Development of models and program solutions for machinery plant production management system^*

Preface.

The problems of industrial processes optimization are widely spread in the field of efficient production management. They belong to the family of complex optimization and combinatorial problems (the real dimension problems belong to the class of NP-complete ones).The development of a common analytical production model still represents an unsolved problem for the scientists of today. But the great progress in the field of computer technologies provides them with means of imitation modeling and net techniques that can be well adopted to the creation of effective production management solutions. Thus, the development of an effective and flexible imitation model of industrial process becomes one of the problems of today.

General scheme of industrial process optimization technique.

Figure 1 represents the outline of industrial process optimization.

Figure 1 shows that the solution of industrial process optimization problem can be represented by 2 main operations:

• development of industrial process model;
• elaboration of optimization algorithm.

The optimization apparatus is currently represented by the techniques that allow of no strict mathematical optimum substantiation. These techniques include heuristic methods (common sense- and expert experience-based algorithms), genetic algorithms, neural networks and diverse hybrid methods.

As for the industrial process modeling, the simulation modeling method stays a strong leader in this field. The development of the analytical production process model is extremely complicated. Therefore the major part of solutions does not result in creation of an accurate model. It’s also inadmissible to experiment on a real model. Thus, today the simulation modeling represents a fundamental tool to obtain the production system parameters.

At the best, the simulation model run-through ensures the results at the only point of solution search space. Therefore it is necessary to conduct a series of experiments on the imitation model that will cover a wide range of solutions. In the traditional simulation modeling systems the purposefulness of such experiments is ensured by a developer specialist.

Analysis and expediency grounding of a modeling method.

A great diversity is not a common feature of the complex system simulation methods in general and of the production process simulation methods in particular. At present the commonly used methods are represented by

• simulation modeling;
• network modeling.

The main goal of this project is to develop an object-oriented library of machinery construction components that will be subsequently used for the elaboration of the production simulation model. The simulation modeling expediency is present together with one of the following conditions:

1. there is no complete mathematical formulation of the problem in question. Or the well stated mathematical model still doesn’t have the analytical solution methods;

2. there are analytical solution methods, but they include the mathematical procedures that are too complicated and time-consuming. In this case the imitation modeling represents a more simple and efficient solution technique;

3. in addition to the evaluation of certain system parameters it is necessary to observe the process run during certain period of time. The imitation modeling serves as a great tool to implement the observation task in this case;

4. it is impossible to find a way of experimenting and conducting the observations on a real process in a real-time environment;

5. the processes in question are of long-duration, so there is a need of time scale compression.

The following statements describe the drawbacks of simulation modeling technique:

1. the development of a high quality simulation model represents an expensive and often time-consuming process requiring work of qualified specialists;

2. simulation model could be inaccurate and there are no techniques that would allow to evaluate a degree of it’s inaccuracy. The only way to overcome this difficulty is to analyze the model’s sensitivity towards certain parameter changes;

3. due to long-duration of an experiment with aid of the simulation model, the simulation modeling technique is not for use in real-time control and management systems;

4. simulation models are not flexible enough to reflect all of the production environment changes. There’s a need of much time and efforts to addopt a certain model for a certain set of changing parameters.

As a part of a production management system the simulation model represents one of the forms of controled object’s data representation. Together with management system’s information maintenance the simulation model serves to support a decision making process. It also serves as a data mining tool. The use of simulation modeling in the process of operational management provides for the solution of the following problems:

• production planning;

• external deliveries planning;

• production process forecasting to detect the probability of a production plan fulfillment and to detect the need of certain corrections;

• evaluation and selection of management strategies.

Project objective.

The main goal of this project is to develop an object-oriented library of machinery construction components and to create the production process simulation model using this library. The simulation model is used to work out optimal time-tables for the industrial plant sub-units. The evaluation of a time-table efficiency is conducted by the following criteria:

• time duration of machine parts production cycle (Tc(time of cycle)->min);
• average machinery load coefficient (Cl(coefficient of load)->max).

• load coefficient for all types of machinery used in production process;
• average queue length for each type of machinery;
• machinery idle time duration;
• time duration of production cycle.

Results.

The use of object-oriented library containing machinery construction components would allow to speed up and to simplify the development and the flexibility of an simulation model in the process of it’s creation.

Literature.

1. Мосталыгин Г.П., Толмачевский Н.Н. Технология машиностроения. – М.:Машиностроение, 1990. - Учебник для вузов по инженерно-экономическим специальностям – 288 с.: с ил.
2. Непомнящий Е.Г. - Экономика и управление предприятием. Конспект лекций. - Таганрог: Изд-во ТРТУ, 1997. Электронный вариант http://www.aup.ru/books/m83/15.htm
3. Соколицын С.А. и др. Многоуровневая система оперативного управления ГПС в машиностроении. – Спб.: Политехника, 1991. – 208 с.
4. Смоляр Л.И. Модели оперативного планирования в дискретном производстве. - М., 1978. - 320 с.: с ил.
5. Шкурба В.В., и др. Планирование дискретного производства в условиях АСУ. - Техніка, 1975. - 296 с.
6. Каменицер С.Е. и др. Автоматизированная система управления машиностроительным предприятием. – М.:«Машиностроение», 1971. - 272 с.
7. Смолин Д. В. Введение в искусственный интеллект: конспект лекций. - М.:ФИЗМАТЛИТ, 2004. - 208 с.
8. Усаков А.А. Принципы построения систем управления с нечеткой логикой. // Приборы и системы. Управление, контроль,диагностика. №6. 2004.
9. Круглов В.В. И др.Нечеткая логика и искусственные нейронные сети: учеб. пособие для вузов. - М.: Физико-математическая литература, 2001. - 224 с.
10. Горнев В.Ф. и др. Оперативное управление в ГПС. – М.: Машиностроение, 1990. – 256 с.: с ил.
11. Емельянов В. В. и др. Теория и практика эволюционного моделирования. — М.: ФИЗМАТЛИТ, 2003. - 432 с.
12. Егорова Т.А. Организация производства на предприятиях машиностроения – Спб.: Питер, 2004. – 304с.: с ил.
13. Климов А.Н., и др. Организация и планирование производства на машиностроительном заводе. – Л.: Машиностроение, 1979. – 463 с.: с ил.
14. Ипатов М.И. и др. Организация и планирование машиностроительного производства: Учеб. Для машиностр. спец. вузов. – М.: Высш. шк., 1988. – 367с.: с ил.
15. Фаулер М., Скотт К. UML.Основы. - Пер. с англ. - Спб.: Символ-Плюс, 2002. - 192с.: с ил.
16. Трофимов С.А. CASE-технологии: практическая работа в Rational Rose - м.: ЗАО "Издательство БИНОМ", 2001г. - 272 с.: с ил.
17. Шеннон Р. Дж. Имитационное моделирование систем - искусство и наука. М.: Мир, 1978