生物医学工程研究
生物醫學工程研究
생물의학공정연구
JOURNAL OF BIOMEDICAL ENGINEERING RESEARCH
2014年
3期
166-169
,共4页
张应云%张榆锋%王勇%李敬敬%施心陵
張應雲%張榆鋒%王勇%李敬敬%施心陵
장응운%장유봉%왕용%리경경%시심릉
极大似然辨识%二房室模型辨识%最小二乘%绝对误差%参数辨识
極大似然辨識%二房室模型辨識%最小二乘%絕對誤差%參數辨識
겁대사연변식%이방실모형변식%최소이승%절대오차%삼수변식
Maximum likelihood algorithm identification%Two-compartment modelIdentification%The least squares method%The absolute error%Parameter identification
提出一种基于极大似然法的二房室模型参数辨识方法。为验证本方法的有效性,我们比较了基于极大似然法和递推增广最小二乘法估计得到的常用二房室模型的参数值及其绝对误差。结果表明,基于极大似然法的非线性二房室模型参数辨识准确性和可行性明显优于递推增广最小二乘法。通过极大似然法获得的较小误差的非线性二房室模型参数估计值可用于相关临床试验,有助于提高建立非线性二房室模型的实用性。
提齣一種基于極大似然法的二房室模型參數辨識方法。為驗證本方法的有效性,我們比較瞭基于極大似然法和遞推增廣最小二乘法估計得到的常用二房室模型的參數值及其絕對誤差。結果錶明,基于極大似然法的非線性二房室模型參數辨識準確性和可行性明顯優于遞推增廣最小二乘法。通過極大似然法穫得的較小誤差的非線性二房室模型參數估計值可用于相關臨床試驗,有助于提高建立非線性二房室模型的實用性。
제출일충기우겁대사연법적이방실모형삼수변식방법。위험증본방법적유효성,아문비교료기우겁대사연법화체추증엄최소이승법고계득도적상용이방실모형적삼수치급기절대오차。결과표명,기우겁대사연법적비선성이방실모형삼수변식준학성화가행성명현우우체추증엄최소이승법。통과겁대사연법획득적교소오차적비선성이방실모형삼수고계치가용우상관림상시험,유조우제고건립비선성이방실모형적실용성。
A approach according to the Maximum Likelihood method was presented in this paper to identify the parameters of the Two-compartment Model.To verify the performance of this method, the estimation parameters of the Two-compartment Model ob-tained from it and their absolute errors were compared with those obtained from the methods based on recursive augmented least -squares algorithm.It could be seen that the accuracy and feasibility of the identification-parameters of the nonlinear two-compart-ment model received by Maximum Likelihood method were obviously better than those from the recursive augmented least-squares method.So those parameters with smaller deviations can be used in correlative clinical trial to improve the practicability of the nonlinear two-compartment model.