ЖУЧЕНКО О. А.

Сортировать по умолчанию названию
  • Врахування теплоти дисипації в моделюванні процесів екструзії полімерів і керуванні ними

    Обрано методику розрахунку потужності дисипації, що дозволяє врахувати її під час синтезу систем керування режимом роботи одночерв’ячного екструдера.

    Переглянути
  • Параметрична схема скловарної печі

    Створено й досліджено параметричну схему ванної скловарної печі з поперечним напрямом полум’я.

    Переглянути
  • Передатні функції елементів скловарної печі

    Розглянуто алгоритм виведення елементів передатних функцій скловарної печі, як об’єкта автоматизації. Результати в подальшому будуть використані для створення системи керування подачею газу.

    The algorithm of deducing of elements of transfer functions of the glass-making furnace, as object of automation is considered. Results will be used further at creation of a control system of giving of gas.

    Переглянути
  • Моделювання теплового режиму скловарної печі

    Розроблено математичну модель теплового режиму скловарної печі. Одержано розподіл температур за різних режимів роботи.

    The mathematical model of thermal conditions of glass furnace is conducted. The temperature profiles are got at different operating conditions.

    Переглянути
  • Моделювання теплового режиму конверсії метану

    Розроблено математичну модель теплового режиму конверсії метану. Одержано розподіл температур за різних режимів роботи.

    The mathematical model of thermal conditions of methane conversion is conducted. The temperature profiles are got at different operating conditions.

    Переглянути
  • Математичне моделювання розігріву одночерв’ячного екструдера

    Досліджено розігрів одночерв’ячного екструдера. Одержано розподіл температур за різних режимів і потужностей нагрівників. Результати дозволяють оптимізувати кількість нагрівників.

    The heating-up of single-screw extruder is conducted. The temperature profiles are got at different operating conditions and heater power. The results allow choosing the optimum amount of heaters.

    Переглянути
  • Дослідження апроксимуючих моделей об’єктів із розподіленими параметрами в частотній області

    Показано, що навіть для простого об’єкта з розподіленими параметрами побудова його передатної функції є непростою задачею. Установлено, що структура апроксимуючої моделі залежить від умов функціонування об’єкта керування, зокрема, від частотної області його роботи. Розглянуто аналітичні методи побудови апроксимуючих моделей.

    It is shown, what even for simple object with the distributed parameters construction of its transfer function is an uneasy problem. Besides, in work it is shown that the structure of approximating model depends on operating conditions of considered object of management, namely, from frequency area of its work. It is considered analytical methods of construction of approximating models.

    Переглянути
  • МОДЕЛЮВАННЯ РЕЖИМУ ПУСКУ ЕКСТРУЗІЇ ПОЛІМЕРІВ

    Наведено новий спосіб моделювання режиму пуску екструзії полімерів, що базується на нейронних мере-жах. Досліджено декілька різних структур моделі. Як критерій оцінювання якості нейромоделей викорис-тано накопичену суму квадратів помилок. Як навчальний алгоритм застосовано алгоритм зворотного роз-рахунку помилки. У процесі синтезу системи керування використано дані реального процесу екструдування. Показано, що найкращим чином працює нейромережа, що включає три нейрони у схованому шарі та запіз-нення на два кроки. Визначено оптимальну структуру та параметри нейромережі.

    The paper presents a new approach to the modeling of the start-up part of a polymer extrusion process, based on a neural network. Some different model structures were researched. As a performance criterion in estimating the quality of the neural network model the cumulated squared error over a whole training sequence of data was used. The backpropagation algorithm was used as a training algorithm. The design steps, which included choosing the network topology, training, validation and testing, were illustrated with data taken from a real process that uses polymer material.
    It was found that for the extruder system presented in the paper, 3 hidden neurons and 2 input and output delays gave optimum performance. As a result the optimal model structure and parameters of the neural network (weights and biases) were determined.
    Simulation results with real data are also presented and confirm good performance of proposed control system. The neural network model approximated the transient part of the process well even when input variables were quite far away from training data.

    Переглянути
  • ПЕРЕДАТНІ ФУНКЦІЇ ТА ЧАСТОТНІ ХАРАКТЕРИСТИКИ ЦИЛІНДРИЧНОЇ ТЕПЛОАКУМУЛЮЮЧОЇ СТІНКИ: НОВІ ВАРІАНТИ

    Наведено передатні функції й частотні характеристики циліндричної теплоакумулюючої стінки як об’єкта з розподіленими параметрами залежно від межових умов на зовнішній і внутрішній поверхнях. Отримані результати можуть бути використані для синтезу систем керування.


    Almost all real objects control in chemical, food, metallurgical, oil and other industries, in fact in the entire field of human activity are objects with distributed parameters.

    From mathematical modeling point of view a lot of plants in different industries can be considered as cylindrical heat storage walls with distributed parameters. Selected research path based on the following considerations: consider only heat objects with distributed parameters as the most common in the industry; Mathematical model objects with distributed parameters obtained in the form of the transfer function as the most suitable for further research in terms of analysis and synthesis of control systems using existing software. The aim of this study was to receive transfer functions and frequency responses as mathematical models of above mentioned plants. To solve this problem the main equation of heat conductivity was put in basis. In addition three kinds of boundary conditions on external and internal wall surfaces were taken into account. Different combinations of boundary conditions on external and internal wall surfaces result in different kinds of transfer functions and frequency responses. These functions are transcendental and perhaps must be simplified for practical using. It was proposed few nontrivial ways in order to calculate frequency responses of cylindrical heat storage walls. To implement this calculation it is necessary to have available a subroutine for computing certain functions. If necessary, these routines can create their own with the above formulas. You can also use the tables of the functions, forming the basis of their structure interpolation (polynomial, cubic splines or B-splines). But most appropriate looks above numerical integration of differential equations with appropriate boundary conditions on single stepped input signal with zero initial conditions.

    Obtained results can be used for researching of plants, which can be considered as cylindrical heat storage walls with distributed parameters, and for control system synthesis.


    Переглянути
  • МОДЕЛЮВАННЯ РЕАКЦІЙНИХ ТРУБ ПЕЧІ ТРУБЧАСТОГО ТИПУ В КОНВЕРСІЇ МЕТАНУ

    Проаналізовано математичну модель реакційної труби печі трубчастого типу – основного апарата на першій стадії конверсії метану. Одержано систему диференціальних рівнянь для подальших досліджень.


    The most famous technologies of making hydrogen are based on chemical and thermal processes also on electrolysis of water. However, they major shortcomings as the use of high potential energy to the costs of fossil fuels and therefore significant pollution. The disadvantage of electrolysis of water is a significant level of power consumption. Electrolytic hydrogen is the available, but more expensive product. Today, the world's largest distribution obtained technology of developing hydrogen or a mixture of hydrogen and other gases by steam reforming of natural gas – methane. But almost half of the initial volume of gas consumed in carrying out endothermic steam reforming process. So the world is making an intensive search for such technologies that produces hydrogen, which would meet the requirements of economic and energy efficiency and environmental safety.

    The process of steam reforming of hydrocarbon compounds designed to produce hydrogen technology and other process gases by high-temperature catalytic reforming in the cycle of production of synthesis gas. The process of conversion is one of the most important stages of the process of obtaining gas in ammonia synthesis loop. Despite the complexity of hardware design, steam catalytic conversion of methane in the furnace tube is currently the most economical way to get the process gas and hydrogen.

    Therefore, to improve it is relevant and necessary scientific and technical challenge. Given that the full-scale study of the process of conversion is expensive, very difficult and dangerous. To solve the problems of design of process equipment, synthesis and study of control systems, simulators technical personnel, the challenge of obtaining the mathematical model of the process object that will best reproduce the properties of real prototype to a wide range of operation. However, creating of a model is based on a compromise between the complexity of the mathematical tools and depth of simplifications and neglects. Today we know a few tube type furnace models, including regression models, which are widely used because of its ability to approximate analytical relationship between input, output variables for multiple passively or actively conducted experiments, and mathematical models are based on the theory of fuzzy logic and fuzzy set theory. The general scientific problem is that with most of them do not allow modeling the dynamics of the process and, therefore, not part of the unsolved scientific problem is that it makes it impossible to optimize dynamic processes.


    Переглянути
  • МОДЕЛЮВАННЯ НАСАДКОВОГО АБСОРБЕРА У ПРОЦЕСІ ОДЕРЖАННЯ АЦЕТИЛЕНУ З РЕАКЦІЙНИХ ГАЗІВ

    Розглянуто та проаналізовано математичну модель насадкового абсорбера у процесі виготовлення ацетилену з реакційних газів, у якому проходить основний процес насичення ацетилену. Одержано систему диференційних рівнянь для подальших досліджень.


    There are many different ways to produce acetylene. Among them, such as manufacturing out of calcium carbide, pyrolysis of hydrocarbons and gas separation from gases which consist acetylene. All of these technologies are based on chemical or mass transfer processes, etc. However, they are energy consuming and pour on the environment. Production using pyrolysis spends a lot of heat in the process of selection of the product, so there is an opportunity to significantly improve the technical and economic performance using plasma process, but these measures have not been ever brought in real life. If we consider the process of production of reactive acetylene gas using dimethylformamide, it is certain that a large amount of dimethylformamide can spend to achieve the level of concentration in the reaction of acetylene gas. Only the following reasons, the quality of the incoming gas mixture can vary significantly. In this regard, we seek such factors by which the process will be more meet feasibility requirements and reduce the impact on the environment.

    The manufacture of acetylene gas from the reaction gases is designed to produce high quality acetylene by mass transfer processes taking place in absorbers.

    The process of mass transfer in the absorber is one of the most important steps in being saturated with acetylene production cycle. Production of acetylene from the reaction gas is the most economical way of obtaining acetylene because all unused products in the process of returning to a pure storage, where they can be used again in the process. The research on real objects is expensive, very difficult and dangerous, so this improvement is relevant and necessary.

    To solve the problems of design of process equipment, synthesis and study of control systems the challenge of obtaining the mathematical model of the process object that best reproduce the properties of real prototype to a wide range of operation. For this purpose, a mathematical model is needed, which would most closely corresponds to the actual and also based on the complexity of the device and the usage of simplifications, which could be neglected .

    Today we know a few models of packed absorbers, including regression models and mathematical models are based on the theory of fuzzy logic and fuzzy set theory. However, most of them do not allow simulating the dynamics of the process, and thus make it impossible to optimize dynamic processes.


    Переглянути