A Knowledge-Based Multiobjective Memetic Algorithm for Green Job Shop Scheduling With Variable Machining Speeds

Nowadays, green manufacturing has become one of the hot topics in both academia and industry because of global warming and the greenhouse effect. Among manufacturing systems, job-shop scheduling plays a key role due to its wide applications. In this article, we study a green job-shop scheduling problem with variable machining speeds (JSPVMS) aiming at minimizing the makespan as well as the total energy consumption (TEC). First, a new mixed-integer linear programming model is formulated for this green JSPVMS. Then, a new knowledge-based multiobjective memetic algorithm (MOMA) is developed to address this problem. In our MOMA, a novel decoding scheme based on the problem property is well designed to obtain better tradeoff solutions between makespan and TEC. Furthermore, a novel local search is proposed to search for promising nondominated solutions by discovering problem-specific knowledge. Additionally, the proposed MOMA utilizes the advantage of the genetic operator and local search to balance the exploration and exploitation. The effectiveness of each improvement component (decoding scheme and local search) in our MOMA is verified by comparing experiments among different MOMA variants. Finally, we compare our MOMA with several well-known multiobjective optimization algorithms (i.e., NSGA-II, SPEA2, and MOEA/D) on JSPVMS instances. Experimental results indicate that our MOMA significantly outperforms the other algorithms on most of the instances.

Automated summary

This article studies a green job-shop scheduling problem with variable machining speeds, aiming at minimizing the makespan as well as the total energy consumption. A new mixed-integer linear programming model and a knowledge-based multiobjective memetic algorithm are proposed to address this problem, and experimental results show that the proposed algorithm significantly outperforms other algorithms on most instances.