基于群智能算法的土壤水分特征曲线模型参数优化

Parameter Optimization of Soil Water Characteristic Curve Model Based on Swarm Intelligence Algorithm

土壤水分特征曲线Van-Genuchten模型(以下简称VG模型),因其拟合精度高、适应性广被广泛应用,但VG模型的参数多(θ,θr,θs,h,α,n,m),其参数拟合属于非线性问题。为提高参数拟合的精度,引入高斯混沌变异理论将蜣螂优化算法(DBO)进行改进,使用Circle混沌序列增加种群多样性,提高初始解质量,形成CDBO优化算法。将DBO,CDBO算法分别应用于VG模型参数优化并计算模拟含水量,使用MATLAB R2021a仿真软件和SPSS26软件进行仿真模拟以及数据分析。结果表明使用CDBO优化后的VG模型含水量模拟值与实测值的误差范围在(0,0.8),而DBO在(0,3.0);最后结合HYDRUS-2D软件进行土壤水分运移模拟,采用SPSS26软件对模拟结果进行曲线估算分析,优化前后均方根误差值(RMSE)分别是0.051、0.039,决定系数(R^(2))分别是0.733,0.859。结果表明CDBO算法优化后的VG模型在含水量模拟及土壤水分特征描述中有更高的精确度和更强的适用性。

Van-Genuchten model of soil water characteristic curve(hereinafter referred to as VG model)is widely used because of its high fitting accuracy and wide adaptability.However,VG model has many parameters(θ,θr,θs,h,α,n,m),and its parameter fitting is a nonlinear problem.In order to improve the accuracy of parameter fitting,Gaussian chaos variation theory was introduced to improve the Dung Beetle optimization algorithm(DBO),Circle chaos sequence was used to increase population diversity,improve the quality of initial solutions,and form a CDBO optimization algorithm.In this paper,DBO and CDBO algorithms were applied to optimize VG model parameters and calculate simulated water content,and MATLAB R2021a simulation software and SPSS26 software were used for simulation and data analysis.The results show that the error range between the simulated water content and the measured water content of the VG model optimized by CDBO is(0,0.8),while that of DBO is(0,3.0).Finally,HYDRUS-2D software was used to simulate soil water migration.SPSS26 software was used to estimate the curve of the simulation results.The root mean square error(RMSE)before and after optimization were 0.051 and 0.039,and the coefficient of determination(R^(2))were 0.733 and 0.859,respectively.The results show that the VG model optimized by CDBO algorithm has higher accuracy and applicability in water content simulation and soil water characteristics description.