摘要
“94·6”华南大水，珠江三角洲网河的顺德、番禹洪水水位异高，引发对同河及近岸环境变异的探讨。近年来，河口同河区的资源开发、环境变异引起各界的关注。河口网河区河床变化，水文环境发生变异。利用人工神经网络（ANN），对网河水位过程进行非线性模拟，输入径流与潮流资料，以历史长序列与近年实测资料训练模型和检验，认为ANN较适应于同河高度非线性水文模拟，也较适应于环境受到人类活动冲击的河口水文模拟；在河口区，洪峰水位与网河洪（潮）水位的一无线性相关，在理论和应用上都遇到挑战。
During' 94.6' floood in Anth China, anomalously high water levels occurred in the river network region near the Pear Estuary . Ih meen po, attenhon has bo paid to the environment changes with relations tothe mpid economic developwnt of the Pearl River Delta. By using Artificial Neural Networks (ANN), aSeries of nonlinear correation analyses of water level fluctuahon in the river netw are mad with the data ofnmof and hde. The edts show that ANN edl is a highly nonlinear boc sysed whch can efficienilsimulate the nonlinear pmis of the hydrologic variables in river networ or estUallne enviaret. It isconsidend that the hyedopc vallables in river netwh should be of estump hydxoboc properties, and is anonlinear dynamic system. The linear Wion applicating to analysis of continental hydIDboc vedablesmight not be used to the compotm hydrologic frequecy analysis of river network.
出处
《热带地理》
北大核心
1998年第2期162-167,共6页
Tropical Geography
