基于IGWO-LSSVM的弹体侵彻地下洞室毁伤效应预测  

Effect Forecast of Projectile Penetrating Underground Arched Structure Based on IGWO-LSSVM

作　　者：张悦 孙惠香[1] 冯盛辉 刘绍鎏 韩博 

机构地区：[1]空军工程大学机场建筑系,西安710038

出　　处：《空军工程大学学报（自然科学版）》2017年第6期95-100,共6页Journal of Air Force Engineering University(Natural Science Edition)

基　　金：国家自然科学基金(51208506)

摘　　要：侵彻的巨大动能会对结构造成毁伤,准确预测弹体侵彻地下洞室的毁伤效应,对地下防护工程的设计具有重要意义。针对当前方法存在的问题,提出了改进型灰狼算法优化最小二乘法支持向量机(IGWOLSSVM)模型,通过对训练样本的训练,形成输入量到输出量之间的映射关系,根据输入量可以得出相对应的拱顶峰值压力,实现了对弹体侵彻下不同跨度、不同埋深的地下洞室毁伤效应的预测。并利用ANSYS/LS-DYNA数值仿真结果对预测结果进行验证,表明该模型对弹体侵彻地下洞室毁伤效应具有良好的预测效果,可以满足工程需要。The great kinetic energy of penetration can cause damage to the structure. It is important to the protection engineering that predicting the penetration damage effect of the underground caverns accurately. In order to solve the problems of current method, an improved IGWO-LSSVM model is put forward. This model can form a mapping relationship between inputs and outputs by training the training samples. It can get corresponding peak pressure of the vault according to the inputs, and predict damage effect of the un- derground caverns with different spans and different burial depths. The results of ANSYS/LS-DYNA nu- merical simulation show that the model has good prediction effect on penetration effect of the underground caverns, which can meet the engineering requirements.

关 键 词：侵彻 灰狼优化算法 最小二乘法支持向量机 峰值压力 预测 

分 类 号：TU312[建筑科学—结构工程]