A novel teaching-learning-based optimization algorithm for energy-efficient scheduling in hybrid flow shop

Abstract

Hybrid flow shop scheduling problem (HFSP) has been extensively discussed and the main objectives are related to completion time. The reduction of energy consumption should be considered fully in HFSP in the era of green manufacturing. In this study, biobjective energy-efficient HFSP is considered, which is made up of three subproblems including scheduling, machine assignment, and speed selection. A three-string coding method is used to indicate solutions of three subproblems. A new teachers’ teaching-learning-based optimization (TTLBO) is proposed to minimize total energy consumption and total tardiness. Total tardiness is regarded as a key objective and a lexicographical method is adopted to compare solutions. TTLBO generates new solutions using a new optimization mechanism and is made up of the self-learning, interactive learning, and teaching of teachers. The learning phase of students are deleted from the algorithm. Multiple neighborhood searches are used to implement the self-learning of teachers and global search based on crossover is chosen to imitate other tivities of teachers. A number of experiments are conducted to test the impact of the new optimization meachanism on the performance of TTLBO and compare TTLBO with other algorithms from the literature. The computational results show that TTLBO is a competitive algorithm for the considered HFSP.