DonNTU   |   FCST   |   Masters portal

RU |  UA

• Resume
• Abstract

1. Paper objectives and tasks
2. Introduction. Justification of paper topicality
3. Alleged scientific novelty
4. Alleged practical value
5. Survey of researches and problem elaborations on the topic
6. Research task formulation
7. The Ant Colony Optimization algorithm (ACO) description
8. Graph model
9. Conclusion
10. USEDLIST OF USED LITERATURE

Paper objectives: to develop a computerized operative job design subsystem of automatic processing cutting department for increasing effective of its work owing to optimal work planning in the conditions of multicriteria problem solving.

Paper tasks:

• to describe the mechanical department work and to assign the planning task correctly.

• to analyze the methods, the models, the algorithms and instruments for the optimization of operational production planning.

• to make the choice and the justification a method for the optimization of operational production planning.

• to develop and realize algorithm for the optimization of operational production planning.

Nowadays we can see the rapid development of the ICT and their adaptation in all the spheres of man’s activity. The maching-building enterprises need badly the integrated adaptation of information control systems. The application of modern IT in the field of records production management can help regulate documents circulation, take into account and minimize losses, to raise productivity of work, to reduce coats and others [1].

For the machine-building enterprises with different industrial structures one of the most important task of intradepartmental planning is the establishment of machine utilization planned schedule [2].

Computerized systems allow to raise essentially the industrial control efficiency at the enterprise. The adaptation of such systems at the enterprise would help control the movement and production of details and blocks, because the task of operational planning influences significantly the results of the enterprise in the whole [1].

Till now there is not studied a driving of kits of parts immediately in the industrial environment, on a production site according to a process flow. Creation of program complexes for modeling and control of operation of production sites, compilation of close to optimal schedules of operation of the equipment, is the actual scientific and technical task [3].

This task already dared by means of ant algorithms where reversal of "ants" could be estimated as driving of kits of parts on a production site. And scientific novelty of my operation consists that for "ant" I will take the equipment (the floppy industrial unit).

It will allow in practice to meet the challenge of operational planning quickly and efficiently. The developed program module will allow to cut the costs of material resources and timing budgets during the production process planning, it will raise the standards of information processing and processing effect, excepting the errors, connected with human factor.

• global (worldwide)

A wide range of software products has been created. They can be referred to ERP-systems, e.g. LEKIN and TACTIC. We can name such foreign authors as Rong Zhou, Jun Zhang, Alan S. Manne and others.

• national (what is done on this topic in the country outside the DonNTU)

It was found a lot of interface papers with workable subsystem. One can also mark out existing elaborations on this topic: production planning system and dispatch control system “Phobos”, (MRP-II) ERP-systems “Compass”, Zenith SPPS and others.

• local (based on the materials of DonNTU lecturers professors, postgraduates and undergraduates)

According to the results of search among the materials of DonNTU masters portal the papers of E.A. Andreyenkova (Podolyaka), V.G. Bashev, D.V. Suhorukov have been founded. The problems of operational planning in different fields were considered there. As for the studies of our university lecturers on this topic, the papers of A.I. Sekirin were founded.

In general, machine utilization planning consists in the following: there are details, which are required to be machined. It is necessary to define such sequence of part intake for treating to each machine that the machine-down time (the total part cutting time) was minimal.

Control task is to ensure productivity according to quantity production program (N_pl) at a stated time, using the resources effectively (R_l) if there is disturbance (V_k). in general, high performance assuring of using the resources (R_l) and department operation is achieved owing to the optimization of equipment operation schedule.

The process of part-production D_i (i=l, ..., п) is divided into production steps О_ij (i=1,..., п; j=l,..., m). Each step can be regarded as:

where Н_ij – group number of the processing equipment; Т_ij – the duration of the operation.

Operational order is the sequence of the operations, which I^th part passes during production process.

If t_ij is the time of beginning of operation О_ij and t_ij' is the moment of finish date, then we have the equality:

The population of data {t_ij} (i=1,..., n; j=l,..., m), which satisfies all technological and temporal constraints is scheduling.

Restrictions:

1. production volume restrictions:

N_pl=N_f ,   (5)

where N_f is produced part count of I^th-type (i=l,...,K);

N_pl is call-off quantity part of I^th-type in the production program.

К – product mix of produced parts.

2. due date restrictions:

Т_pl ≥ Т_f ,   (6)

where Т_f is real due date of the I^th part (i=l,...,K);

Т_pl is schedule due date of the I^th part.

3. restriction of the process equipment running time:

,   (7)

where Т_ij is production step duration;

R_i is the resource of the I^th group equipment.

Performance test:

1. Minimization of the machine-down time

T_оpt = Т → min,   (8)

where Т is total time of the production cycle.

,   (9)

where Т_ij is the cycle duration of the production step of the I^th kit of parts;

α_ij – downtime duration of the J^th production step over the I^th kit of parts.

2. Maximization of the machine utilization

К_{оpt u} = 1/m∑K_u → max,   (10)

where m is equipment amount;

K_u is duty factor of the I^th equipment.

,   (11)

3. Minimization of part disuse time

,   (12)

,   (13)

where α_i is storage expenses of the I^th kit of parts.

There are also other criteria for the choice of the optimal schedule of machine usage [2].

The main idea of the simple Ant Colony Optimization algorithm is modeling of ants’ links, collective adaptation. The colony represents a system with very simple rules of individual autonomic links. However, in spite of the primitive links of every individual ant, the links of all the colony are quite sagacious. So, the source of the links of ant colony is low-level interaction, due to which, in the whole the colony represents sagacious multiagent system. The interaction is determined through the chemical substance, pheromone, which they lie down on their paths. While choosing the movement direction, the ant wants to travel the shortest path as well as the experience of other ants. This ant gets information through the pheromone’s level. With time, the pheromone evaporates and this is a negative feedback [5].

Planning task refers to the optimization of discrete process. Recently, for the search of optimal solution, the Ant Colony Optimization algorithms have been widely used. They are always based on the graph model of the object or of the task. So, let us consider the graph model construction of automatic putting work section.

The elements of the graph are vertex (point) and lines which connect individual points of the graph (Fig. 1). The graph reference vertex specifies the beginning of plan fulfillment (start point). It is supposed to place as many ants in this vertex as many machines they have at the production area. The other graph vertexes are divided into levels, each of them corresponds to individual production step. The number of vertexes on the first and the second level equals the number of parts types which were planed to produce. They can have less vertexes on the other levels if the checklist of parts output contains only two operations.

Thus, the vertex is reference designation of the operation at this stage, which was determined by flexible machining cell and the line is characterized by the probability of the ant’s pass from one operation to another one.

Figure 1 – Graph for operational scheduling assignment of the production area.

The probability of ants’ pass from the start point to the first level vertexes can be calculated with the use of checklist time parameter and with a glance of order timing.

where T_i is the operation time;  T_i^tim is order timing.

The probability of the further passes can be calculated with a glance of work executed and remaining time until the kit of parts is produced.

If it is possible to pass from the start point only to the first level vertexes, the further passes provide the combination of the vertexes of the same level and occurrence of the bails that signifies the continuation of this operation with the next kit of parts [6].

The automation of compute-assisted manufacturing machinery and auxiliary equipment with the development of machine-building is one of the main directions of technological progress in machine-building industry. The equipment must guarantee the fulfillment of preplanned assortment of production steps at the factory at a stated time and be able to respond efficiently to the change of equipment status and factory order.

Working on this research paper we have analyzed existing methods, models and approaches for operational planning problem solving. The algorithm and the graph model of working processes and work effectiveness increase of cutting department on basis of suboptimal schedule of run-time have been developed.

Hereafter, the algorithm for optimization of the production area work will be developed and worked in practice by means of software.

1. Первозванский А.А. Математические модели в управлении производством // Главная редакция физико-математической литературы издательства «Наука», М. – 1975, 616 с.

2. Сытник В.Ф. АСУП и оптимальное планирование // Издательское объединение «Вища школа», Киев. – 1977, 312 с.

3. Секирин А.И. Программный комплекс для моделирования, анализа и оптимизации работы автоматизированных технологических комплексов механообработки // Донецкий национальный технический университет, кафедра автоматизированных систем управления.

4. Ходашинский И.А., Горбунов И.В., Дудин П.А. «Алгоритмы муравьиной и пчелиной колонии» // М. - 2007.

5. Штовба С. Статья из журнала Exponenta Pro: Муравьиные алгоритмы [электронный ресурс]. – Режим доступа: http://rain.ifmo.ru/.../ant-algo-2006/article.pdf

6. Габалис Е.Ю., Савкова Е.О., Жукова Т.П. Графовая модель планирования работы автоматизированного технологического участка механообработки деталей // Збірка матеріалів IІ всеукраїнської науково-технічної конференції студентів, аспірантів та молодих вчених – 11-13 квітня 2011 р., Донецьк, ДонНТУ – 2011, с. 49-53.

7. Лаздынь С.В., Секирин А.И. Оптимизация расписаний работы автоматизированных технологических комплексов механообработки с использованием генетических алгоритмов // Международный сборник научных трудов “Прогрессивные технологии и системы машиностроения”, выпуск 25. –Донецк: ДонНТУ.-2003, c. 198-203.

8. Исполнительный производственные системы. Официальный сайт компании Fobos-mes [электронный ресурс]. – Режим доступа: http://www.fobos-mes.ru/russian-MES/index.php

9. Независимый ERP-портал [электронный ресурс]. – Режим доступа: http://www.erp-online.ru/.../kompas/

10. Официальный сайт LEKIN [электронный ресурс]. – Режим доступа: http://www.stern.nyu.edu/.../lekin/index.htm

11. Официальный сайт TACTIC [электронный ресурс]. – Режим доступа: http://www.waterloo-software.com/

IMPORTANT REMARK

During the writing this abstract the master's work wasn’t finished yet. The final ending is planning on December, 2009. The complete text of work and all the information about work you can receive from the author or him scientific adviser after the mentioned date.