Inline Simulation and Knowledge Management in Film Extrusion and Coating Processes, 2005 European PLACE Conference

In this paper, innovative approaches for improved information management to support operators in extrusion processes are discussed. This will be realised through inline simulation systems, knowledge identification and usage based on data mining technologies. To support operators steering and controlling extrusion processes, context related information is generated by simulation calculations and action hints to solve problems are given based on knowledge retrieval from process manipulation history. These methods have to be integrated seamlessly into highly heterogeneous IT environments consisting of PLC automation systems, databases, ERP systems, autonomous quality inspection systems and more. Thus, a well structured architecture of integrators, wrappers, bridges and controllers has to be developed and realised as an integration framework to automated information flow between the systems components.

Inline simulation can be used to analyse parts and components of the plant which cannot be controlled by sensor data (’soft sensor’), to predict and analyse process states by extrapolation of process parameters and to pre-estimate process modification effects by simulation calculations. Additionally to these simulation based information generation, the context based definition of situations and their online identification during the production process makes previous process manipulation, actions and interactions and their effects available to the operators. It helps to assess a process state based on similar situations from the past and to find the best modification actions to solve the problems.