Simulation of a Multi-connected Process in iThink Program

摘要

Quite often, multiply connected processes with parameters that affect each other can be observed in operation at many industrial facilities. In the article, the item to be automated is a reactor for the synthesis of a ready product, in which there are parameters that have a mutual influence on each other. To build an effective control system, the use of classical PID controller for this object is not possible. In the course of the research, a control system for the electrical heating of the reactor was developed, taking into account the mutual influence of the control loops on each other. The article describes the development of a control system for electrical heating of the reactor with the joint use of a fuzzy controller and neural networks in the Matlab environment to compensate for the mutual influence of control loops. The paper presents a software implementation of the reactor electric heating control system, which can, depending on the given input parameters, monitor and control the reactor electric heating process using neural networks, displaying the dynamics in form of parameters with respect to time. The control system can be implemented for reactors of different applications. It is also possible to use similar solutions for different kinds of complex objects.