【摘要】 [目的]土壤表面模型是分析土壤与车辆耦合作用的基础,为获取相对逼真的土壤表面模型,本文改进并分析了一种表面插值重构方法。[方法]首先利用不平度测试仪获取3种农业土壤的三维数字高程数据,使用一定的剔除率对原土壤高程数据进行剔除,得到3组剔后数据,求出剔后数据中各行、各列2D数据的Hurst指数,并应用于随机中点位移法过程参数的求解中。改进随机中点位移法,使用此法对剔后数据进行一次和二次分形插值重构,求解并分析3种土壤原始测量数据及重构前和重构后数据的平均值、峰度系数、相对误差、分形维数和RMS高度,同时,也使用传统的随机中点位移法和迭代函数法对3种土壤的剔后数据进行分形插值,并与改进随机中点位移法进行对比分析。[结果]改进的随机中点位移法是一种高效、稳定的分形插值方法,使用此法对矩形区域高程数据进行的插值重构能使重构后的数据在分形计盒维数、平均值及峰度系数方面与原数据之间的误差基本保持在5%以内,说明使用改进的随机中点位移法重构后能较好保持原数据的特性。[结论]该研究为进一步分析土壤与轮胎耦合作用提供了精细和高效的土壤表面模型建立方法。更多还原

【Abstract】 [Objectives]Soil surface model is the foundation of soil and vehicle interaction research. In order to obtain realistic surface models,a modified method for surface interpolation and reconstruction was pioneered and analysed. [Methods]Through a laser profiler,the 3D digital height elevation data of three kinds of agricultural soil surfaces were obtained,which were reduced under a reduction rate. Based on the reduced data,hurst exponents of each line and column of the data matrix were calculated and then used for one time and quadratic interpolation. The RMD method is modified and then the fractal boxing dimension,RMS height,mean,kurtosis coefficient and the errors of the one time and quadratic modified RMD interpolating data are compared to that of the original data,reduced data,traditional RMD interpolating data and IFS interpolating data. [Results]The results showed that modified random midpoint displacement method is an efficient and stable way to reconstruct the height elevation data in the rectangular area,the errors of the fractal boxing dimension,RMS height,mean and kurtosis coefficient stay within 5% with respect to the original data. In other words,the modified RMD reconstruction method keeps the original features very well. [Conclusions]The research provided a precise and efficient agricultural soil surface reconstruction approach which makes sense for analysing the reaction of soil and tires.更多还原