A hybrid genetic algorithm for the finite horizon economic lot and delivery scheduling in supply chains

In this paper, we investigate the lot and delivery scheduling problem in a simple supply chain where a single supplier produces multiple components on a flexible flow line and delivers them directly to an assembly facility. It is assumed that all of parameters such as demand rates for the components are deterministic and constant over a finite planning horizon. The main objective is to find a lot and delivery schedule that would minimize the average of holding, setup, and transportation costs per unit time for the supply chain. We develop a new mixed integer nonlinear program and an optimal enumeration method to solve the problem. Due to difficulty of obtaining the optimal solution in medium and large-scale problems, a hybrid genetic algorithm (HGA) is also developed. The proposed HGA incorporates a neighborhood search into a basic genetic algorithm that enables the algorithm to perform genetic search over the subspace of local optima. The two proposed solution methods are compared on randomly generated problems, and computational results show that the performance of HGA is very promising because it is able to find an optimal or near-optimal solution for majority of the test problems.