Search

This paper describes a solution method to the problem of processing n jobs on m non-related parallel machines. It is a linear and combinatorial generalized allocation problem that considered a sequencedependent setup time and dynamic job entry. A genetic algorithm with integer coding and random generation of population, parent selection, crossover and mutation is proposed. There are two descendants per generation that are compared against the worst existing element to enter to population. After a number of generations that is proportional to the product of nxm the solution is generated. The jobs are sequenced on each machine by due date and computational times are acceptable. It is concluded that the proposed genetic algorithm is an effective and efficient solution that focuses on reducing processing time and on meeting deadlines. [ABSTRACT FROM AUTHOR]