论文部分内容阅读
目的应用人工神经网络结合粒子群优化法(ANN-PSO)筛选地锦草结肠靶向给药微丸制备工艺参数。方法以pH依赖-时滞型地锦草结肠靶向给药微丸的制备为研究模型,薄膜包衣增重、增塑剂与成膜材料用量比为变量,微丸体外释放总评归一值为因变量,应用BP(back-propagation,反向传播)人工神经网络建模,并结合粒子群优化法筛选微丸的处方工艺参数。结果按优化的处方工艺参数制备的地锦草结肠靶向给药微丸,在体外实验中可满足结肠定位释放的要求。结论人工神经网络建模与粒子群法寻优相结合,为解决药物制剂工艺涉及的多维复杂非线性系统的优化问题提供了有效的途径。
OBJECTIVE To screen Artemisia fuscus colon-targeted drug-loaded pellets by artificial neural network and particle swarm optimization (ANN-PSO). Methods The preparation of pH-dependent and time-retarded pellets of Colletotrichum colonum-targeted drug-loaded pellets was used as a model, the weight gain of film coating, the ratio of plasticizer to film-forming material were used as variables, As the dependent variable, BP (back-propagation) artificial neural network was modeled, and the formulation parameters of the pellets were screened by particle swarm optimization. Results According to the optimized prescription process parameters, the radix duranth pubescens colon targeted drug-loaded pellets could meet the requirements of colon-targeted release in vitro. Conclusion The combination of artificial neural network modeling and particle swarm optimization can provide an effective way to solve the optimization problem of multi-dimensional complex nonlinear systems involved in pharmaceutical preparation technology.