摘要
西南岩溶地区地下河系统的多层次和多级性特征,决定了其输入因子与响应因子之间为非线性关系,传统的统计方法在揭示此类关系时效果欠佳,而人工神经网络模型(A rtific ia l N eura l N etw ork——ANN)正好弥补了此项不足,其在原理和构模上均表现出与岩溶地下河系统十分相似的特点。通过对广西地苏地下河系统水量数据的重塑发现,ANN模型重塑的效果明显优于传统的回归分析法,证明了运用ANN模型重塑岩溶地下河系统流量数据是完全可行的。
The nonlinear relationship between input and output factors is determined by the multi-layered and multi-order characteristics of groundwater system in Southwest China karsts area. Artificial Neural Network -ANN Which is similar to groundwater in principle and model establishment, fills up a gap in traditional statistics that cannot express the relatiortship satisfactorily. Results of ANN is better than traditional regression analysis method through remolding the groundwater data in Disu underground river system, Guangxi province. Therefore, using ANN model to remold the flux data of groundwater system is feasible.
出处
《中国岩溶》
CAS
CSCD
北大核心
2006年第2期121-125,共5页
Carsologica Sinica
基金
国家地调项目"西南岩溶地区地下水与环境地质调查综合研究"(编号:200310400043)
关键词
人工神经网络
地下河系统
观测数据
重塑
Artificial neural network
Underground river system
Observed data
Rebuilding
