SYNTHESIS OF PLANTWIDE QUALITY PREDICTION SYSTEM FOR A SINTERING PLANT

Abstract To substantially improve operational control of complex industrial plants, it is important to be able to predict the influences which process variables have on quality-relevant target variables. The determination of influencing relationship and also the delay-time behaviour of various industrial processes makes it difficult to synthesis a quality prediction system. A hybrid modeling method combined with a suitable classification of process characteristics ensures a widespread model synthesis for quality prediction.