The present study optimizes various extraction conditions for better yield of asiaticoside in Centellaasiatica. Response surface methodology (RSM) and artificial neural network (ANN) were used for the first time here in order to model and optimize the ultrasonic extraction parameters of asiaticoside from C.asiaticaleaves for comparing and establishment of effective prediction models.The quantitative determination of asiaticoside was carried out on silica gel 60 F₂₅₄HPTLC plates by using the mobile phase consisting of butanol: ethyl acetate: water (4:1:5). The optimum sonication parameters solid:solvent ratio (1:15), sonication time (18 min), solvent composition (35% aqueous-ethanol), the experimental maximum yield obtained for asiaticoside were 0.198% and the maximum predicted yield were found to be 0.201% i.e closely related to the experimental yield. The results showed that RBF gives better performance as compared to MLP and RSM. The study suggests that RSM and ANN model system can be manipulated for the optimization and production of valuable bioactive compounds.

本研究优化了各种提取条件，以提高积雪草中积雪草苷的产量。本文首次采用响应面法（RSM）和人工神经网络（ANN）对积雪草叶积雪草苷的超声提取参数进行建模和优化，以比较和建立有效的预测模型。积雪草苷在硅胶 60 F ₂₅₄ HPTLC 板上进行，流动相由丁醇组成：乙酸乙酯：水（4:1:5）。最佳超声处理参数固体:溶剂比 (1:15)、超声处理时间 (18 分钟)、溶剂组成 (35% 水-乙醇), 积雪草苷的实验最大收率为 0.198% 并且发现最大预测收率为0.201% 即与实验产量密切相关。结果表明，与 MLP 和 RSM 相比，RBF 具有更好的性能。该研究表明，可以操纵 RSM 和 ANN 模型系统来优化和生产有价值的生物活性化合物。