数据断面中突变点的快速剔除方法

THE FAST METHOD OF MUTING VALUE OF DATA SECTION

针对在高密度电阻率法测量过程中,数据断面易出现突变点,而造成电阻率拟断面图虚假异常,难以准确解释的情况提出了一种快速、高效的剔除突变点的方法。该方法以曲线的曲率公式为理论基础,用相邻点间的半二阶差分对突变点进行自动识别,并将其剔除。充分利用断面数据资料,然后用一元三点插值法或趋势面拟合加残差叠加算法对缺失点进行内插,以补齐数据断面中损失的数据点,使其便于进行定性或半定量解释。最后用该方法对含有突变点的剖面模型数据和野外实测数据断面进行处理,处理结果表明,突变点被完全剔除,但不损失有用信息,易于对其进行推断解释。通过实际应用,该方法具有一定的应用价值,值得推广使用。

In the measurement of the high density resistivity method,there is often some false data in the data section because of the influence of poor earth contact of electrodes and other interference factors etc.,which can cause false abnormality of data section and make data interpretation difficult.This paper puts forward a kind of fast and effective method of muting value based on the curvature formula,and it can automatically discern mutating data using half two rank difference method and get rid of the false data.In order to make full use of the materials of data section and compensate the losing data,then it was interpolated with the method of one unit three point or trend surface stacking and residual error interpolation,which make it to be convenient for the qualitative or half fixed quantity interpretation.At last,the model and practical data including mutating data are processed with this method and the processing results show that the false points of data section are completely rejected without losing the useful information,and it is easy to be interpreted.So that it is feasible and worthwhile for application.