基于数据挖掘的仿真模型验证  

Validation of Simulation Model Based on Data Mining

作　　者：白瑞阳 吴晓燕[1] 陈永兴[1] 卜祥伟[1] 姚春明[1] 

机构地区：[1]空军工程大学防空反导学院,陕西西安710051

出　　处：《现代防御技术》2015年第1期168-172,共5页Modern Defence Technology

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.

关 键 词：数据挖掘 模型验证 模式表示 模式距离 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]