求解双边装配线第Ⅰ类平衡问题的改进离散蝙蝠算法

Study on Effect of Lubrication and Cooling Parameters on Temperature Rise of Electromagnetic Self-balanced Motorized Spindle

在分析双边装配线第Ⅰ类平衡问题(Two-sided Assembly Line Balancing Problem of Type-Ⅰ,TALBP-Ⅰ)离散性、序列相关性等特点后,提出了一种改进离散蝙蝠算法(Improved Discrete Bat Algorithm,IDBA)。为在总工位数相同情况下筛选出更优质的解,增加了启发式目标,引导种群向更优方向搜索。标准蝙蝠算法不能直接求解离散问题,针对TALBP-Ⅰ,设计了基于任务拓扑排序矩阵的编码策略,利用双重编码映射机制,实现蝙蝠飞行的连续物理空间到TALBP-Ⅰ离散解空间的映射。采用改进的"工位-操作"解码方法代替传统的"操作-工位"解码方法,减少工位的空闲时间。针对蝙蝠算法后期收敛速度慢,易陷入局部最优,设计了4种插入邻域算子,进行变邻域搜索。通过基准问题的数值实验验证了算法的有效性。

After analyzing the discrepancy and sequence related of two-sided assembly line balancing problem of type-I(TALBP-I),an improved discrete bat algorithm(IDBA)was proposed.In order to filter out better solutions under the same total number of workstations,a new heuristic objective was added to guide the population to search in a better direction.The standard bat algorithm can’t solve the discrete problem directly.For TALBP-I,the coding method based on task Topology sorting matrix and the dual coding mapping mechanism was designed,which mapped the continuous flight physical space of bats to the discrete solutions space of TALBP-I.The improved"station-operation"decoding method was used to replace the traditional"operation-station"decoding method to reduce station idle time.To avoid optimization process trapped in local optimum at later stage,IDBA implemented variable neighborhood search by four neighborhood structures proposed.It measured the proposed algorithm by the different scale benchmark problems with comparison of several algorithms.Simulation results show that DBA is efficient for solving TALBP-I.