Background. Due to the absence or limited information about the dynamics of the object, the problem of synthesizing control algorithms of nonlinear technological objects, such as a chemical reactor occurs. At the same time, the static characteristic of the object in the control channel is given and has an extreme nature. Information about the form of the static characteristic and the structure of substance transfer flows in the reactor makes it possible to develop a simplified model that reflects the nonlinearity and inertia of the object. Thus, it is necessary to develop methods of synthesizing a control system using a simplified model and to study its performance.
Materials and methods. To synthesize algorithms for controlling the concentration of the target product in a chemical reactor, the authors have used Analytical Design of Aggregated Regulators method, applying a nonlinear static model of the object and a linear model characterizing the inertia. The issue of the extremum point on the static characteristic has been addressed by introducing a piecewise-constant function into the control algorithms.
Results. Three control algorithms have been synthesized. They differ in having an integral component and structure. The performance of all three algorithms has been demonstrated. The authors have synthesized control algorithms for an extreme nonlinear system – a liquid-phase chemical reactor with a consecutive-parallel reaction. The problem has been addressed under conditions of incomplete mathematical modeling and the availability of only an extreme static characteristic in the control channel.
Conclusions. An unconventional approach to the synthesis of control systems for extreme objects (using a chemical reactor as an example) has been proposed. It provides a simplified dynamic model of the object and analytical synthesis of control algorithms using Analytical Design of Aggregated Regulators method. The performance of control system is shown. The proposed synthesis method is recommended to solve problems of control systems development for similar objects of different physical nature with limited information on the dynamics of the object.