基于ARMA模型的水下爆炸冲击谱预测  

Shock response spectrum prediction of underwater explosion based on an ARMA model

作　　者：梁利华[1] 荀盼盼[1] 王保华[1] 

机构地区：[1]哈尔滨工程大学自动化学院,哈尔滨150001

出　　处：《振动与冲击》2013年第13期23-26,共4页Journal of Vibration and Shock

摘　　要：针对舰船设备抗冲击设计冲击环境预测难的问题,提出了利用自回归滑动平均模型(ARMA模型)预测水下爆炸冲击响应谱的方法,并引入遗传算法优化ARMA模型阶数。理论介绍了舰船设备水下爆炸冲击谱模型及AR-MA模型。建模所需样本数据通过对有限元软件仿真输出的冲击响应信号进行去趋势化和平稳化处理获得,根据该样本数据的自相关函数和偏自相关函数统计特性分析,验证了ARMA建模的可行性。在此基础上,利用遗传算法优化后的ARMA模型,对水下爆炸冲击响应信号进行预测,并通过分析预测误差特性评估预测效果。研究结果表明,基于遗传算法优化的ARMA模型可以很好的预测水下爆炸冲击响应信号,从而为舰船设备抗冲击设计提供帮助。An auto-regressive and moving average（ARMA） model was presented to predict an underwater explosion shock environment for ship equipments.And genetic algorithm（GA） was used to optimize the ordersof the ARMA model.Spectrum model and ARMA model for underwater explosion shock response of ships were theoretically introduced.Sampling data required for modeling were obtained from simulation results of explosion shock signals using a finite element analysis software and treated with the methods of off-trend and stabilization.The feasibility of the ARMA model was verified with the statistical characteristics of autocorrelation function and partial autocorrelation function of the sampling data.Then,the explosion shock spectrum was predicted with the ARMA model optimized using GA and its error characteristics were evaluated to judge effects of the method.Results showed that the ARMA model optimized using GA can predict the shock response spectrum of underwater explosion well.The results provided a help for shock resistance design of shipboard equipments.

关 键 词：冲击谱预测 ARMA模型 遗传算法 时间序列分析法 抗冲击 

分 类 号：O384[理学—流体力学]