Optimization control method with system condition online monitor for cleaning process

The cleaning process is an important step in sugar refinery and features strong nonlinearity, a long sampling time to obtain quality data, multiple constraints imposed by practical requirements, its control goal is to acquire high density products. This paper presented an intelligent optimization control method, using improved back-propagation neural networks modeling of quality prediction and system condition online monitor, optimization be realized by c-means clustering, genetic and chaos approaches. The control system is implemented by DCS. The results of actual runs demonstrate the validity of the method.