摘要
根据建筑物实测沉降利用人工神经网络理论，结合膨胀顶升法纠偏加固理论，建立了前馈网络模型并提出新的算法，结合某建筑物纠偏工程实例对建筑物纠偏加固所需的生石灰量进行了预测。预测结果表明神经网络方法是可行且有效的。
According to the field survey of building settlement and the rectifying methods of expanding method, the artificial neural network theory and method is adopted to predict the quantity of quicklime, build the feed forward network forecasting model and presented the new learning algorithm. It can be seen from the forecasting result on the example that the ANN prediction method is feasible and effective.
出处
《建筑技术开发》
2006年第7期29-30,33,共3页
Building Technology Development
基金
甘肃省科技攻关资助项目（编号：ZS011-A50-013-G）
关键词
建筑物沉降
生石灰桩
预测
人工神经网络
Building settlement
Quicklime pile
Prediction
Artificial neural network
