Oil refining is a multi-stage process of physical and chemical crude oil processing, the result of which is the production of a petroleum products complex. The resulting products are directly related to the crude oil processed characteristics. Requirements for petroleum products quality are increasing, the need for heavy fuels is decreasing, and the raw materials quality is decreasing. All this hinders global oil refining development, and deep petroleum raw materials processing using secondary processes is of particular importance. Increasing competition in the field of fuels and lubricants and modern standards create a need to increase the depth of oil refining and improve the quality of produced oils, coke, and, above all, fuels. The optimizing problems operating modes of complex technological systems and their main apparatuses of oil refining, chemical, and other industries based on their mathematical models are among very relevant scientific and practical problems. In practice, solving the optimizing problem operating modes of complex technological systems is complicated by the fact that these systems are characterized by complexity, many interrelated parameters, multi-criteria, which are usually contradictory, as well as an initial information lack. All this greatly complicates the developing mathematical model's process necessary for preparing and making decisions in the face of conflicting criteria for choosing optimal operating modes of technological systems. Optimal modes of their operation, which are determined by making effective decisions based on mathematical models, allow for increasing production efficiency. In the article, based on a process study occurring in a catalytic reforming unit, attention is paid to an urgent problem - the search for optimal conditions for carrying out the process under consideration and the objective function optimum. For this purpose, by the specifics, optimization problem statement, and the mathematical models formed for the reactors, an effective optimization method, the gradient method, is justified and selected. Here we propose an algorithm for identifying and adapting the resulting models to current situations (states). Next, to optimize the functioning of the reactor block as a whole, by the existing mathematical models of reactors, an optimization algorithm is developed that provides a quick and effective search of the optimal regime parameters for each reactor of the block under consideration. Keywords: technological process, oil refinery, petroleum products, control object, optimal control, mathematical model, optimization problem.
Read full abstract