METHODOLOGY BASED ALGORITHMS VEGA AND MOGA TO SOLVE A MULTIOBJECTIVE PROBLEM IN A JOB SHOP PRODUCTION SYSTEM

This paper presents a methodology that aims to minimize simultaneously, in a "Job Shop" production system the following variables: process time (makespan time), cost of direct labor and also the fraction defective generated by operator fatigue. For this purpose, are taken and fused elements of genetic algorithms Vega and Moga, through the following steps: generating the initial population, form the new population, obtaining the appropriate analysis of variance and finally compared with a hybrid method of weighted sums and genetic algorithms. According to the above, when evaluating the solution faster processing time corresponding to the method based on algorithms Vega and Moga, respect to the solution faster processing time calculated from the method based on weighted sums and genetic algorithms, states that the first one exceeds the second performance as: for process time variable (in hours) at 27.86%, for variable in process time (in weeks) at 1.25%, in terms of the variable cost of direct labor in 6.73% and, as to the variable defective fraction in 25.85%.