Abstract: A malfunction detection device provided with an acquisition unit, a cutout unit, a feature extraction unit, a learning unit, and a deviation degree computation unit. The acquisition unit is configured to acquire device data. The cutout unit is configured to cut out data from the acquired device data based on predetermined conditions. The feature extraction unit is configured to extract a feature vector from the cut-out device data. The learning unit is configured to analyze the extracted feature vector. The learning unit is configured to generate a model of the feature vector based on analysis results. The deviation degree computation unit is configured to compute a degree of deviation between a feature vector of newly acquired device data and the model of the feature vector.

Abstract: A system includes: a monitoring data obtainer configured to obtain real-time monitoring data regarding equipment; a deterioration index calculator configured to calculate an estimated deterioration index of the equipment from an equipment model and the obtained monitoring data, the equipment model being created by modeling a relationship between the real-time monitoring data of the equipment in a normal state; a time change predictor configured to predict a time change of the estimated deterioration index; and a scheduler configured to estimate a cost in a future caused by the abnormal event, the cost being estimated based on the predicted time change of the estimated deterioration index and the relationship between the estimated deterioration index and the probability of the abnormal event.

Abstract: Provided is a glass film ribbon manufacturing device (1), including: a transverse conveyance unit (4), which is configured to convey a glass film ribbon (G) in a transverse direction; and a cleaving unit (5), which is arranged on a conveyance path of the transverse conveyance unit (4), and is configured to cleave the glass film ribbon (G) along a preset cleaving line extending in a longitudinal direction, the transverse conveyance unit (4) including a wrinkle-smoothing unit (14) configured to smooth a wrinkle generated in the glass film ribbon (G) before the glass film ribbon (G) is cleaved by the cleaving unit (5).

Abstract: Provided is a glass film ribbon manufacturing device (1), including: a transverse conveyance unit (4), which is configured to convey a glass film ribbon (G) in a transverse direction; and a cleaving unit (5), which is arranged on a conveyance path of the transverse conveyance unit (4), and is configured to cleave the glass film ribbon (G) along a preset cleaving line extending in a longitudinal direction, the transverse conveyance unit (4) including a wrinkle-smoothing unit (14) configured to smooth a wrinkle generated in the glass film ribbon (G) before the glass film ribbon (G) is cleaved by the cleaving unit (5).

Abstract: Provided is a glass film ribbon manufacturing device (1), including: a transverse conveyance unit (4), which is configured to convey a glass film ribbon (G) in a transverse direction; and a cleaving unit (5), which is arranged on a conveyance path of the transverse conveyance unit (4), and is configured to cleave the glass film ribbon (G) along a preset cleaving line extending in a longitudinal direction, the transverse conveyance unit (4) including a wrinkle-smoothing unit (14) configured to smooth a wrinkle generated in the glass film ribbon (G) before the glass film ribbon (G) is cleaved by the cleaving unit (5).

Abstract: A plant model management device according to one embodiment includes a generation unit and a combining unit. The generation unit receives a first set of prior probability distributions of a first set of automatic operation parameters. The first set of automatic operation parameters define a condition required for modeling a plant model that is a model based on data concerning a plant. The generation unit generates a first set of posterior probability distributions of the first set of automatic operation parameters and a first set of probabilistic plant models based on the first set of prior probability distributions. The first set of posterior probability distributions reflect compatibility with learning data that is past time-series data concerning a first explanatory variable and a first objective variable. The combining unit combines the first set of probabilistic plant models using the first set of posterior probability distributions to generate a first plant model.

Abstract: A manufacturing method for a glass roll includes drawing a glass film downward vertically from a forming device, converting a delivery direction of the glass film from a vertical direction to a horizontal direction, continuously delivering the glass film to a downstream side by a horizontal conveyance section, and rolling the glass film into a roll shape on a downstream side of the horizontal conveyance section. While the glass film is supported in surface contact by an endless belt serving as a support section, which is provided in the horizontal conveyance section, a rolling force equal to or less than a static friction force, which is generated by a self-weight of the glass film between the glass film and the endless belt, is imparted to the glass film.

Abstract: A system includes: a monitoring data obtainer configured to obtain real-time monitoring data regarding equipment; a deterioration index calculator configured to calculate an estimated deterioration index of the equipment from an equipment model and the obtained monitoring data, the equipment model being created by modeling a relationship between the real-time monitoring data of the equipment in a normal state; a time change predictor configured to predict a time change of the estimated deterioration index; and a scheduler configured to estimate a cost in a future caused by the abnormal event, the cost being estimated based on the predicted time change of the estimated deterioration index and the relationship between the estimated deterioration index and the probability of the abnormal event.

Abstract: A manufacturing method for a glass film includes a cleaving step of dividing a glass film having a band shape into a product glass portion and an unnecessary glass portion, and a collecting step of sequentially shattering the unnecessary glass portion, and collecting the shattered unnecessary glass portion. The cleaving step is followed by vibration regulating processing of regulating vibrations of the unnecessary glass portion in a thickness direction thereof, the vibration regulating processing being executed in at least one location in the longitudinal direction of the unnecessary glass portion, and the collecting step being executed on a downstream side of the at least one location of the vibration regulating processing.

Abstract: Provided is a glass film ribbon manufacturing device (1), including: a transverse conveyance unit (4), which is configured to convey a glass film ribbon (G) in a transverse direction; and a cleaving unit (5), which is arranged on a conveyance path of the transverse conveyance unit (4), and is configured to cleave the glass film ribbon (G) along a preset cleaving line extending in a longitudinal direction, the transverse conveyance unit (4) including a wrinkle-smoothing unit (14) configured to smooth a wrinkle generated in the glass film ribbon (G) before the glass film ribbon (G) is cleaved by the cleaving unit (5).

Abstract: A plant model management device according to one embodiment includes a generation unit and a combining unit. The generation unit receives a first set of prior probability distributions of a first set of automatic operation parameters. The first set of automatic operation parameters define a condition required for modeling a plant model that is a model based on data concerning a plant. The generation unit generates a first set of posterior probability distributions of the first set of automatic operation parameters and a first set of probabilistic plant models based on the first set of prior probability distributions. The first set of posterior probability distributions reflect compatibility with learning data that is past time-series data concerning a first explanatory variable and a first objective variable. The combining unit combines the first set of probabilistic plant models using the first set of posterior probability distributions to generate a first plant model.

Abstract: A manufacturing method for a glass roll includes drawing a glass film downward vertically from a forming device, converting a delivery direction of the glass film from a vertical direction to a horizontal direction, continuously delivering the glass film to a downstream side by a horizontal conveyance section, and rolling the glass film into a roll shape on a downstream side of the horizontal conveyance section. While the glass film is supported in surface contact by an endless belt serving as a support section, which is provided in the horizontal conveyance section, a rolling force equal to or less than a static friction force, which is generated by a self-weight of the glass film between the glass film and the endless belt, is imparted to the glass film.

Abstract: The manufacturing method for a glass film includes, in this order: a cleaving step of dividing a glass film (G) having a band shape into a product glass portion (Ga) and an unnecessary glass portion (Gb); and a collecting step of sequentially shattering the unnecessary glass portion (Gb), and collecting the shattered unnecessary glass portion. The cleaving step is followed by vibration regulating processing of regulating vibrations of the unnecessary glass portion (Gb) in a thickness direction thereof, the vibration regulating processing being executed in at least one location in the longitudinal direction of the unnecessary glass portion (Gb) together with the collecting step executed on a downstream side of the at least one location of the vibration regulating processing.