基于IAO-SVM的软岩隧道变形预测模型

Deformation Prediction Model for Soft Rock Tunnels Based on IAO-SVM

为提高软岩隧道变形预测分析精度,提出一种改进的青蒿素优化算法(IAO)优化支持向量机(SVM)的软岩隧道变形预测算法。针对青蒿素优化算法(AO)搜索效率低下、易陷入局部最优的缺点,融合Circle混沌映射初始化、正弦药物浓度因子、自适应莱维飞行局部清除策略提出IAO算法,以提高算法的寻优性能,进一步将IAO应用于优化SVM模型的惩罚因子和核参数,并以最优参数构建IAO-SVM软岩隧道变形预测模型。以在建铁路某特长隧道工程为例,验证IAO-SVM模型的预测性能。结果表明:IAO-SVM隧道变形预测模型的预测精度优于BP、SVM、AO-SVM模型,进一步说明IAO-SVM预测模型适用于实际工程隧道变形预测评估,可为隧道安全施工提供可靠的理论支撑。

In order to improve the accuracy of deformation prediction analysis for soft rock tunnels,an improved artemisinin optimization algorithm(IAO)is proposed to optimize the support vector machine(SVM)in the soft rock tunnel deformation prediction algorithm.Aiming at the disadvantages of low search efficiency and easy to fall into local optimization of artemisinin optimization algorithm(AO),an IAO algorithm is proposed by using Circle chaotic mapping initializing population,Sine drug concentration factor and adaptive Levy flight local clearance strategy to improve the optimization performance of the algorithm.IAO algorithm is further applied to optimize the penalty factor and kernel parameters of SVM model,the IAO-SVM prediction model of soft rock tunnel is constructed with the optimal parameters.Taking the construction of a special long tunnel project on a railway under construction as an example,the predictive performance of IAO-SVM model is verified.The results show that the IAO-SVM tunnel deformation prediction model has better prediction accuracy than BP,SVM,and AO-SVM models.The further explanation demonstrates that the IAO-SVM prediction model is applicable for tunnel deformation assessment in practical engineering,providing reliable theoretical support for safe tunnel construction.