ЧИСЛЕННЫЙ ПОИСК ПАРАМЕТРОВ МАТЕМАТИЧЕСКИХ МОДЕЛЕЙ ПИЩЕВЫХ ПРОДУКТОВ ИЗ ГИДРОБИОНТОВ ДЛЯ ПРОЦЕССОВ ПАСТЕРИЗАЦИИ

Abstract

The article highlights the numerical search for the mathematical model parameters of food products from hydrobionts put into hermetically sealed canning containers. The mathematical description of the canned products was given as transfer functions. The input value is a retort short-term temperature-time dependence and the output value is the product temperature dynamics. The search for the transfer function coefficients was performed by the numerical optimization method based on the plant response. The product temperature response data was obtained experimentally with thermologgers of the “iButton Data Loggers” series during canned food pasteurization. The numerical search for the mathematical model parameters of food products was carried out using the software AutoCont Lite: SeekerC developed by the author at the Automation and Computer Engineering Department of Murmansk State Technical University. A first-order factor was chosen as 
 a candidate for a product model in a tin can, and for a product in a glass container there was chosen a third-order factor. To solve the problem of numerical search for model parameters the integral estimate was employed as an optimization criterion. It is calculated as the square deviation between the output values of the model and the plant over the investigation period. The research resulted 
 in mathematical models of fishery-canned products obtained in the transfer function form. The obtained result was applied for modeling and subsequent automatic control system implementation for the industrial autoclave ASCAMAT-230 during the canned food pasteurization.