Using Guided Initial Chromosome of Genetic Algorithm for Scheduling Production-Distribution System

Abstract

Production and distribution system in a company should be managed carefully. Delay in product delivery not only results in a late penalty due to customer dissatisfaction or breach of contract, but also causes a supply chain failure. Of course, all these impacts will also reduce the reputation of a company. Scheduling integrated production-distribution is classified as NP-Hard problem. Genetic algorithm can be used to solve complex problem. In this paper, genetic algorithm is used for scheduling production-distribution in make to order system where each job has a different deadline and volume (size). This problem is represented on mixed integer programing model. We verify the genetic algorithm’s performance by comparing the results with the total cost calculated by lower bounds of the problems. Experiments show that the traditional initial random cannot produce good result with more than 15 job size problem. We proposed guided initial chromosome to tackle this problem. From further experiments shows that the proposed method approach can increase the performances of genetic algorithm in more than 15 job size problem. In general, proposed genetic algorithm with guided initial chromosome shows better solution quality and better time efficiency compared to previous related research.