论文部分内容阅读
为了解决滑坡地质灾害传统预测方法中出现的综合性、实用性不强等问题,本文研究用基于优化参数设置的BP神经网络模型来预测滑坡地质灾害。该方法基于BP神经网络,顾及与滑坡地质灾害产生紧密相关的地质条件和环境因素,对BP神经网络的输入层、隐含层、输出层的参数进行优化;再由历史的经验数据通过训练、泛化建立基于BP神经网络的地质灾害预测模型;最后,按照0和1的组合结果对滑坡地质灾害进行预测。本文利用该模型对汶川地震诱发的滑坡地质灾害进行分析预测,结果表明:该模型的预测结果与实际结果吻合度达到86%~90%,预测精度较高,验证了基于改进的BP神经网络预测滑坡地质灾害的方法是实际可行的。
In order to solve the problems of comprehensiveness and poor practicability in the traditional prediction methods of landslide geological disasters, this paper studies BP neural network model based on optimization parameters to predict landslide geological disasters. The method is based on BP neural network to optimize the parameters of input layer, hidden layer and output layer of BP neural network, taking into account the geological conditions and environmental factors that are closely related to the landslide geological disasters, and then through the training of historical experience data, Geological disaster prediction model based on BP neural network is established by generalization. Finally, landslide geological disasters are predicted according to the combination of 0 and 1. The model is applied to the prediction and prediction of the landslide-induced geological disasters caused by the Wenchuan earthquake. The results show that the prediction results of this model are in agreement with the actual results of 86% -90%, and the prediction accuracy is high. The results show that the prediction based on the improved BP neural network The method of landslide geological disaster is practically feasible.