Порівняння Роботи Підсистем Прогнозування В Автоматизованій Системі Керування Випарним Апаратом

Abstract

The purpose of the presented research is to substantiate the methods of steam pressure regulation in the evaporator with the forecasting subsystem, which will allow predicting the behavior of the system and the study of the influence of the number of time intervals of the forecasting algorithm on the accuracy of the forecast of the evaporating plant operation. Development of a data exchange system between the level of the technological process and the level of production. The operation of the evaporator with the forecasting subsystem of steam pressure regulation is investigated. In the steam pressure control automation scheme, PC-28 pressure transducers are used as sensors. Pneumatic seat valves with a built-in throttle and an electro-pneumatic transducer are used as executive mechanisms. The use of neurofuzzy regulators occurs only in some specific cases of intelligent control of the evaporation process, there are no data comparing the use of intelligent regulators with classical ones, the possibility of combining the ope­ration of several types of intelligent regulators, as well as clear means of predicting their operation. Therefore, in this work, the forecasting method was used to compare the methods of regulating the steam pressure in the apparatus, which made it possible to predict the behavior of the system du­ring the formation of the control action and to display the ready forecast on the operatorʼs screen and, thus, to increase the efficiency of the evaporation station. Statistical data on the behavior of automation system circuits in different operating modes using intelligent and classical regulators were collected, and a forecasting model of the operation of the evaporation plant was built using the local trend method, and the forecasting algorithm was modified. The advantage of this method is its easy and quick implementation, which does not require large economic and energy costs. A model for forecasting the operation of the evaporation plant was built using the local trend method and a forecasting algorithm was developed. The accuracy of the obtained prediction model was also evaluated. The accuracy of the prediction model was 97 % for the PID controller, 97 % for the neurofuzzy controller, and 96,5 % for the neural network when using 9 intervals, which is higher than the accuracy when using 6 intervals. The proposed model for predicting the operation of the evaporation plant is characterized by high accuracy as a whole, but during the occurrence of fluctuations in the transient process, there is an insignificant delay in forecasting these fluctuations. The accuracy of this model directly depends on the increase in the number of time intervals during the development of the forecasting algorithm.