摘要
介绍了BP神经网络的基本概念与结构,提出了计算和预测混凝土碳化深度的神经网络模型。建立1-3-1单因子(时间)输入向量网络与传统回归分析方法进行比较;建立6-4-1多因子输入向量网络计算及预测混凝土碳化深度。分析结果表明该模型计算和预测精度都能达到工程要求,适合在工程中应用。
Introduced the primary concepts and components of BP neural network. A model based on BP neural network is preented to calculate and forceast carbonation depth. Build 1 - 3 - 1 network with single inputs（time） to calculaste carbonation depth, compared with th（ result of traditional recur,sire analysis; build 6 - 4 - 1 network with multiple inputs to ealeulate and forecast earbonation depth. It is concluded through analysis that the BP network model has high precision in calculation and forecast and the model is suitable for practical use.
出处
《计算机技术与发展》
2006年第11期231-233,共3页
Computer Technology and Development
关键词
神经网络
BP算法
混凝土
碳化
neural network
BP algorithm
conerete
carbonation
