论文部分内容阅读
以BP反传理论为基础，建立了对Osprey过程的前向多层神经网络，并对其进行测试．利用这一方法研究了Osprey过程中部分参数对孔隙度的影响．结果证明该网络较好地实现了学习和预测．
Based on the back-propagation theory of BP, a forward multilayer neural network for Osprey process is established and tested. Using this method, the influence of some parameters on the porosity in Osprey process was studied. The results show that the network is better to achieve learning and prediction.