利用改进的和声搜索算法求解Van Genuchten方程参数

Solving parameters of Van Genuchten equation by improved harmony search algorithm

Van Genuchten方程是最常用的土壤水分特征曲线方程,运用该方程的关键是4个参数的取值精度。为了精确地求解这些参数,引入和声搜索(HS)算法进行求解,提出一种基于全局信息的和声搜索优化计算方法——IGHS。IGHS算法具有如下特点:利用当前和声记忆库中的全局最优解产生新解,改变了和声搜索算法新解的产生方式;通过对和声记忆库中当前最优解的扰动避免算法早熟,增强算法的全局搜索能力;IGHS算法结构简单,容易实现。实验结果表明IGHS算法求解Van Genuchten方程参数的精度与随机微粒群结果相似,但其收敛速快、计算量小,因此可以作为计算Van Genuchten方程参数的新方法。

Van Genuchten equation is the most commonly used soil water characteristic curve equation, and its parameter value precision is the key to the use of the equation. In order to solve these parameters accurately, the Harmony Search （HS） algorithm was introduced, and a new HS algorithm based on the current global information named IGHS was proposed. IGHS algorithm has the following characteristics： firstly, IGHS employs a new method for generating new solution vectors, which uses the current global optimum in the harmony memory. Secondly, in order to avoid premature and enhance global search ability, IGHS disturbs the current global optimum at a certain probability. Lastly, the algorithm is simple, and easy to implement. The experimental results show that the solution accuracy of IGHS is similar to the random Particle Swarm Optimization （PSO） algorithm, but the convergence of IGHS is faster than PSO and the calculated amount is smaller, so IGHS can be used as a new method to calculate Van Genuchten equation parameters.