The Optimization of Experimental Fitting Curves Based on Rosin-Rammler Distribution

Based on the Rosin-Rammler distribution function, this research presents a new optimization method in fitting distribution functions. It also presents the analysis and treat- ment in experiment data of Laser particle size analyzer by taking for example four kinds of materials´╝Üsilica, potassium feldspar powder, mineral powder and quartz sand. Then the Mat- lab software is used to realize the automation of analysis and processing of the experimental data by programming. Through comparing the literature with graphic curve fitting and error analysis, the results show that the optimization method is correct and reasonable. They also show that the fitting result is better than the method of direct regression analysis and of the non – optimized linear regression analysis and the curve of optimization and fitting is closer to experimental measurement curve. This method has much significance for the accurate cal- culation of the average particle size and the variance, and for the objective evaluation of powder material properties, etc.