摘要
将智能算法引入渗透系数参数反演中,可以实现实时率定渗透系数,更好反映工程实际。本文将采用随机森林算法对支持向量机进行了修正,构造了改进支持向量机的土石坝渗透系数反演计算模型,并运用于某粘土心墙石碴坝计算中,计算结果表明:改进支持向量机方法较支持向量机法更为精确,整体较实测值偏大。该方法更好地反映渗透系数的实时性,具有一定的工程适用性与实际意义。
The intelligent algorithm is introduced into the parameter inversion of permeability coefficient, which can realize the real-time calibration of permeability coefficient, and reflect the engineering practice better. This paper use the random forest algorithm an improved support vector machine for earth rockfill dam to modify support vector machine, construct permeability coefficient inversion calculation model, and apply it to a clay core embankment dam calculation. The calculation results show that the improved method of support vector machine with support vector machine method is more accurate and the overall measured value is too large. This method has a better response to the real time of permeability co- efficient, and it has a certain practical significance in engineering practice.
出处
《低温建筑技术》
2017年第3期87-89,共3页
Low Temperature Architecture Technology
关键词
土石坝渗流计算
改进支持向量机
有限元
渗透系数
earth rock dam seepage calculation
improved support vector machine
finite element method
seepage coefficient
