摘要
利用模糊神经网络直接从实验数据中提取规则 ,进行材料性能建模与预测 ,作为应用示例 ,建立了基本成分和组织参数的灰铁预测模型。与多元统计分析、模糊回归和广义回归网络所得的结果相比 ,该方法所得的模型具有学习精度高 ,且具有更好的泛化能力。
A fuzzy neural metwork is developed to extaract fuzzy rules directly from experimental data for material property modeling.As an application example,a model based on compositions and microstructures is developed to predict strength of gray iron.Comparing with the results obtained by multiple statistic analysis,fuzzy regression and generalized regerssion neural network,the fuzzy neural network show good learning precision and generalization.
出处
《重庆工学院学报》
2001年第5期5-10,共6页
Journal of Chongqing Institute of Technology
基金
This work was supported by the Science & technology Funds of CAEP under Grand No .2 0 0 0 0 32 9 and No .2 0 0 10 6 6 8
