Abstract: Aiming at the complexity of rational balance between product family's commonality and difference without transcendental information, a multi-objective constraints optimization model of scale-based product family with commonality maximization and average performance lose minimization was proposed under multi-platform design of scale-based product family. Based on the ideas of non-dominated sort and crowding distance sort, the compare operation and the update operation of traditional firefly brightness were improved, and a modified firefly algorithm adapt to solve multi-objective constraint optimization problem was presented. To solve the optimization model of scale-based product family, a expression pattern of two-layer artificial firefly structure with common variable ability was constructed. A simulation study on optimization design of motor product family was provided to evaluate the effectiveness of proposed model and algorithm.

Key words: scale-based product family, multi-objective optimization, firefly algorithms, mass customization, optimization design

摘要： 在多平台可调节产品族优化设计下，针对无先验信息时产品族共性与差异性合理平衡的复杂性，提出以共性最大及平均性能损失最小的可调节产品族多目标约束优化模型。借鉴"非支配排序+拥挤度排序"思想，对传统的萤火虫亮度比较操作及亮度更新操作方法进行改进，提出一种适用于求解多目标约束优化模型的改进型萤火虫算法；构造具有共性可变能力的双层结构人工萤火虫个体表达模式，用于求解可调节产品族优化模型。通用电动机产品族优化设计案例的仿真实验，验证了所提模型和算法的有效性。

关键词: 可调节产品族, 多目标优化, 萤火虫算法, 大规模定制