改进的概率积分函数模型及曲线拟合求参研究

Research on Curve Fitting Based on Improved Probability Integral Function Model

为了克服利用概率积分函数进行曲线拟合求取沉陷参数时需对坐标原点进行假定的缺点,在概率积分函数中引入坐标原点修正参数,实现坐标原点自动搜索,使所求的沉陷参数进一步优化。利用MATLAB软件的曲线拟合工具箱,建立自定义非线性函数模型。通过对31071工作面的地表移动实测数据的曲线拟合,得出了相应的沉陷参数,印证了该方法的可行性,同时指出对于坐标起算点不能确定的其他预计函数同样可以利用增加坐标修正参数予以改进。

When using the probability integral function to fit curve for subsidence parameters, in order to overcome the disadvantage of the need to assume the coordinate origin, the correction parameters of origin of coordinate are introduced in the probability integral function, realize the coordinate origin automatic search, to further optimize the subsidence parameters. Using curve fitting toolbox of MATLAB software, create a custom nonlinear function model. By fitting curve of the surface mobile data of 31071 working face, obtained the subsidence parameters, proves the feasibility of the method. Pointed out that,when coordinate origin cannot be determined, other expected functions also could be improved by using coordinate correction parameters.