摘要
结合四川省金沙江某水电站工程实例,应用BP人工神经网络方法建立3层BP网络模型,选取岩石单轴抗压强度等6个影响因素为输入变量,对坝基复杂岩体进行质量分级。通过机算机Visual C++语言编程实现神经网络模型,进行网络的学习和运算。以神经网络合理结构分析方法选取合理结构,确定合理隐层单元的数量,提高网络测试的精度。对测试结果的分析发现,经过优化的BP网络模型经多次学习后,测试精度提高,结果可靠,取得较好的实际应用效果。
Applying the method of the back propagation artificial neural network of a hydraulic power station on the Jinsha River, Sichuan Province, choosing six influential factors on the rock mass quality as the input variables, such as one axis compressive strength of rock, the complicated rock mass of dam foundation is classified. With Visual C++ program language, network model is realized and can be used to learning and calculating. Through selecting rational network structure and setting the rational number of implicit layers and their units, network and training procedure are optimized, testing precision is also improved. The analysis indicats that the test result is exact and credible. The application has gained a good effect.
出处
《世界地质》
CAS
CSCD
2004年第1期64-68,共5页
World Geology
基金
教育部资助优秀年轻教师基金项目(120413133)
关键词
人工神经网络
岩体质量
网络模型
网络结构
artificial neural network
rockmass quality
network model
network structure
