Condition monitoring based on up-to-date process thermal models

• Applicant: Professor Dr.-Ing. Steffen Ihlenfeldt
• Subject Area: Production Automation and Assembly Technology
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 563565782

The goal is to detect the need for maintenance of machine components based on a change in the thermal behavior of the components. The indicator for the need of maintenance should be an abnormal thermal behavior of the machine during regular operation. The anomalies should be detected using temperature sensors and physically based thermal models that can be computed in parallel with the process. Data that can be recorded in the control system will be used as model input variables. For the first time, anomaly detection methods will be coupled with thermal concentrated node models and MOR-FE models. This will provide new insights for hybrid, model-based anomaly detection methods. Furthermore, a concept for considering production-typical labeling situations (supervised, unsupervised, semi-supervised) will be developed and exemplarily implemented.

• DFG Programme: Research Grants