摘要
以B-P网络为例介绍了人工神经网络模型的应用原理。以开拓人工神经网络在环境科学中的应用范围为目的,综述了几种常用神经网络模型在环境预测与评价2方面的应用性能。结果表明:人与环境关系的科学量度是多方位、多因素的非线性模糊问题,而人工神经网络模型具有解决该类问题的独特优点。针对不同环境问题,优选模型,开拓人工神经网络方法在环境科学中的应用范围,为环境问题的解决提供新的技术手段。
This article introduced the applied theory of artificial neural network (ANN) by the example of Error Back-Propagation network, in order to develop the applied field of ANN in environmental science, it summarized the applied function of several ANN models in common use of environmental prediction and assessment using synthesis method of comparison and assessment. The scientific measurement of relations between human being and environment is a nonlinear fuzzy problem in numerous positions and multiple elements, while ANN models possess peculiarities to solve the problem. Optimum seeking the model with different environmental problem, developing the applied fields of ANN in environmental science may offer new technological method to solve environmental problem.
出处
《上海环境科学》
CAS
CSCD
1999年第11期510-512,共3页
Shanghai Environmental Sciences
