面向维修的复杂装备模块智能聚类与优化求解技术

摘要

To overcome the shortage that single maintenance department participated in existing maintenance activi- ties, the driven factors and strategy selection of maintenance stage were introduced in design stage of complex equip- ment, and modular design method of complex equipment oriented to maintenance was proposed. From the perspec- tives of maintenance cost, maintenance complexity and maintenance efficiency, the criteria for modular design were discussed, and specific calculation methods of maintenance properties were obtained. The modular design optimiza- tion model oriented to maintenance was constructed by considering constraint conditions. The model was optimized and solved by using hybrid multi-objective shuffled frog leaping algorithm which combined shuffled frog leaping al- gorithm with bacteria optimization, therefore a series of Pareto optimal solution which represented modular design schemes were obtained. The final modular design scheme was gained by using information entropy theory based Pareto optimized method. GMC precise 5-axis machining center in Shenyang machine tool factory was taken as an example to verify the effectiveness and feasibility of proposed method.%为解决传统维修过程中由单独的维修部门被动应对既成事实的复杂装备检修问题，同时加强其他部门对于维修活动的协同能力，在复杂装备的设计阶段引入维修相关的驱动要素和维修阶段的策略选择，提出一种面向维修的复杂装备模块化设计方法。从维修成本、维修复杂度、维修效率等方面探讨复杂装备模块化设计准则，得到各维修特性的量化计算方法，通过综合考虑约束条件建立面向维修的模块化设计模型。采用青蛙跳跃算法和细菌优化相结合的混合多目标蛙跳算法对模型进行优化求解，从而得到一系列代表模块化设计方案的Pareto最优解，并利用基于信息熵理论的Pareto优选方法获取最终的模块化设计方案。以沈阳某机床厂设计生产的GMC型精密五轴加工中心为例，运用数值仿真手段验证了该方法的有效性和可行性。