关键词: 列文伯格-马夸尔特神经网络, 工程地质, 预测评价, 步长自适应调整

Abstract: On the reliable remote sensing geological information，it establishes the evaluation and forecast model of railway route geological hazard using the adaptive adjustment learning step Levenberg-Marquardt neural network.It achieves effective assessment and prediction of engineering geological conditions and geological hazard.It studies the LM neural network’s building，training and simulation methods.Through the quantitative analysis of the evaluation unit，evaluation results are integrated into three-dimensional geographical environment of railway location system and implement the evaluation results visualization.It helps the engineers achieve the visual geological analysis and evaluation.Theoretical analysis and case study show，the Levenberg-Marquardt neural network based on the adaptive adjustment learning step has high accuracy and speed advantages，is an ideal geological hazard assessment method.

Key words: Levenberg-Marquardt neural network, engineering geology, prediction and evaluation, step adaptive adjustment