微粒群优化求解van Genuchten方程参数

Calculate the van Genuchten Equation Parameters Based on PSO

van Genuchten方程参数的求解是复杂的非线性拟合问题,用传统方法难以有效解决。应用微粒群优化目的在于找到一条求解van Genuchten方程参数新的有效途径。以粉壤土吸湿和脱湿实测数据为实验,利用VisualBasic.NET语言进行编程,首先建立目标函数,然后计算每个微粒的适应值,通过微粒的自动寻优过程,找到最小误差平方和,最后记录下与之对应的参数,即为方程所求得解。通过最小误差平方和与其他方法进行比较,表明微粒群优化程序设计简洁、运行时间短、拟合精度高,是一种求解van Genuchten方程参数的有效方法。

The calculation of the van Genuchten equation parameters belongs to the complex nonlinear fitting problem, which is difficult to be resolved by the traditional methods. The particle swarm optimization （PSO） was applied in this paper to introduce a new method of effectively calculating the van Genuchten equation parameters. By means of the written programs of using Visual Basic. NET language, the experiments were performed based on the imbibition and drainage silt loam soil measure data. First, the objective function was established. Second, the fitting values of every particles were calculated. Third, the minimum error of square sum （MESS） was found by the particles searching optimal value automatically after the iterative computation. Finally, the parameters corresponding to the minimum error of square sum were equation solutions. Compared with other methods using MESS, the PSO was proved to be an effective algorithm for calculating the van Genuchten equation parameters with the simpler program, higher speed running and higher fitting precision.