Robust optimization of the continuous annealing process based on a novel Multi-Objective Dragonfly Algorithm. (November 2021)

• Record Type: Journal Article
• Authors: Tian, Huixin; Yuan, Chang; Li, Kun
• Is Part Of: Engineering applications of artificial intelligence. Volume 106(2021)
• Publication Date: 2021-11
• Subjects: Continuous annealing process -- Robust optimization -- Multi-objective dragonfly algorithm -- Adventure Circuitous; Engineering -- Data processing -- Periodicals; Artificial intelligence -- Periodicals; Expert systems (Computer science) -- Periodicals; Ingénierie -- Informatique -- Périodiques; Intelligence artificielle -- Périodiques; Systèmes experts (Informatique) -- Périodiques; Artificial intelligence; Engineering -- Data processing; Expert systems (Computer science); Periodicals; 620.00285
• DOI: 10.1016/j.engappai.2021.104448
• Languages: English
• ISSNs: 0952-1976

Abstract

Abstract: Aiming at the large fluctuations in system and low efficiency caused by disturbance factors in the continuous annealing process, we established a robust multi-objective optimization model. Finding an optimal parameter setting scheme making the strip quality and capacity utilization more stable and robust is the fundamental task. For this problem, we first proposed a new robust index that combines the standard deviation and average value for guiding the model. Second, Multi-objective Dragonfly Algorithm based on Adventure Circuitous (ACMODA) is employed to solve the model. To improve the search efficiency of the algorithm, an Adventure Circuitous strategy is adopted. The strategy enables the individuals to search for new breakthroughs basing on the weight cross-combination hybrid flight pattern when disturbance intense. The experimental results basing on the test functions show that the proposed index can evaluate the robustness precisely and ACMODA can effectively improve the convergence performance comparing with others. Experimental results basing on actual industrial problems show that ACMODA can effectively solve the robust optimization problem in the continuous annealing process.