摘要
目的 将多目标同步优化技术应用于药物剂型的处方筛选中。方法 通过硫酸沙丁胺醇渗透泵型控释片的处方设计 ,将两种同步优化技术 :反应曲面法 (responsesurfacemethod ,RSM )与人工神经网络 (artificialneuralnetwork ,ANN)应用于药物剂型的优化筛选过程中 ,并将两种方法进行比较。结果 两种方法筛选的最优处方结果较为接近 ,但ANN的预测结果误差较小。结论 在处理多目标同步优化问题上 ,人工神经网络技术是值得推广应用的一种新型的处方优化筛选技术。
AIM To apply simultaneous optimization technique in pharmaceutical dosage form design. METHODS By sieving the optimal formulation of salbutamol sulfate osmotic pump tablets, two simultaneous optimization techniques: response surface method (RSM) and artificial neural networks (RSM) were employed and their generalization ability was compared. RESULTS The optimal formulations proposed by the two methods were alike, but the results estimated by ANN showed a smaller error. CONCLUSION In the comprehensive experimental design, ANN and RSM have almost the same estimate ability. The ANN is useful in solving optimization problems and should be widely applied in formulation design.
出处
《药学学报》
CAS
CSCD
北大核心
2000年第8期617-621,共5页
Acta Pharmaceutica Sinica
关键词
同步优化
反应曲面法
硫酸沙丁胺醇
渗透泵
simultaneous optimization
response surface method
artificial neural network
salbutamol sulfate
osmotic pump
