A system and method for labelling normal and abnormal regions in data related to a paper machine for web break prediction and labelling individual parameters for root cause analysis, using machine learning models, includes using the machine learning models in real-time to predict breaks in the paper web, analyzing root cause for the breaks in the paper web, and estimating a time to break. An auto-data-labeling framework helps in adaptive learning for autonomous model improvement of the deployed model, transfer learning, shortlisting parameters and automating feasibility study.

Aspects of the present invention provide methods, systems, and/or the like for: (1) receiving current thermal imaging data for a portion of a paper web in the paper manufacturing process; (2) determining force data for the portion of the paper web; (3) processing, to produce a first data analysis result, current thermal imaging data, the force data, and paper profile data using a machine-learning model trained with respective historical thermal imaging data, respective historical force data, and respective paper profile data for each respective prior paper breakage event in a set of prior paper breakage events; (4) generating, based on the first data analysis result, a prediction as to an occurrence of the paper break on the portion of the paper web; (5) identifying a preventative action based on the prediction; and (6) facilitating performance of the preventative action.

An assembly includes: a seal strip with an upper wear surface and an opposed lower surface, the seal strip configured to provide a seal for a suction roll, the seal strip comprising a first material having a first thermal conductivity; a seal strip holder, the seal strip residing in the seal strip holder and movable relative thereto; and a temperature monitoring system. The temperature monitoring system comprises: a heat-conducting rod at least partially embedded in the seal strip, the heat-conducting rod comprising a second material, wherein the second material has a second thermal conductivity that is higher than the first thermal conductivity; a temperature sensor connected with the heat-conducting rod for sensing temperature in the heat-conducting rod, and a controller operatively connected with the temperature sensor, the controller configured to receive signals from the temperature sensor and process the signals to indicate a temperature of the upper surface of the seal strip.

Aspects of the present invention provide methods, systems, and/or the like for: (1) receiving current thermal imaging data for a portion of a paper web in the paper manufacturing process; (2) determining force data for the portion of the paper web; (3) processing, to produce a first data analysis result, current thermal imaging data, the force data, and paper profile data using a machine-learning model trained with respective historical thermal imaging data, respective historical force data, and respective paper profile data for each respective prior paper breakage event in a set of prior paper breakage events; (4) generating, based on the first data analysis result, a prediction as to an occurrence of the paper break on the portion of the paper web; (5) identifying a preventative action based on the prediction; and (6) facilitating performance of the preventative action.

Automated parameter tuning techniques for cross-directional model predictive control for paper-making under user-specified parametric uncertainties to reduce variability of the actuator and measurement profiles in the spatial domain is proposed. Decoupling properties of the spatial and temporal frequency components permit separate controller design and parameter tuning. CD-MPC design that explicitly accounts for parametric model uncertainty while finding MPC cost function weighing matrices that prevent actuator picketing and guarantee robust stability of the spatial CD profile. Picketing refers to periodic variation patterns in the actuator array. The inventive technique includes: (i) determining the worst case cutoff frequency of all process models, given parametric uncertainty, (ii) designing a weighing matrix to penalize high frequency actuator variability based on the process model and worst case cutoff frequency, and (iii) finding a multiplier for the spatial frequency weighted actuator variability term in the MPC cost function that assures robust spatial stability.

An object is to provide a monitoring system capable of preventing the occurrence of a defect in paper by monitoring the device. The present invention is directed to a monitoring system A having a papermaking machine 1 for manufacturing paper X, an applying device 2 for applying a chemical solution to a site of the papermaking machine directly or indirectly in contact with the paper while the papermaking machine 1 is operated, a control panel 3 for setting an application condition of the applying device 2, a monitoring camera 4 for monitoring a monitoring target site, and a control device 5 connected to the monitoring camera 4 via a network. The monitoring target site is the site of the papermaking machine 1 directly or indirectly in contact with the paper X and/or the applying device 2, and the control device 5 has a computing unit S1 which converts a state of the monitoring target site into a numerical form by using video taken by the monitoring camera 4, a display unit S2 which displays detection data obtained by conversion into the numerical form at the computing unit S1, and a storage unit S3 which stores the detection data.

A raw material supply device includes a plurality of supply portions for supplying the raw material, a detection portion for detecting a state of the supply portion, and a control portion for controlling the supply portion, in which when the plurality of supply portions are in a normal state, the control portion switches the supply portions at a predetermined timing, and causes each of the supply portions to supply the raw material.

A system, method, and/or apparatus is provided for establishing targets or setpoints of speed parameters for the operation of a papermaking machine that is based on the square point of the paper sheet being produced. The square point of the paper sheet is determined in response to one or more speed parameters and associated one or more fiber distribution parameters for the paper sheet that are measured a plurality of times during the papermaking process and analyzed to determine the square point based on the measurements.

Automated parameter tuning techniques for cross-directional model predictive control for paper-making under user-specified parametric uncertainties to reduce variability of the actuator and measurement profiles in the spatial domain is proposed. Decoupling properties of the spatial and temporal frequency components permit separate controller design and parameter tuning. CD-MPC design that explicitly accounts for parametric model uncertainty while finding MPC cost function weighing matrices that prevent actuator picketing and guarantee robust stability of the spatial CD profile. Picketing refers to periodic variation patterns in the actuator array. The inventive technique includes: (i) determining the worst case cutoff frequency of all process models, given parametric uncertainty, (ii) designing a weighing matrix to penalize high frequency actuator variability based on the process model and worst case cutoff frequency, and (iii) finding a multiplier for the spatial frequency weighted actuator variability term in the MPC cost function that assures robust spatial stability.

Estimating or predicting runnability or end product quality risk level for a pulp or papermaking process is disclosed. The method includes measuring hydrophobicity values of samples originating from a same aqueous process stream. A hydrophobicity measurement signal is produced of measured hydrophobicity values as a function of time. A risk level is calculated for the process. At least one mathematical index is calculated based on the hydrophobicity measurement signal, and optionally based on the amount of particles in the sample, other property of the aqueous stream and/or production data. The mathematical index and optionally the amount of the particles, other property, and/or production data is used as a risk indicator input in the calculation. Based on the risk level calculated for the pulp or papermaking process, the runnability and/or end product quality risk level for the pulp or papermaking process is indicated.