基于BP神经网络的ETO产品配置设计方法

按订单设计(engineering-to-order, ETO)的定制产品因产品族结构比较复杂，产品间结构差异较大，设计过程涉及个人经验和灵感，并大量应用人机交互处理，难以实现设计自动化、程序化。人工神经网络模仿人脑结构及智能行为，具有大规模并行处理、容错、自组织和自适应能力及联想功能，符合ETO配置设计的特点。通过对ETO定制产品需求的分析，构建并训练具有一定结构和功能的BP神经网络，训练好的网络蕴含着ETO配置设计规则和经验。实例证明了该方法的可行性。

Configuration design based on BP neural network for ETO product

Customized product of ETO(engineering-to-order) is hard to program and realize to automation due to the complex structure of its product family,big difference between product structures,involvement of personal experience and inspiration during design process,and a large amount of interaction between human and machine.Artificial neural network can simulate the structure of human brain and intelligent behavior,and have the ability of parallel processing,redundancy,self-organization,self-adaptation and association,which accords with the features of configuration design.Through analysis of ETO customized product requirements,a BP artificial neural network with some structure and function is established and trained.The trained network contains rules and experiences of ETO configuration design.Illustrations have demonstrated its feasibility.