Implementing a Hybrid Push-Pull System Using Genetic Algorithms

Abstract

This paper proposes a Hybrid Push-Pull production system, which optimizes the value of the Inventory and Not Demand Satisfaction. The hybrid production system proposed integrates characteristics of a Pull system into a Push system. The optimization technique applied was the Nondominated Sorting Genetic Algorithm using the Simulated Binary Crossover with a probability of 90%, and Parameter-based Mutation with probability of 17%. To validate the proposal, simulations were performed with production capacities ranging from 5.000 to 500.000 products per month, with random demands. The system performance was compared against a Push System, due to the nature of its construction. The results showed that the proposed production system maintains a balance between the Inventory and Not Demand Satisfaction.