Lexicographical dynamic goal programming approach to a robust design optimization within the pharmaceutical environment
Lexicographical dynamic goal programming approach to a robust design optimization within the pharmaceutical environment
复制标题
DOI：
10.1016/j.ejor.2013.02.017
复制
发表时间：
2013-09
期刊：
影响因子：
--
通讯作者：
Vo Thanh Nha;Sangmun Shin;S. Jeong
中科院分区：
文献类型：
--
作者：
Vo Thanh Nha;Sangmun Shin;S. Jeong
本文的主要目的是开发一种新的稳健设计(RD)优化方法，该方法基于词典编纂的动态目标规划(LDGP)方法来实现基于时间序列的多响应，而传统的实验设计形式和框架可能实现静态响应。首先，利用双重响应估计的概念，分别估计均值和方差的响应函数，作为响应面方法的一部分，提出了一种时变药物响应(即药物释放和凝胶动力学)的参数估计方法。其次，通过在动态建模环境中加入时间序列分量，提出了一种利用过程均值和方差的估计响应函数的多目标RD优化模型。最后，为了验证的目的，进行了与仿制药开发过程相关的制药案例研究。基于案例研究的结果，我们得出结论，与其他模型相比，所提出的LDGP方法能够以显著较小的偏差和MSE值有效地提供最优药物配方。
The primary objective of this paper is to develop a new robust design (RD) optimization procedure based on a lexicographical dynamic goal programming (LDGP) approach for implementing time-series based multi-responses, while the conventional experimental design formats and frameworks may implement static responses. First, a parameter estimation method for time-dependent pharmaceutical responses (i.e., drug release and gelation kinetics) is proposed using the dual response estimation concept that separately estimates the response functions of the mean and variance, as a part of response surface method. Second, a multi-objective RD optimization model using the estimated response functions of both the process mean and variance is proposed by incorporating a time-series components within a dynamic modeling environment. Finally, a pharmaceutical case study associated with a generic drug development process is conducted for verification purposes. Based on the case study results, we conclude that the proposed LDGP approach effectively provides the optimal drug formulations with significantly small biases and MSE values, compared to other models.