基于最小二乘支持向量机和人工鱼群算法的预应力锚杆布置间距优化

Range interval optimization of prestressed anchors based on the least squares support vector machine and artificial fish swarm algorithm

为了以有限的实验数据确定预应力锚杆布置的合理间距,结合山东省境内的104国道界河立交桥加筋土挡土墙的失稳加固工程实例,首先通过最小二乘支持向量机拟合优化对象与优化目标之间的复杂函数关系建立模型,然后采用现场实验数据样本进行模型训练,最后采用人工鱼群算法对模型进行优化,获得合理的布置间距,并通过加固效果监测验证了参数的合理性.结果表明,该法具有建模容易、收敛快和计算精度高等特点,说明该模型是合理可行的.

The practical engineering of reinforced soil retaining walls in Jiehe Overpass,which locates on No.104 National Highway in Shandong Province of China,was taken as an example to determine the reasonable range interval of a prestressed anchor layout on the base of limited test data.Firstly,a nonlinear functional relation between the optimization parameters and optimization object was fitted by the least squares support vector machine（LSSVM）.Then,the model was trained through data samples from in-situ test.Finally,the model was optimized by the artificial fish swarm algorithm（AFSA） to get the reasonable range interval,and the same time the rationality of optimization parameters was verified by monitoring the reinforcement effect of post-construction.The results show that this model is reasonable and feasible,with the characteristic of easily modeling,rapid convergence rate and high inversion accuracy.