摘要
为支持产品方案设计阶段设计信息反馈指导方案优化,提出基于BP神经网络的设计方案敏感性分析方法。建立了绿色信息模型,从方案设计中提取与环境影响相关的绿色信息;建立了神经网络敏感性分析模型,将各个生命周期阶段的绿色信息作为敏感性分析的输入参数,将对应的生命周期评价结果作为输出参数;训练BP神经网络模型使相对误差达到预期精度;分别提取该BP网络输入层与隐含层、隐含层与输出层各单元之间的连接权值,使用Tchaban算法计算得到绿色信息对生命周期评价结果的敏感性系数,即影响贡献值。以牵引电机为例,验证了所提敏感性分析方法,确定了主要设计变量对评价结果的影响趋势,从而对设计方案的变更提出了改进方向。
To give feedback directing to scheme optimization in conceptual design stage by using design information,a sensitivity analysis method based on Back Propagation（BP）neural network was proposed.Green features model was established to extract green information which was related to environmental impact.The sensitivity analysis model of BP neural network was established by taking the green information of each life cycle period as input parameters while the results of its life cycle assessment as output parameters respectively.The BP neural network was trained to reach desired accuracy with the researched data.Tchaban theory was used to calculate the sensitive coefficient between green information and life cycle assessment results using link weights exacted from the input layer to the hide layer and the hide layer to the output layer of BP neural network.A traction motor was taken as an example to verify the proposed method,and the impact trend of design variables on evaluation result was determined to provide guidance for design scheme modification.
出处
《计算机集成制造系统》
EI
CSCD
北大核心
2016年第3期666-671,共6页
Computer Integrated Manufacturing Systems
基金
国家自然科学基金资助项目(51175312)
国家863计划资助项目(2014AA041503)~~
关键词
生命周期评价
绿色特征
BP神经网络
敏感性分析
方案设计
life cycle assessment
green features
BP neural network
sensitivity analysis
conceptual design
