Dynamic multi-objective optimization for uncertain order insertion green shop production scheduling

Abstract

Efficient production scheduling of forgings is critical to manufacturing. However, production orders in manufacturing can change dynamically, making it challenging to quickly track changes between orders and ensure productivity and environmentally friendly production in the shop. This paper presents a dynamic multi-objective green shop production scheduling optimization problem that addresses uncertain order insertion, considering the objectives of total completion time, energy consumption, and carbon emission. When a new batch of orders arrives, changes in the dimensionality of decision variables for production scheduling lead to environmental alterations. By integrating the workpiece completion rates of orders from historical environments with the workpiece quantities of orders in the new environment, the scheduling plan is dynamically adjusted and promptly responded to. Therefore, we design a discrete matrix-based dynamic multi-objective optimization algorithm (DM-DMOEA), which can measure the similarity of the orders before and after the dynamic changes, and reconstruct a high-quality scheduling solution under the new environment, which solves the problem of variable dimensionality changes due to the dynamic environment. Finally, experiments were conducted in a real case of a flange manufacturing company, and the results proved the validity of the proposed model and the superior performance of the algorithm.