Rolling schedule design for the ESP rolling process based on NSGA-II-DE.

Area of Science:

• Materials Science
• Manufacturing Engineering
• Operations Research

Background:

• Endless strip production (ESP) involves interconnected processes, making single-process optimization insufficient for global solutions.
• Coupled parameters in ESP rolling require integrated approaches for effective optimization.
• Existing methods struggle to achieve optimal outcomes due to the complexity of multi-objective modeling in ESP.

Purpose of the Study:

• To develop a multi-objective optimization model for ESP rolling schedules.
• To improve product quality, minimize power consumption, and ensure loading balance.
• To provide a practical method for selecting optimal solutions from Pareto sets.

Main Methods:

• Proposed a multi-objective optimization model integrating power consumption, product quality, and loading balance.
• Utilized thickness and heating temperature as decision variables to couple temperature and loading.
• Applied a non-dominated sorting genetic algorithm-II based on differential evolution (NSGA-II-DE) to find Pareto solutions.
• Designed a satisfaction function to select the best solution from the Pareto front.

Main Results:

• The NSGA-II-DE method successfully generated Pareto optimal solutions for ESP rolling schedules.
• The satisfaction function effectively selected a superior solution from the Pareto set.
• Compared to online scheduling and standard NSGA-II, the selected solution demonstrated improved product quality and loading balance.

Conclusions:

• The proposed integrated optimization approach effectively addresses the complexities of ESP rolling.
• The developed method provides a precise and efficient way to optimize ESP rolling schedules.
• The findings offer valuable guidance for enhancing real-world industrial production processes.