Research On The Model And Application Of Landslide Risk Assessment Based On Association Rules And Neural Networks

The topography is complex and the climate is changeable in our country,and the mountainous area is large.In recent years,with the continuous development and progress of water conservancy construction,occurrence of geological disasters is frequent.Landslide disaster accounted for a large proportion,not only seriously hindered the construction of water conservancy projects,but also caused countless economic losses and casualties.As the premise of landslide disaster prediction and prevention,landslide risk assessment is always an important research topic in water conservancy construction.Based on the analysis of the status quo of landslide hazard research at home and abroad,data mining and deep learning are applied for landslide risk assessment research,and the research on landslide risk assessment model and method based on association rules and neural network are carried out:Firstly,in order to overcome the traditional subjective factors caused by the selection of evaluation indexes based on experts' opinions,the data mining technology of association rules is introduced into the evaluation of landslide risk,so as to improve the rationality of the selection of evaluation indexes.Then,it tries to apply the LSTM neural network in deep learning to landslide risk evaluation.BP neural network and LSTM neural network are used as the basic network structure to construct landslide risk evaluation model,and the feasibility of applying LSTM neural network model to landslide risk evaluation is discussed.After that,association rules are combined with BP or LSTM neural network models respectively.According to the basic principle of association rules,Apriori algorithm is used for data mining,and landslide risk factors are selected specifically to reduce the uncertainty of the selection of impact factors.Finally,the model is trained,validated and its applicability is analyzed by using the landslide point data of the known landslide risk level in wanzhou of the Three Gorges Reservoir.The results show that the association rules can provide a basis for the determination of landslide risk assessment factors that because of their powerful data mining function.LSTM neural network because of its memory function,can enhance the learning ability of landslide risk evaluation model,and the combination of the comprehensive utilization of association rules and LSTM neural network model,can effectively improve the accuracy and efficiency of landslide risk assessment,not only has important theoretical significance for landslide risk assessment research,but also has higher application value in the landslide risk assessment practice.