Geological Disaster Risk Assessment Of Oil Pipeline Landslide Based On GIS

In mountain area of China,the research on oil pipeline vulnerability evaluation of the risk evaluation of landslide hazard is less.The quantitative degree of the risk evaluation result is not high.In addition,the damage of landslide to pipe is lack of consideration.In view of the above problems,this article takes the Guangyuan section(82km)of Lanzhou-Chengdu-Chongqing long-distance oil and gas pipeline as the research object,all the slope ranges which may have an impact on the section of the pipeline as the study area and the slope unit as the evaluation unit of the regional landslide hazards.Based on the regularity analysis of landslide hazard in the small watershed along the pipeline,the risk evaluation system of regional landslide hazard and the regional landslide hazard risk evaluation model based on LM-BP neural network are established by using GIS technology and the theory of landslide,oil and gas storage and transportation to complete the risk calculation of each slope unit and the risk classification.Then combined with the pipeline body and its laying information,the vulnerability evaluation model of pipeline based on entropy weight method is established to calculate the vulnerability of each pipe segment and divide the vulnerability.We can carry out the risk assessment and risk zoning of landslide in mountainous oil pipeline based on the results of regional landslide hazard evaluation and pipeline vulnerability evaluation.And a risk evaluation system for landslide hazard of mountain oil pipeline is formed.The evaluation index system for regional landslide hazard of the study area,which including elevation,slope,aspect,elevation,terrain profile curvature,NDVI,annual precipitation,lithology,distance,is established.Based on the analysis of the relationship between the evaluation factor and the landslide,the LM-BP neural network standard sample matrix is generated by using the interpolation theory on the basis of the LM-BP neural network theory.And the risk evaluation model of regional landslide hazard based on BP neural network is established.And then we evaluate the hazard of regional landslide in the study area with 315 slopes as hazard assessment objects,and assign the hazard of each slope unit.The spatial relationship index between pipeline and landslide is put into the evaluation index system of pipeline landslide vulnerability.Based on entropy weight method,a pipeline vulnerability evaluation model is established to carry out the vulnerability assessment of pipeline sections in the study area and divide the vulnerability grade.Based on the theory of pipeline landslide risk assessment,the matching of the landslide risk degree and the pipe section vulnerability is realized by ArcGIS.The risk degree of each pipe section in the study area is obtained and the risk classification of pipe section is completed.