摘要
介绍了神经网络中的BP网络的模型及其学习算法原理 ,并将其应用于活性粉末混凝土强度预测。基于MATLAB神经网络工具箱 ,进行了结果分析 ,发现应用该网络在活性粉末混凝土 (RPC)强度预测方面具有很高的精度 ,可以用来对高强度混凝土强度进行预测 ;结果表明 ,神经网络方法是一种可以定量分析、简便易行的预测方法 ,随着新型建筑材料科学领域各种实验数据不断丰富完善 ,计算机语言的发展 ,神经网络可为广大技术人员提供科学的理论分析方法 ,指导生产实践。
The model and learning algorithms of BP(Back Propagation)are applied to strength forecast of reactive powder concrete(RPC).Through the analyses based on Matlab-NNT,it is found that BP network has high accuracy in strength forecast of RPC, and can be used in the forecast of high performance concrete. The results show that BP neural network is a convenient approach for the quantitative forecast. With the development of the new construction material and computer language, the neutral network will provide scientific analysis method for the technicians and also supervise the practice of production.
出处
《河南科技大学学报（自然科学版）》
CAS
2004年第2期75-77,98,共4页
Journal of Henan University of Science And Technology:Natural Science
