| Long distance oil and gas pipelines often go across large piedmont alluvial fan,As the pipeline is a shallow buried structure and has a wide range of regions,it is more damaged by geological hazards than similar linear projects?highways and railways?,thus affecting the operation of the pipeline.The precondition for the prevention and control of pipeline geological disasters is to understand the spatial distribution characteristics of soil surface properties.Spatial interpolation method is one of the important tools for soil study to study the spatial variation of erosion factors.The method is divided into two categories,one is deterministic method and the other is geostatistical method.The deterministic method creates the fitting surface through the similarity of information points or the smoothness of the surface,mainly including the inverse distance weighting method?IDW?,the trend surface method and the spline function method.The geostatistical interpolation is based on the statistical law of the sample points to analyze the spatial structure of the data,and by quantifying the autocorrelation between samples.Thus,the measured point around the establishment of space structure model,such as Kriging interpolation?Kriging?method.It is a hot topic in soil science to study the variation of surface properties of the region with less sample points.In this paper,to explore the influence on the valuation of the alluvial fan across oil and gas pipelines between two interpolation methods?anti-distance weight and Kriging?,choosing 35 sample points for test along the pipeline and pipeline near field experiment,which lies on south of saka salt lake of qinghai province.extracted the surface attributes:impact resistance,permeability,density is used for interpolation analysis,and used the reserved testing point for cross validation method for precision evaluation of two interpolation methods.The main findings include:?1?Establish the semivariable function model of resistance,permeability and density,and analyzed the spatial structure characteristics of the data.Erosion factors exist strong spatial autocorrelation in research area,through the analysis of half variant function,in the interpolation of resistance,density of ordinary kriging by using the spherical model,while permeability by using the gaussian model.?2?Through the analysis of data by Normal QQPlot tool,it is found that the density data does not need to be converted to logarithm,and itself basically conforms to the normal distribution,while the impact of resistance and permeability need to be transformed logarithmically to be a straight line.It shows that the three factors accord with normal distribution,and can be interpolated.Through the spatial trend analysis of GIS,it is found that the three attribute distributions of the surface have strong spatial trend.The index of impact of resistance and permeability coefficient all have the basic decrease trend in derections of"west to East"and"North to South".Density in East?West,South and north directions all showed an inverse"U"shape,with high middle height and low sides.?3?Cross validation is performed to compare the accuracy of two interpolation methods through the reserved test points.For resisting impact index,IDW interpolation method more accurate than ordinary kriging with a relatively small mean absolute error?MAE?and root mean square error?RMSE?,4.34?10-4 and 6.00?10-4;For permeability coefficient and density,the common kriging interpolation method is superior to IDW interpolation method.Based on the comparison of the spatial interpolation method of erosion factors on alluvial fan across oil and gas pipeline,choose the optimal method,so as to provide basic data for the line selection on alluvial fan across oil and gas pipelines,which makes cost reduction.According to the distribution diagram of erosion factors,it can give some theoretical supports on the main inspection area and where needs to strengthen the location of hydraulic protection measures in operation period. |
PDF Full Text Request