面向个性化定制需求的异构制造资源服务组合优化方法

Optimization Method for Heterogeneous Manufacturing Resource Service Composition Catering to Personalized Customization Demands

工业制造任务日趋复杂,组件服务组合优化问题相关的指标日益增多,需要综合考虑各个评价指标,从备选服务中筛选出最优服务组合。本文针对工业制造的特点,从服务成本、服务时间等服务质量(QoS)指标构建了组件服务评价指标体系。为了处理高维多目标优化问题并针对非支配排序遗传算法(NSGA-Ⅱ)只能求得最优解集的特点,本文提出改进ε-约束策略融合Pareto支配改进NSGA-Ⅱ算法,并将NSGA-Ⅱ和模糊决策相结合,利用模糊决策从最优解集中寻找最优解。

With the increasingly complex tasks in industrial manufacturing, the metrics related to the optimization of component service combinations are growing. It’s essential to comprehensively consider various evaluation metrics to filter out the optimal service combinations from alternative services. This paper, targeting the characteristics of industrial manufacturing, constructs a framework of evaluation metrics for component services based on QoS indicators such as service cost and service time. To address high-dimensional multi-objective optimization problems and considering the lim-itation of the Non-dominated Sorting Genetic Algorithm II (NSGA-II) in obtaining only the optimal solution set, this paper proposes an improved ε-constraint strategy integrated with the Pareto dominance-enhanced NSGA-II algorithm. Additionally, it combines NSGA-II with fuzzy deci-sion-making to search for the optimal solution within the obtained optimal solution set using fuzzy decision-making methods.