摘要
对试验数据估计通常需要作正态假设,获得正态分布样本值信息。但对不符合正态分布假设的许多工程实际问题,例如多管火箭炮的弹丸落点坐标分布与正态分布相差较大,本文介绍了一种非正态分布样本值的新方法———密度函数小波估计方法。通过用正交对称小波基函数估计概率密度函数,对含有噪声的试验数据进行阈值处理。用此方法进行射弹散布估计,避免了试验数据人为进行正态分布假设带来的较大误差,实例分析表明该方法有效。
In order to obtain the sample message of normal distribution, a normal hypothesis is used for the estimation of test data in general. But many engineering practice problems are not in accord with normal distribution hypothesis, for example, there is a world of difference between the impact point distribution of multiple launch rocket and the normal distribution. A new method, a wavelet estimation method of density function, is introduced, which estimates the non-normal sample distribution. Probability density function was estimated with orthogonal symmetrical wavelet base function, test data with noise were dealt with by threshold. The dispersion of projectile is estimated with this method so that the deviation of test data from normal hypothesis is avoided. The instance analysis indicates that this method is effective.
出处
《兵工学报》
EI
CAS
CSCD
北大核心
2006年第3期394-398,共5页
Acta Armamentarii
关键词
函数论
小波基函数
正交对称
密度函数
阈值
散布
function
wavelet base function
orthogonal symmetry
density function
threshold
dispersion
