现阶段中药制剂的研发与生产仍多依赖于老药工的经验，制剂的稳定性以及制剂工艺的放大效果往往难以得到保证，从而造成了资金、资源的极大浪费。因此，如何准确设计中药制剂的处方与工艺是目前研究中的瓶颈问题。本文将人工智能中专家系统的概念引入中药制剂的研究中，通过对化学药制剂专家系统研究现状的归纳与分析，提出了一条以中药制剂原料物理特性表征为基础，运用神经网络技术构建制剂工艺模型，并结合“物理改性技术”构建意外情况解决方案的中药制剂专家系统研究思路。该系统能够预测小试的处方、工艺，甚至中试放大效果，为实现现代化的中药制剂研发与生产提供了一些参考。

At present, the pharmaceutical development as well as production process of traditional Chinese medicine (TCM) is mostly relied on the practical experience of skilled technicians. Consequently, it is hard to maintain product stability and guarantee product quality when production process is put into pilot or large-scaled production, thus leading to a great waste of capital and resources. As a result, the appropriate design of formulation and process become the bottleneck problem for TCM pharmaceuticals. In this paper, the concept of artificial intelligence expert system was firstly introduced into the research of TCM pharmaceuticals. Artificial intelligence expert system, based on the physical characterization of raw herbs and the pharmaceutical process model construction via neural network technology (ANN), and combined with physical modification technology, is aimed to make a solution to the unexpected incidents in the practical production. Accordingly, artificial intelligence expert system can make predictions on the amplification effect of pilot production as well as the formulation and process of labscaled production, thus providing some guidance in the R&D and production of modernized TCM pharmaceuticals.

上海市教委重点学科（J50302）：中药药剂学，负责人：冯怡。