针对常规克里金插值算法中的不足之处,通过改变粒子群算法中粒子多样性,结合地质变量的特征和数据特征,提出了一种改进的插值方法——基于约束粒子群优化的克里金插值算法,在粒子群优化过程中,通过高斯变异、样本点权重系数设定、搜索范围约束等方式提高了插值精度。实验结果表明:基于约束粒子群优化的克里金插值算法可以获得高精度的插值效果,优于常规的克里金插值。
According to the deficiency of the conventional Kriging interpolation algorithm,this paper proposes an improved interpolation method,that is,the Kriging interpolation algorithm based on the constraint particle swarm optimization(PSO)by changing the diversity of the particles in the PSO and combined with the characteristics of the geological variables and the data features.This method improves the precision of interpolation by means of Gaussian variation,setting the weight coefficient of sample points and limiting the search scope in the process of PSO.The experiment result indicates that the Kriging interpolation algorithm based on the constraint PSO can obtain a high-precision interpolation result superior to that of the conventional Kriging interpolation algorithm.