基于遗传算法的BP神经网络供应商评价模型研究  

Study on Supplier Assessment Model of BP Neutral Network Based on Genetic Algorithm

作　　者：孟林丽[1] 郑宁[1] 汪沁[1] 

机构地区：[1]上海师范大学天华学院管理学院,上海201815

出　　处：《物流技术》2015年第14期199-201,210,共4页Logistics Technology

摘　　要：利用基于遗传算法的BP神经网络建立了一种供应商评价模型,实现了供应链中的供应商多指标评价,提高了供应商评价过程的客观性,同时克服了BP神经网络具有泛化能力较弱、较多的依赖初始值等缺点,该模型为供应商的评价研究提供了模型和方法的支撑。实证结果表明,该模型具有较强的自组织、自学习和自适应能力,模型评估结果比较客观合理。In this paper, we utilized the BP neutral network of the genetic algorithm to establish a supplier assessment model. With a number of indexes, this model improved the objectivity of the supplier assessment process, overcame the shortcomings of weak generalization ability and initial value dependence and offered models and methods to the supplier assessment. The empirical findings showed the model was able to organize, learn and adapt by itself, which made the assessment results more objective and reasonable.

关 键 词：供应商评价 BP神经网络 遗传算法 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]