Model based monitoring and diagnostic procedure for supporting condition based maintenance strategies

In heavy industries where the production processes are integrated, high demands are placed on the system reliability and availability. In case of system breakdown, high downtime costs occur due to production loss. Suitable maintenance strategies are required to increase the system availability and to minimize downtime costs. In order to have effective maintenance strategies, information about the system component status i.e. when the system component will fail and why it has failed is required. Dedicated sensors are used to continuously monitor the system condition but rarely convey the information about the component status. In order to bridge this deficit, the capability of the online simulation together with the condition monitoring system is investigated in this research work. As a system example, slab sizing press drive used in steel mills is chosen. The electromechanical model of the drive with electrical and mechanical components is implemented on a real time system and is fed with the measured inputs. With these real time inputs, the stress acting on the system components is calculated. Appropriate algorithms are used for the estimation of remaining service life time of the components using the information obtained from the model. Further, this information can be directly sent to a maintenance planning system, which is at the dispose of a maintenance engineer. In this manner, maintenance measures can be planned earlier and unscheduled downtime due to the component breakdown can be reduced.