بررسي استواري روش‌هاي فراابتكاري براي برنامه‌ريزي مسيله‌ي جريان كارگاهي انسدادي در حالت تصادفي

Investigating the Robustness of Heuristic and Metaheuristic Algorithms for Stochastic Blocking Flow Shop

برنامه‌ي زمان‌بندي و توالي عمليات در مسايل برنامه‏ريزي توليد يكي از عوامل كليدي در موفقيت‌ سازمان‌هاي توليدي است كه باعث جلوگيري از انباشت سرمايه، تقليل ضايعات، كاهش يا حذف بيكاري ماشين‏آلات و كاربرد بهينه‌ي آن‌ها و پاسخ‌گويي به‌موقع به سفارش‏هاي مشتريان مي‏شود. در روش‌هاي استاندارد برنامه‌ريزي و زمان‌بندي فرض بر آن است كه اجراي برنامه در يك شرايط قطعي و معين اتفاق مي‌افتد؛ اين فرض به‌دليل تنوع منابع مورد ‌نياز و شرايط عدم قطعيت حاكم بر ‌آن، فرض درست و عملي نيست. در چنين شرايطي بررسي استواري روش‌هاي حل به‌جاي بهينه‌سازي از اهميت بالايي برخوردار است. هدف اصلي اين تحقيق حل مسيله‌ي برنامه‌ريزي كارگاهي انسدادي در حالتي است كه زمان پردازش كارها تغيير‌پذير است. در اين تحقيق هشت روش فرا‌ابتكاري براي حل مسيله‌ي مذكور مورد استفاده قرار گرفته است. از معيار ‌«استواري بتا» براي ارزيابي عملكرد روش‌هاي پيشنهادي در شرايط نامعين استفاده‌ شده‌ است. نتيجه‌ي بررسي‌ها نشان‌دهنده‌ي استواري بالاتر روش ژنتيك با جست‌وجوي غير‌همزمان نسبت به ساير روش‌هاست.

Production sequencing and scheduling can be considered as an effective tool in optimally utilizing the resources. In the researches the production environment is normally considered as deterministic. In deterministic environment parameters of problem is predefined and fixed while Such an assumption is not true in a real life manufacturing firms. uncertain events make the processing time of jobs uncertain, which may disturb the completion time of jobs and produce makespan variations. These make span variations cause the jobs to accomplish after the expected completion time. Therefore, the researches on deterministic environment are not practically used and uncertainties need consideration in permutation flowshop scheduling. Therefore it is highly important to take the robustness of the schedules into consideration. In most of the studies, robustness measure used is to minimize the variations of a performance indicator in a schedule. These robust scheduling does not assure that the performance indicator will not exceed a certain limit when uncertain events occur. On the other side in the classical flow shop problem is assumed that intermediate buffers have infinite capacity and a job can be stored for unlimited amount of time. In practice, there are many production factories, in which the buffer capacity has to be taken into account, and there may be limits on the capacity of buffers so we consider blocking flow shop problem with beta Robustness of Make span performance measure.The performance of a number of heuristic and meta heuristic algorithms to solve a blocking flow shop is studied in 9 problems in different sizes. The results reveal the robustness of the proposed algorithms and the most robust algorithm in each size is identified. Generally we can say the Asynchronous genetic local search algorithm give better and more robust sequences with respect to other algorithms. It is followed by proposing potential areas for further researches.