CAJ | 학술논문

对目前参数估计的方法进行了总结与对比.针对不完整数据参数估计时的情况,分析了目前关于参数估计方法存在的问题.利用Matlab求解非线性方程,提出了一种高精度数值解析法.利用少量的实验数据便可求出一组参数的解,然后排列组合试验所得到的数据,可以求出多组参数的解,求解多组参数的平均值,从而得到参数的精确解.并以两种工程中常用的拟合函数正态分布和威布尔分布为例,列出了数值解析法的具体求解步骤,通过与极大似然法的比较,验证了本文方法的可行性.
대목전삼수고계적방법진행료총결여대비.침대불완정수거삼수고계시적정황,분석료목전관우삼수고계방법존재적문제.이용Matlab구해비선성방정,제출료일충고정도수치해석법.이용소량적실험수거편가구출일조삼수적해,연후배렬조합시험소득도적수거,가이구출다조삼수적해,구해다조삼수적평균치,종이득도삼수적정학해.병이량충공정중상용적의합함수정태분포화위포이분포위례,렬출료수치해석법적구체구해보취,통과여겁대사연법적비교,험증료본문방법적가행성.
The current parameter estimate methods are summarized and compared. Considering data incompletion of parameter estimate, the problems existing in current parameter estimate methods are analyzed. One high accuracy method on parameter solving is put forward by using Matlab to solve nonlinear equation. A small quantity of experiment data can get a set of parameter solutions. The experiment data is organized and combined to get several parameter solutions, and their average value of parameter, then, the precise parameter solutions are deduced. Two kinds of function fitting that are popular for the engineering weibull distribution and normal distribution are taken as examples, and the parameter solution methods are listed. Compared with maximum likelihood, the feasibility of these methods are confirmed.