摘要
本文通过采用并加以改进的误差反向传播人工神经网络(BPN)算法,使网络收剑速度加快和避免局部极小。在此基础上建立了岩溶水模拟模型,同时以济南为例对模型进行了应用验证。拟合结果表明,该方法对于岩溶水资源评价具有较大的适用性。
A new neural network based model for karst ground water resources is presented. In the model, the backpropagation algorithm is used to train the network, and the conjugate gradient method is used to speed up the convergence and to improve performance. To demonstrate the procedures and performance of the model, the case of Jinan area is selected for analysis and discussion. The results indicate that the method is feasible.
出处
《中国岩溶》
CAS
CSCD
1999年第4期337-341,共5页
Carsologica Sinica
基金
国家自然科学基金!项目批准号:49772162
