A genetic algorithm for minimizing total tardiness/earliness of weighted jobs in a batched delivery system

This paper endeavors to solve a novel complex single-machine scheduling problem using two different approaches. One approach exploits mathematical modeling, and the other is based upon genetic algorithms. The problem involves earliness, tardiness, and inventory costs and considers a batched delivery system. The same conditions might apply to some real supply chains, in which delivery of products is conducted in a batched form and with some costs. In such delivery systems, the act of buffering the products can have both positive effects (i.e., decreasing the delivery costs and early jobs) and negative ones (i.e., increasing the number of tardy and holding costs). Accordingly, the proposed solution takes into account both effects and tries to find a trade-off between them to hold the total costs low. The suggestions are compared to existing solutions for older non-batched systems and have illustrated outperformance.

► We consider batched delivery system and its advantages in single machine scheduling.
► We conclude that integer programming methods are not able to solve this new problem.
► A novel genetic algorithm is provided for this new problem.
► The proposed method performs very well for different settings and sizes of problems.