Paper presents automated method of supervising of complex technological processes. Advanced control algorithms were used with expert knowledge written inside fuzzy models and controllers. The main idea of described supervising method is a multistage compensation of prognosed product's quality deviations, caused by disturbances of low-speed variability variables, by changing values of high-speed controlled input variables of process. Prognosis and process correction are performed in real-time mode, so no defective products are made. Proposed method was applied to supervise PELD molding injection process.
展开▼