摘要
针对复杂系统输出具有数据量大、高度复杂性、非线性和非平稳等特点,将数据挖掘思想和有关方法引入到导弹系统仿真模型模型验证,通过对仿真数据和试验数据进行模式表示和特征提取,将2时序数据的模式距离作为评价指标来衡量模型的有效性。仿真分析表明,该方法相比传统方法计算量小,但意义明确,具有一定的实用性和参考价值。
Aiming at the situation that the output of complex system are huge in data,complicated in structure,nonlinearity and smoothness,the idea and relevant method of data mining are introduced into model validation of missile system. Based on mining the pattern and feature of simulation and experiment data,the validity of model is measured by pattern distance between the two time series. The simulation analysis shows that the given method is simple in computation but specific in significance,practical and valuable of reference.
出处
《现代防御技术》
北大核心
2015年第1期168-172,共5页
Modern Defence Technology
关键词
数据挖掘
模型验证
模式表示
模式距离
data mining
model validation
pattern representation
pattern distance
作者简介
白瑞阳（1991-），男，甘肃庆阳人。硕士生，主要研究方向为系统建模与仿真。通信地址：710051陕西省西安市长乐东路甲字1号空军工程大学防空反导学院研1队 E-mail：brynuaa@163．com
