Solving the Single Machine Total Weighted Tardiness Problem Using Bat-Inspired Algorithm

This paper aims to present an application of a recently proposed metaheuristic approach, namely bat algorithm (BA), for solving the single machine with total weighted tardiness (SMTWT) scheduling problem. In this paper, a guided population and two-swap local search are introduced to integrate with the BA. Four variants of BA methodology, including classical BA, BA with two-swap, BA with guided population, and BA with guided population and two-swap, have been developed. All BA variants are executed on the selected benchmark instances for the SMTWT problem taken from OR-library and compared search performances on deviation of solution obtained from best-known solution and computation time. The results show that two proposed BAs with a guided population scheme achieve a good performance. The BA with guided population and two-swap techniques outperforms among four proposed designs on solution quality, and it can achieve a trade-off between solution quality and execution time.