摘要
采用BP神经网络预测技术建立少于3层的国毛毛条加工质量预测模型,利用贝叶斯判别规则提高网络的学习能力,通过实验对比选出最优隐层神经元数。模型预测结果表明,用神经网络方法预测毛条加工性能与实际结果有相当高的一致性。
This paper adopts BP neural network predicting technology to set up the model on predicting domestic-top quality in processing,which is less than three layers.It uses Bayesian regularization to improve network's study ability.By contrast in test,the proper number of latent nerve cell is selected.The result of model shows that the method of using neural network to predict top's processing performance is consistent with the real result.
出处
《毛纺科技》
CAS
北大核心
2004年第12期9-13,共5页
Wool Textile Journal
基金
国家科技创新项目"毛条制造与纺纱加工过程的质量预测与控制"
2003 3~2003 6
国家经贸委。