МЕТОД АДАПТАЦИИ МОДЕЛЕЙ ОПЕРАТИВНОЙ ОЦЕНКИ ПОКАЗАТЕЛЕЙ КАЧЕСТВА НЕФТЕХИМИЧЕСКИХ ПРОИЗВОДСТВ (НА ПРИМЕРЕ ПРОИЗВОДСТВА ЭТИЛЕНПРОПИЛЕНОВЫХ КАУЧУКОВ)

Abstract

Problems of production quality indexes efficient determination and ensuring technical and economic effectiveness of process control based on these indexes are to be solved with the use of the models having the parameters which should be periodically updated in accordance with the online analyzer or laboratory data. For the model correction a considerable amount of information is usually needed which is often impossible to obtain within a short period of time. Due to above reason, with technological parameters, raw material quality and equipment parameters being changed, the accuracy of quality indexes evaluation goes down, producing a negative effect on the efficiency of process control. This article suggests an efficient adaptation method which requires minimum amount of experimental data. On the basis of ethylene propylene rubber production, impact of technological parameters and raw material quality (factors) on quality index (Mooney viscosity) has been studied. It is demonstrated that the model coefficients determining the quality index response to the factors are changed insignificantly. This makes it possible to execute model adaptation by changing the bias value with the coefficients remaining unchanged against the factors. Structure of automatic control system based on quality indexes with adaptation contour and algorithm of obtaining quality index values with predetermined accuracy are presented.