Indian Journal of Fundamental and Applied Life Sciences ISSN: 2231-6345 (Online) An Online International Journal Available at http://www.cibtech.org/jls.htm 2012 Vol. 2 (1) January-March, pp.312-316/Akhshabi et al. Research Article 312 GENETIC ALGORITHM AND ITS APPLICATION IN INDUSTRIAL MACHINERY SCHEDULING WITH FUZZY DUE DATE * Mohammad Akhshabi 1 , Mostafa Akhshabi 2 and Javad Khalatbari 3 1 Department of Industrial Engineering, Ramsar Branch, Islamic Azad University, Ramsar, Iran 2 Department of Computer Engineering, Ramsar Branch, Islamic Azad University, Ramsar, Iran 3 Department of Management, Ramsar Branch, Islamic Azad University, Ramsar, Iran ABSTRACT Appropriate programming in a gorge of plant or the key and strategic machine has much effect to increasing efficiency. Due to increasing global market competitiveness the regarded targets have become complex. Thus one criterion is not enough and scheduling with multiple criteria is more realistic. The main difficulty of these scheduling problems is extensive solving time needed to it. This paper proposed a new fuzzy scheduling model for single machine scheduling problem and aims to improve it to a real-world application through fuzzy set theory. For this purpose, due dates of jobs are defined as fuzzy numbers. Key Words: Genetic Algorithms, Single Machine Scheduling, Fuzzy Sets, Due Dates INTRODUCTION In just-in-time (JIT) environment, each job should be completed as close as possible to its due date. It is involved producing goods only when necessary. Owning to the wide adoption of this philosophy in recent decades, scheduling problems for meeting the due date requirement have been investigated extensively, including those with general earliness-tardiness penalties about a common due date. Missing a Job’s due date may result in loss of customer or the need to compensate for the delay along the production or assembly line. On the other hand, producing a job much earlier than its due date may cause unwanted inventory and/or deterioration of the product. in the modern competitive environment the cost of tardy deliveries, such as a company’s goodwill, future sale and rush shipping cost, and the cost of early include holding cost for finished goods, deterioration of perishable goods and opportunity cost will significantly decrease a company profits. Therefore minimizing total weighted completion time, tardiness and earliness is not only a measure of academic interest but also useful and important in practice. Indeed flow shop, flexible flow shop, job shop and open shop scheduling problem are often addressed decomposing the original planning process into many sub problem that can be solved by using single machine techniques. Genetic algorithm (GA) is a powerful search technique based on natural biological assessment which is applied for finding an optimal or near-optimal solution. The idea of GA was first suggested by Holland Holland (1975) in the early 1970s and since then has been widely used in solving optimization problem. In contrast to other optimization methods, GA functions by generating a large set of possible solutions to a given problem instead of working on a single solution. The technique such as the process of selection, crossover, mutation and evaluation has been implemented successfully in many scheduling problems, in particular job shop scheduling. Job shop scheduling problem (JSSP) is a difficult NP-hard combinatorial optimization problem. Earlier work on solving JSSP centered on exact algorithms such as branch-and- bound approach Applegate and Cook (1991). However, the work focused on small sized instances which can be solved in a reasonable computation time. As the problems become more complex, the research focused various other techniques such as simulated annealing and genetic algorithms Akhshabi et al., (2011). Some other papers considered the corresponding identical parallel machine scheduling problem in which the machines have the same speed. The literature in recent years mainly focused on the problem with unequal release dates, i.e. Pm|rj|Lmax problem. Carlier et al., (1998) and Néron et al., (2008) developed