Abstract: To solve the human-robot collaborative assembly line balancing problem with limited number of cobots,a mixed integer linear programming model with the optimization objective of minimizing the cycle time was constructed,and the correctness of the model was verified.In view of the characteristics of complex task allocation,multiple assembly modes and limited robot resources,a hybrid genetic simulation annealing algorithm was proposed.Considering the scheduling relationship between workers and robots in the station,a three-stage encoding and decoding strategy based on task allocation,task mode and resource constraints was constructed to improve the quality of the initial solutions,a mapping crossover operator and an insertion mutation operator based on precedence constraints were designed to enhance the global search ability of the proposed algorithm,and eight new solution generation methods combining task,assembly mode and robot allocation were introduced to improve the local search ability of the proposed algorithm.Through 32 test cases and compared with four classical algorithms,the effectiveness and superiority of the proposed algorithm were verified.In the resource-constrained cases of different scales,the cycle time of the human-robot collaborative assembly mode was shortened by 17.83%,15.00% and 14.71% respectively compared with the manual assembly mode,which improves the assembly efficiency.

Key words: assembly line balancing, human-robot collaboration, mixed integer linear programming, genetic algorithms, simulated annealing algorithm

摘要： 针对协作机器人数受限的人机协作装配线平衡问题,构建了以最小化节拍为优化目标的混合整数线性规划模型,并验证了模型的正确性。针对问题任务分配复杂、装配模式多、机器人资源受限等特征,提出一种混合遗传模拟退火算法。考虑工位内工人和机器人的调度关系,构造了基于任务分配、任务模式和资源约束的三段式编解码策略,以提高初始解的质量;设计了一种基于优先约束的映射算子和插入变异算子,以增强算法的全局搜索能力;引入模拟退火操作,并设计了任务、装配模式与机器人分配组合的8种新解产生方式,来提高算法的局部搜索能力。通过32个测试案例并与4种经典算法对比,验证了所提算法的有效性和优越性。在资源受限的不同规模案例上,相比手工装配模式,人机协作装配模式的节拍分别缩短了17.83%,15.00%,14.71%,有效提高了装配效率。

关键词: 装配线平衡, 人机协作, 混合整数线性规划, 遗传算法, 模拟退火算法