Hrama, MykhailoSidletskyi, VictorElperin, Igor2023-10-302023-10-302022Hrama, M. Intelligent automatic control of sugar factory evaporator operation using behavior prediction subsystem / M. Hrama, V. Sidletskyi, I. Elperin // Ukrainian Food Journal. – 2022. – Vol. 11, no. 1. – Pp. 148–163https://dspace.nuft.edu.ua/handle/123456789/41297Introduction. The aim of the presented research was to substantiate the intelligent automatic control of the sugar juice evaporation with the subsystem for behavior prediction, which allows to determine the behavior of the automatic system. Materials and methods. The operation of the evaporator unit with system behavior prediction to regulate the sugar juice level was investigated. Capacitive level gauges were used as a sensor in the automation scheme of sugar juice level control. Pneumatic seat valves with a built-in throttle and an electro¬pneumatic converter were used as actuators. Results and discussion. The use of neuro-fuzzy regulators occurs only in some specific cases of intelligent control of the evaporation process. There is no data comparing the use of intelligent regulators with classical ones and the possibility of combining several types of intelligent regulators, as well as clear means of predicting their work. Therefore, in the present study, a prediction method was used to compare methods to regulate the level of sugar juice in the evaporator. This made it possible to predict the behavior of the system during the formation of the control action and display the finished forecast on the operator's screen, which made it possible to increase the efficiency of the evaporative station. Statistical data on the behavior of the automation system contours in various operating modes were collected using intelligent and classical controllers, and a model was built to determine the operation of the evaporator using the local trend method and the modified algorithm of prediction. The advantage of this method is its easy and fast implementation, which does not require large economic and energy costs. The accuracy of the prediction model was 98% for the PID-controller, 95% for the fuzzy-controller and 96% for the neural network. The obtained model of the system prediction is stable because the absolute error does not change when dividing the time series into intervals. Conclusions. The proposed system of intelligent automated control of the evaporation of sugar juice with a modified prediction method based on local trends has an insignificant delay, while prediction is performed with high accuracy and stability.ensugarevaporatorneuro-fuzzyregulatorscontrolbehaviorpredictionцукорвипарникнейронечіткийрегуляторикеруванняповедінкапрогнозкафедра інформаційних технологій, штучного інтелекту і кібербезпекикафедра автоматизації та комп'ютерних технологій систем управління ім. проф. А.П. ЛаданюкаArticle