A Modeling and Optimization Tool for the Expandable Polystyrene Batch Process

Expandable Polystyrene (EPS), a material of the insulation productions and packages, is commonly produced in a batch process. The control of the batch process is based on predefined process recipes and process parameters such as mixing properties. The EPS production has to be able to satisfy the aims and quality requirements of the markets, which causes additional demands on the process control. In this paper we demonstrate the optimization and modeling application for the EPS-batch process. The application consists of a production optimization tool and a simulation tool for the process parameters based on the Multi-Layer Perceptron network (MLP) with retrain properties. The software can be used at the operational level as well as at the process management level. The features are programmed into standalone software built up in the Matlab environment and it is tailored for the needs of the EPS production company StyroChem Ltd. The results show that the application can offer an efficient tool to economize production and storage costs.