دانلود مقاله ISI انگلیسی شماره 18944
ترجمه فارسی عنوان مقاله

مشکل برنامه ریزی تولید کارگاهی انعطاف پذیر در دنیای واقعی مبتنی بر الگوریتم تکاملی ترکیبی: رویکرد ارائه دهنده خدمات کاربردی

عنوان انگلیسی
Hybrid evolutionary algorithm-based real-world flexible job shop scheduling problem: application service provider approach
کد مقاله سال انتشار تعداد صفحات مقاله انگلیسی
18944 2004 14 صفحه PDF
منبع

Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)

Journal : Applied Soft Computing, Volume 5, Issue 1, December 2004, Pages 87–100

ترجمه کلمات کلیدی
انعطاف پذیر برنامه ریزی تولید کارگاهی - اولویت دیسپاچینگ قوانین - الگوریتم تکاملی - ارائه دهنده خدمات کاربردی -
کلمات کلیدی انگلیسی
Flexible job shop scheduling, Priority dispatching rules, Evolutionary algorithm, Application service provider,
پیش نمایش مقاله
پیش نمایش مقاله  مشکل برنامه ریزی تولید کارگاهی انعطاف پذیر در دنیای واقعی مبتنی بر الگوریتم تکاملی ترکیبی: رویکرد ارائه دهنده خدمات کاربردی

چکیده انگلیسی

This paper presents an approach for scheduling of customers’ orders in factories of plastic injection machines (FPIM) as a case of real-world flexible job shop scheduling problem. The objective of discussed work is to provide FPIM with high business speed which implies (a) providing a customers with convenient way for remote online access to the factory’s database and (b) developing an efficient scheduling routine for planning the assignment of the submitted customers’ orders to FPIM machines. Remote online access to FPIM database, approached via delivering the software as a Web-service in accordance with the application service provider (ASP) paradigm is proposed. As an approach addressing the issue of efficient scheduling routine a hybrid evolutionary algorithm (HEA) combining priority-dispatching rules (PDRs) with GA is developed. An implementation of HEA as a database stored procedure is discussed. Performance evaluation results are presented. The results obtained for evolving a schedule of 400 customers’ orders on experimental model of FPIM indicate that the business delays in order of half-an-hour can be achieved.

مقدمه انگلیسی

Until recently, the role of the production factories had been associated with the manufacturing of a high volume of low-cost and high-quality goods. However, an evolution of these features is lately observed as a result of the recently emerged trend in the major world’s economies of decreasing the rate of economic growth. Still maintaining the importance of producing low-cost and high-quality goods, the relevance of the high manufactured volume is going to be gradually replaced by the role of the high business speed—the ability to react quickly in submitting and modifying the customers’ orders. The high business speed implies that factories should provide the customers with services such as remote submission of orders in operative mode; prompt feedback allowing for customers’ awareness of the anticipated due dates of their orders as well as awareness of the expected ratio of tardy orders and their respective delays; and services providing the ability to promptly follow the current state of the customers’ orders. Within this context, the objective of our research is to investigate the feasibility of developing a scheduling system for factories of plastic injection machines (FPIM), emphasizing on providing the mentioned above customers services needed for achieving factory’s high business speed. Fulfilling our objective implies addressing of the following two main tasks. First, allowing for submission of orders and tracking their statuses requires providing a convenient way for remote online access to the factory’s database. And second, allowing for prompt customers’ awareness about the anticipated due dates of their orders assumes developing of efficient (both in terms of runtime and quality of solution) scheduling routine for planning the assignment of the submitted customers’ orders to the factory’s equipment. Our work is intended to address these main tasks, and its contents could be viewed from three different aspects, representing the following three layers of abstraction of the developed scheduling system: • Problem aspect—the task from the specific problem domain intended to be solved. • Aspect of algorithmic paradigm—the algorithmic paradigm employed to solve the problem. • Implementation aspect—the system architecture used to solve the considered problem exploiting the adopted algorithmic paradigm. The discussion, presented in this document, is attempting to highlight these aspects of our work, and the remaining of the paper is structured as follows. Section 2 briefly explains the problem aspect—a real-world problem of scheduling of factories of plastic injection machines (FPIM) as an instance of the class of flexible job shop scheduling problem (FJSS). Section 3 discusses the aspect of algorithmic paradigm—the main attributes of the hybrid evolutionary algorithm, we developed to solve the targeted FPIM FJSS. Section 4 considers the implementation aspect—the application service provider (ASP) approach, focusing on developing of three-tiered Web-based system architecture. Performance evaluation results are given in Section 5. Finally, Section 6 draws a conclusion and discusses some directions for future work.

نتیجه گیری انگلیسی

We proposed an approach for solving the problem of scheduling the customers’ orders in FPIM as a case of real-world JSSP. The objective of our work is to provide the FPIM with high business speed which implies addressing of the following two main issues: (a) providing a convenient way for remote online access to the factory’s database and (b) developing an efficient (both in terms of runtime and quality of solution) scheduling routine for planning the assignment of the submitted customers’ orders to the FPIM machines. The first issue is addressed by the proposed approach of delivering the software as a service in accordance with the ASP paradigm, which offers significant commercial and organizational benefits such as easy software maintenance and future upgrade, low cost of entry into the business (especially for small and medium scaled FPIM), and considerably less expensive pay-as-you-go model. The issue of efficient scheduling routine is addressed by developed HEA which combines the approaches of using PDR with GA. PDR-based approaches offer the advantage of simplicity, featuring low computational cost and can therefore be applied to complex real-world problems such as FPIM FJSS. GA, incorporated into proposed HEA addresses the issues of the myopic nature of PDR and the necessity to empirically evolve the most suitable PDRs and their combination. Implementing HEA as a database SP offers the benefits of reduced communication network overhead and improved performance characteristics Performance evaluation results obtained for evolving a desirable (without tardy orders) schedule of 400 customer’s orders on experimental model of FPIM indicate that the business delays are in order of half an hour. We are intending to explore the following two approaches to future reduce the business delays. The first approach is aimed at reducing the computational effort of HEA and it would exploit the continuous nature of the scheduling process. Taking into consideration the empirical observation that newly submitted orders are unlikely to be scheduled in a way that requires significant modifications to the orders, scheduled earlier, we are interested in the feasibility to incorporate few of the best schedules from previous run into the initial population of the current run. The second approach is intended to improve the overall performance of HEA by inducing a noise [15] in fitness evaluation -instead of creating and evaluating the whole schedule, it is much faster to create and evaluate only the initial part of it and to make a judgment about the fitness of the whole schedule. The preliminary obtained results are encouraging in that varying the amount of the induced noise a tradeoff between the improved computational performance and the deteriorated computational effort can be achieved, leading to the better overall performance of HEA.