Method and application of slope stability prediction based on GA-BP neural network
In order to more effectively predict the safety factor of slopes,a slope stability prediction model based on genetic algorithm optimized BP neural network is constructed based on the six main characteristics of slopes(weight γ,cohesion c,in-ternal friction angle φ,slope angle a,slope height H,and pore water pressure ru).Firstly,205 sets of slope cases were collect-ed to establish a sample dataset.The distribution violin plot and Pearson correlation analysis coefficient test matrix were used to visualize the distribution characteristics and correlation of characteristic parameters;Then,use the constructed prediction model for training and testing;Finally,validate the test results.The research results indicate that the violin plot distribution of each feature parameter is similar,and the correlation between feature parameters is not significant.The sample dataset is rela-tively reasonable;The prediction results of GA-BP and BP neural networks are generally close to the true values,while the model optimized by genetic algorithm has better accuracy and stability in prediction.The research results can provide certain reference for the judgment of slope stability status.
NETL
NSTL
万方数据
