The continuous development of steel products generates new challenges for the maintenance of manufacturing machines in steel mills. Substantial mechanical stress is inflicted on the machines during the processing of modern high-strength steels. This increases the risks of damage and flaws in the processed material may appear if the capability of a machine is exceeded. Therefore, new approaches are needed to prevent the machine condition from deteriorating. This study introduces an approach to the prediction of mechanical stress inflicted on a roller leveler during the processing of cold steel strips. The relative stress level is indicated by features extracted from an acceleration signal. These features are based on the calculation of generalized norms. Steel strip properties are used as explanatory variables in regression models to predict values for the extracted vibration features. The models tested in this study include multiple linear regression, partial least squares regression and generalized regression neural network. The models were tested using an extensive data set from a roller leveler that is in continuous operation in a steel mill. The prediction accuracy of the best models identified indicates that the relative stress level inflicted by each steel strip could be predicted based on its properties.
Digital twins have gained a lot of attention in modern day industry, but practical challenges arise from the requirement of continuous and real-time data integration. The actual physical systems are also exposed to disturbances unknown to the real-time simulation. Therefore, adaptation is required to ensure reliable performance and to improve the usability of digital twins in monitoring and diagnostics. This study proposes a general approach to the real-time adaptation of digital twins based on a mechanism guided by evolutionary optimization. The mechanism evaluates the deviation between the measured state of the real system and the estimated state provided by the model under adaptation. The deviation is minimized by adapting the model input based on the differential evolution algorithm. To test the mechanism, the measured data were generated via simulations based on a physical model of the real system. The estimated data were generated by a surrogate model, namely a simplified version of the physical model. A case study is presented where the adaptation mechanism is applied on the digital twin of a marine thruster. Satisfactory accuracy was achieved in the optimization during continuous adaptation. However, further research is required on the algorithms and hardware to reach the real-time computation requirement.
The development of steel products with various characteristics increases the need for timely and preventive maintenance and condition monitoring of the production machinery. For instance, the roller levelers at modern steel factories are exposed to a high variation of forces due to the large range of steels leveled. In this study, the vibration measured from a steel leveler used for cold steel strips is analyzed with the goal to identify the effects different operational conditions have. Features such as generalized norms, generalized norm sums and the crest factor are computed from the vibration signals. The effects of the steel strip properties and the operational parameters of the machine on these features are then analyzed. The obtained information can be utilized in the models that are used as planning tools for the preventive maintenance of the steel leveler.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.