Model based online monitoring and control of batch processes in disturbance situations

We present an approach of arc-timed Petri nets to online monitoring and predictive simulation for resource allocation in disturbance situations in recipe-driven batch processes. Our approach assumes underlying sequential controllers which control the process and provide signals indicating the state transitions of the real process. We monitor the sequence of process steps and their duration. Predictive simulation is performed to help the operator to find a reasonable resource allocation strategy in case of a deviation. The system has been implemented within a process control system and will be tested on a laboratory batch process