Robotic assembly line becomes popular in recent years for its ability to outperform labor assembly lines in terms of efficiency and flexibility. This paper focuses on the mixed-model U-shaped robotic assembly line balancing and sequencing problem (MURALBSP). Different from most reported works, this paper introduces the energy consideration into the problem. Two conflict objectives, i.e., energy consumption and makespan, are studied for realizing energy efficiency manufacturing and lean production. To this end, a hybrid multi-objective dragonfly algorithm (HMODA) is proposed. First, the mathematical model of this bi-objective problem is formulated. Second, the basic dragonfly algorithm is modified and improved to solve the problem. The specific encoding and decoding method is designed and the chaotic map is used to improve algorithm randomness. Besides, the solution update method of basic dragonfly algorithm (DA) is amended and multi-point crossover mechanism is employed. Finally, several multiple size benchmark problems are designed and comparisons are conducted, the sensitivity of decision variables is analyzed to provide managerial insights. The results demonstrate that HMODA is more efficient in solving the energy oriented MURALBSP than five other well-known algorithms.

近年来，机器人组装线因其在效率和灵活性方面优于劳动力组装线的能力而变得流行。本文重点研究混合模型U形机器人装配线平衡和排序问题（MURALBSP）。与大多数报道的作品不同，本文将能源考虑因素引入了问题中。为了实现节能制造和精益生产，研究了两个冲突目标，即能耗和制造期。为此，提出了一种混合多目标蜻蜓算法（HMODA）。首先，建立了这个双目标问题的数学模型。其次，对基本的蜻蜓算法进行了修改和改进以解决该问题。设计了特定的编码和解码方法，并使用混沌映射来提高算法的随机性。此外，修正了基本蜻蜓算法（DA）的解法更新方法，并采用了多点交叉机制。最后，设计了多个尺寸基准问题，并进行了比较，分析了决策变量的敏感性以提供管理洞察力。结果表明，与其他五种众所周知的算法相比，HMODA在解决面向能量的MURALBSP方面效率更高。