دانلود گزارش کارورزی و کار آموزی

چکیده :

A mathematical programming model is proposed for the two parallel machines scheduling

problem where one machine is periodically unavailable, jobs are non-preemptive, and the

objective is minimizing the makespan. The model is established by transforming the two

parallel machine setting into a single machine setting. Average-case analyses of the classical

Longest Processing Time first (LPT) algorithm and the List Scheduling (LS) are presented.

Computational experiments show that the LPT algorithm beats the LS algorithm in all the

96 combinations of two main parameters from an average-case error point of view and that

the average-case error of the LPT algorithm is less than 2% when the number of jobs is

greater than twenty. Unexpectedly, there also exist instances showing that the LS algorithm

may beat the LPT algorithm from the average-case error point of view.