course under the condition of the phenomenon of conservative perturbed equilibrium (CPE) are considered. It is shown that the onset of the CPE point - an extremum of the over-equilibrium concentration - occurs at a certain point in time. Therefore, one of the tasks of optimal control is to calculate the time of occurrence of this point and determine the effect of temperature and pressure on its position. The modeling of the TRM process is performed for an ideal displacement reactor. The DWSIM software product was used to model the TRM process. The kinetic equations were programmed in the form of scripts. The position of the over-equilibrium concentration point was also investigated depending on the initial composition of the mixture. According to the modeling results, it was concluded that the position of the CPE point for product H₂ is constant, and depends only on pressure and temperature and does not depend on the initial composition of the mixture. This allows us to conclude that the control can ensure that the product yield with the maximum concentration is achieved. The problem of optimizing the initial composition of the mixture for the methane tri-reforming process using CPE is formulated: the objective function and constraints are defined, and the solution method is chosen. The optimization problem is to maximize the target product of methane tri-reforming, H2. The solution to the problem is the composition of the mixture (concentrations of CH4, H2O, CO2). As for the constraints, the pressure and temperature are limited by the physicochemical properties of the process and the design features of the reactor, the elemental balance must be maintained, the key component at the input must have an equilibrium concentration at a given pressure and temperature, and the concentrations of other components must be deviated from their equilibrium values. A software implementation of the optimization problem based on the Optimization Toolbox Matlab was performed. The research results indicate an increase in the efficiency of the process (increase in the output of the target component) and the possibility of implementing optimal solutions in real time, which opens up prospects for implementation in existing control systems and development of decision-making systems.
Read full abstract