偏最小二乘法与神经网络耦合的大坝监测模型

Dam Monitoring Model Based on Coupling of Partial Least Square Regression and Artificial Neural Network

针对人工神经网络在大坝变形监测模型应用中所出现的收敛慢和稳定性差等问题,提出了偏最小二乘法与人工神经网络耦合的大坝变形监测模型,提高了神经网络的学习速率和稳定性。首先运用偏最小二乘法对多维自变量进行主成分提取和降维处理,解决了变量之间的多重相关问题,而后把降维的数据输入神经网络进行训练。对比实例应用结果表明,偏最小二乘神经网络耦合模型的拟合速度和精度都高于传统的神经网络。

Because of the bad convergence and poor stability of artificial neural network, a new model based on coupling of partial least square re-gression and artificial neural network was introduced to solve the monitoring of dam in detail. First, the partial least square regression was used to find the most important components with strong interpretation capacity for dependent variables and satisfactory depiction for independent variables. Then, the extracted outputs were regarded as the input of the artificial neural network. Application results show that the learning efficiency and in- spection accuracy of the network are improved.