列を示す。
ΔxTは、Δxの転置ベクトルを示す。
RSSは、残差平方和である。
上記に説明したように、目的関数処理部114は、品質変化量推定部115、品質推定部116、および収率推定部117を用いて収率を推定する。
図6は、本実施形態に係る回帰モデル生成処理のフローチャートである。回帰モデル生成処理は、例えば所定の周期や製造条件を変更した後に所定数の製品の検査を終えた後など、所定のタイミングで実行される。
ステップS11において学習データ生成部111は、製造管理装置430(図1参照)から製造管理データ440(図5参照)を取得する。
ステップS12において学習データ生成部111は、取得した製造管理データ440から学習データ140(図4参照)を生成する。
ステップS13において回帰モデル生成部112は、学習データ140から回帰モデル150(式(1)参照)を生成する。
図7は、本実施形態に係る目的関数処理部114が実行する目的関数処理のフローチャートである。目的関数処理は、最適化処理部113が最適化問題を解く(後記する図8のステップS31参照)際に、必要なタイミングで呼び出されて実行される。
ステップS21において目的関数処理部114は、収率を推定するM個の製品ごとにステップS22~S26を繰り返す処理を開始する。
ステップS22において目的関数処理部114は、N個ある品質(品質項目)ごとにステップS23~S25を繰り返す処理を開始する。
ステップS24において品質推定部116は、品質(品質項目)の予測値であるyi,jの分布(式(3)参照)を求める。
ステップS26において収率推定部117は、製品が合格する確率Πi=1,・・・,NP(lL,i≦yi,j≦lU,i)を求める。なお、本実施形態では、Πi=1,・・・,NP(lL,i≦yi,j≦lU,i)に替わりΣi=1,・・・,Nlog(P(lL,i≦yi,j≦lU,i))を求める。
ステップS27において収率推定部117は、収率score(式(2)参照)を算出する。
図8は、本実施形態に係る最適化処理のフローチャートである。最適化処理は、例えば所定周期や所定数の製品の製造後など、所定のタイミングで実行される。
ステップS31において最適化処理部113は、収率推定関数を目的関数として、収率score(式(2)参照)が最大となる最適化問題を解く。
ステップS32において最適化処理部113は、最適解である製造条件変化量を製造管理装置430に送信する。最適解を受信した製造管理装置430は、製造条件を製造条件変化量分だけ変更するように製造設備410に指示する。
製造条件最適化装置100は、直近に製造した製品の品質の実測値(y0,i,j)に基づいて、収率が最大となる製造条件の変更量(製造条件変化量)を求める。詳しくは、製造条件最適化装置100は収率が最大となるような品質の変化量となる製造条件の変更量を求める。測定していない、ないしは制御不能な外乱に起因する品質について、直近に製造した製品の品質に基づいて製造条件を変更するため、直近の外乱と同じ、ないしはほとんど変わらないと考えられる現在の外乱の状態に応じて品質が向上するように製造条件を設定できるようになる。延いては、制御不能な外乱の状態に応じて製造条件を調整することになり、製品の収率を最大化して製造指標を向上させることが可能となる。
上記した実施形態における回帰モデル150は線形回帰モデルであるが、別のモデルであってもよい。例えば、(一般化)線形モデルに替えてガウス過程回帰モデルであってもよい。または、ニューラルネットワークモデルなどの機械学習モデルであってもよい。詳しくは、学習データ140(図4参照)を教師データとして、製造条件変化量から品質変化量を予測する機械学習モデルを生成して、収率推定関数(目的関数)で用いてもよい。ガウス過程回帰モデルやニューラルネットワークモデルを用いることで、説明変数(図4記載の製造条件変化量141参照)間に高次の交互作用がある場合に予測精度を高めることができる。
また、回帰モデル150の生成に用いる学習データ140における説明変数として、製造条件変化量141に加えて変化量に換算する前の製造条件(図5記載の製造条件443参照)を追加してもよい。この場合、回帰モデル150は、製造条件と当該製造条件からの変化量とから品質変化量を推定するモデルとなる。
上記した実施形態では、製造条件最適化装置100は、直近に製造した所定数M個の製品の品質(y0,i,j)に基づいて、製品の品質の分布(N(Δyi+y0,i,j,SE(Δyi)2+si 2))を求めて、収率を推定している。これに限られず、収率推定部117は、製品の品質の平均値(y0,i)を求め、製品によらない(1つの製品の)品質の分布(N(Δyi+y0,i,SE(Δyi)2+si 2))を求めて、収率を推定するようにしてもよい。製品ごとの繰り返し処理(j=1,・・・,M)がなくなる(M=1とみなす)ので、製造条件最適化装置100は、高速に最適化処理ができるようになり、より頻繁に最適化処理が可能になる。
上記した実施形態では製造条件最適化装置100は、検査結果として品質の測定値を取得している(図6のステップS11参照)。品質(品質項目、検査項目)の測定値の替わりに、外界検査などの合格または不合格の2値で表される検査結果または1つの製品にある欠点個数を取得するようにしてもよい。2値で表される検査結果の場合は、回帰モデル150はロジスティック回帰モデルとなり、欠点個数の場合には、回帰モデル150はポアソン回帰モデルとなる。ロジスティック回帰モデルもポアソン回帰モデルも一般化線形モデルに含まれる。
以上、本発明のいくつかの実施形態や変形例について説明したが、これらの実施形態は、例示に過ぎず、本発明の技術的範囲を限定するものではない。例えば、上記実施形態では、説明の都合により目的関数処理部114、品質変化量推定部115、品質推定部116、および収率推定部117に分けているが、収率推定関数として1つの機能部、例えば収率推定部、としてもよい。つまりは、収率推定部が、回帰モデルを用いて、製造設備410による製品460の製造条件を変化させたときに検査設備420による検査に合格する品質を有する製品470の収率を推定するとしてもよい。
上記した実施形態では、品質の予測値であるyi,jの分布(式(3)参照)は、正規分布であるが、二項分布またはポアソン分布であってもよい。
110 制御部
111 学習データ生成部
112 回帰モデル生成部
113 最適化処理部
114 目的関数処理部(収率推定部)
115 品質変化量推定部(収率推定部)
116 品質推定部(収率推定部)
117 収率推定部
140 学習データ
150 回帰モデル。
410 製造設備
420 検査設備
430 製造管理装置
440 製造管理データ
460 製品
Claims (10)
- 製造設備による製品の製造条件を変化させたときに検査設備による検査に合格する品質を有する前記製品の収率を推定する収率推定部と、
前記収率が最大となる前記製造条件の変化量を算出する最適化処理部とを備える
製造条件最適化装置。 - 前記製造条件の変化量から前記製造設備が製造した前記製品の品質の変化量を推定する品質変化量推定部を、さらに備え、
前記収率推定部は、
前記品質変化量推定部が推定した前記製品の品質の変化量から前記製品が前記検査設備による検査に合格する確率を推定し、
前記検査に合格する確率から、前記検査設備による検査に合格する品質を有する前記製品の収率を推定する
請求項1に記載の製造条件最適化装置。 - 回帰モデル生成部を、さらに備え、
前記回帰モデル生成部は、
過去の前記製造条件の変化量と当該製造条件の変化量における前記製品の品質の変化量との対応データに基づき、前記製造条件の変化量から前記製品の品質の変化量を推定する回帰モデルを生成し、
前記品質変化量推定部は、
前記回帰モデルを用いて前記製造条件の変化量から前記製品の品質の変化量を推定する
請求項2に記載の製造条件最適化装置。 - 前記回帰モデルは、
一般化線形モデル、ガウス過程回帰モデルまたはニューラルネットワークモデルである
請求項3に記載の製造条件最適化装置。 - 前記最適化処理部は、
導関数を用いない最適化アルゴリズム、局所探索法、焼きなまし法、タブー探索法、および遺伝的アルゴリズムのなかの少なくとも1つを用いて前記収率が最大になる前記製造条件の変化量を算出する
請求項1に記載の製造条件最適化装置。 - 前記収率推定部は、
前記品質の変化量から前記製品が前記検査設備による検査に合格する確率を推定する際に、直近に製造された所定数の製品の品質に基づいて、前記製品が前記検査設備による検査に合格する確率を推定する
請求項2に記載の製造条件最適化装置。 - 前記収率推定部は、
前記製品の品質の変化量から当該製品の品質の確率分布を求め、
当該確率分布を基づいて当該製品が前記検査設備による検査に合格する確率を推定する
請求項2に記載の製造条件最適化装置。 - 前記確率分布は、正規分布、二項分布またはポアソン分布である
請求項7に記載の製造条件最適化装置。 - コンピュータを請求項1~8の何れか1項に記載の製造条件最適化装置として機能させるためのプログラム。
- 製造条件最適化装置の製造条件最適化方法であって、
製造設備による製品の製造条件を変化させたときに検査設備による検査に合格する品質を有する前記製品の収率を推定するステップと、
前記収率が最大となる前記製造条件の変化量を算出するステップと、を実行する
製造条件最適化方法。
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202180090647.6A CN116806329A (zh) | 2021-01-19 | 2021-10-25 | 制造条件最优化装置、程序以及制造条件最优化方法 |
JP2022576979A JPWO2022158066A1 (ja) | 2021-01-19 | 2021-10-25 | |
US18/261,314 US20240077854A1 (en) | 2021-01-19 | 2021-10-25 | Manufacturing condition optimization apparatus, computer program product, and manufacturing condition optimization method |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2021-006500 | 2021-01-19 | ||
JP2021006500 | 2021-01-19 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2022158066A1 true WO2022158066A1 (ja) | 2022-07-28 |
Family
ID=82548707
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2021/039329 WO2022158066A1 (ja) | 2021-01-19 | 2021-10-25 | 製造条件最適化装置、プログラムおよび製造条件最適化方法 |
Country Status (4)
Country | Link |
---|---|
US (1) | US20240077854A1 (ja) |
JP (1) | JPWO2022158066A1 (ja) |
CN (1) | CN116806329A (ja) |
WO (1) | WO2022158066A1 (ja) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH08305763A (ja) * | 1995-04-28 | 1996-11-22 | Nippon Steel Corp | 生産計画作成方法 |
JP2012226511A (ja) * | 2011-04-19 | 2012-11-15 | Hitachi Ltd | 歩留まり予測システムおよび歩留まり予測プログラム |
JP2020086784A (ja) * | 2018-11-21 | 2020-06-04 | 株式会社日立製作所 | 製造条件特定システムおよび方法 |
JP2020166749A (ja) * | 2019-03-29 | 2020-10-08 | 株式会社カネカ | 製造システム、情報処理方法、および製造方法 |
-
2021
- 2021-10-25 US US18/261,314 patent/US20240077854A1/en active Pending
- 2021-10-25 CN CN202180090647.6A patent/CN116806329A/zh active Pending
- 2021-10-25 JP JP2022576979A patent/JPWO2022158066A1/ja active Pending
- 2021-10-25 WO PCT/JP2021/039329 patent/WO2022158066A1/ja active Application Filing
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH08305763A (ja) * | 1995-04-28 | 1996-11-22 | Nippon Steel Corp | 生産計画作成方法 |
JP2012226511A (ja) * | 2011-04-19 | 2012-11-15 | Hitachi Ltd | 歩留まり予測システムおよび歩留まり予測プログラム |
JP2020086784A (ja) * | 2018-11-21 | 2020-06-04 | 株式会社日立製作所 | 製造条件特定システムおよび方法 |
JP2020166749A (ja) * | 2019-03-29 | 2020-10-08 | 株式会社カネカ | 製造システム、情報処理方法、および製造方法 |
Also Published As
Publication number | Publication date |
---|---|
US20240077854A1 (en) | 2024-03-07 |
CN116806329A (zh) | 2023-09-26 |
JPWO2022158066A1 (ja) | 2022-07-28 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
KR101917006B1 (ko) | 머신 러닝 기반 반도체 제조 수율 예측 시스템 및 방법 | |
US11599803B2 (en) | Soldering process parameter suggestion method and system thereof | |
Xu et al. | PHM-oriented integrated fusion prognostics for aircraft engines based on sensor data | |
Liu et al. | Unevenly sampled dynamic data modeling and monitoring with an industrial application | |
CN111191726B (zh) | 一种基于弱监督学习多层感知器的故障分类方法 | |
Wang et al. | Transient analysis and real-time control of geometric serial lines with residence time constraints | |
JP2019016039A (ja) | プロセスの異常状態診断方法および異常状態診断装置 | |
JP2022132895A (ja) | 合金材料の特性を予測する製造支援システム、予測モデルを生成する方法およびコンピュータプログラム | |
Losi et al. | Gas turbine health state prognostics by means of Bayesian hierarchical models | |
Lee et al. | In-line predictive monitoring framework | |
Li et al. | A wiener-based remaining useful life prediction method with multiple degradation patterns | |
CN109213057A (zh) | 智能诊断装置和方法 | |
WO2022158066A1 (ja) | 製造条件最適化装置、プログラムおよび製造条件最適化方法 | |
Garcia et al. | Automatic generation of digital-twins in advanced manufacturing: a feasibility study | |
WO2021157667A1 (ja) | 予測装置、予測方法及びプログラム | |
CN101118423A (zh) | 虚拟测量预估模型的适用性选择方法与系统 | |
JP2011145905A (ja) | 予測関数生成装置、方法、及び、プログラム | |
Polyiam et al. | A hybrid forecasting model of cassava price based on artificial neural network with support vector machine technique | |
CN113807606B (zh) | 可解释集成学习的间歇过程质量在线预测方法 | |
WO2021157670A1 (ja) | 予測装置、予測方法及びプログラム | |
EP4102319A1 (en) | Control device, control method, and program | |
TWI830193B (zh) | 預測系統、資訊處理裝置以及資訊處理程式 | |
Okoro et al. | An approach to reliability analysis of aircraft systems for a small dataset | |
WO2023085195A1 (ja) | モデル生成装置、モデル生成方法及びデータ推定装置 | |
Miguelez et al. | An Adaptive Learning Approach to Multivariate Time Forecasting in Industrial Processes |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 21921169 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2022576979 Country of ref document: JP Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 18261314 Country of ref document: US |
|
WWE | Wipo information: entry into national phase |
Ref document number: 202180090647.6 Country of ref document: CN |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 21921169 Country of ref document: EP Kind code of ref document: A1 |