Landslide Displacement Prediction Method Based on GA-Elman Model

The deformation process of landslide displacement has complex nonlinear characteristics. In view the problems large error, slow convergence and poor stability traditional neural network prediction model, in order to better realize accurate effective displacement, this research proposes a model based on Genetic Algorithm (GA) optimized Elman network. This combines GA with optimize weights, thresholds number hidden neurons It gives full play dynamic memory function network, overcomes that single can easily fall into local minimums neuron data is difficult determine, thereby effectively improving performance model. monitoring slow-varying Guizhou karst mountainous area are selected predict verify results compared results. show GA-Elman good agreement actual landslide. average error low accuracy high, which proves role provide reference for early warning deformation.