人工神经网络在岩土参数反分析中的应用

Application of artificial neural network in back-analyzing soil parameter

在运用人工神经网络对深基坑岩土参数进行反分析的基础上,将PSO与BP算法相结合,充分发挥PSO全局寻优的能力和BP算法局部细致搜索优势,并通过实例验证了方法的可行性。可以看出,运用该方法可以使学习效率增高,收敛速度加快,预测结果更加准确。

Based on the soil parameter back-analysis using artificial neural network, the method which combined PSO with BP and took full use of the global optimization of PSO was discussied. It is obvious that the new way increase the efficiency and the velocity of converging, also the prediction was more accurate.