颗粒材料粒度分布特征参数拟合方法研究

Study on Fitting Methods of Particle Size Distribution Characteristic Parameters for Particles

应用Matlab软件、高斯-牛顿法和最小二乘法原理,以高炉渣、氧化镁及菱镁矿三种材料粒度测量结果为研究对象,以Rosin-Rammler分布函数为例,编程实现颗粒材料粒度分布特征参数拟合的自动计算。通过实验数据和拟合值的离差平方和、剩余标准偏差计算分析以及拟合曲线对比表明,高斯-牛顿法优于文献中的拟合方法,计算方法合理、可行。

Blast furnace slag, magnesium oxide and magnesite were selected as samples in this paper. The characteristic parameters of Rosin-Rammler distribution function were fitted with the Gauss Newton method and the least square principle. The algorithm programming was developed with MATLAB software. The results for the sum of squares of deviations and the residual standard deviation for the particles showed that the iteration method is superior to the fitting methods mentioned in the references and the algorithm is rational and valid.