[Multi-index optimization of extraction process of Fengyin Decoction based on BAS-GA-BP neural network combined with entropy weight method].

To optimize the ethanol extraction technology parameters of Fengyin Decoction by orthogonal experiment combined with beetle antennae search(BAS)-genetic algorithm(GA)-back propagation neural network(BPNN). Based on single factor investigation, the extraction temperature, ethanol volume, extraction time, and ethanol concentration were used as orthogonal experiment factors, and entropy weight method was used to calculate the comprehensive scores of aloe-emodin, glycyrrhizic acid ammonium salt, rhein, emodin, chrysophanol, physcion, cinnamaldehyde, 6-gingerol, extraction ratio and fingerprint similarity. BAS-BPNN model was established, and then, GA was used to predict the optimal extraction process. The results showed that BAS-BPNN was optimized to obtain the optimal ethanol extraction process of Fengyin Decoction as follows: extraction temperature of 87 ℃, adding 9 times of 75 % ethanol, and extracting for 47 minutes, with a comprehensive score of 1.052 9. Meanwhile, the optimal process parameters obtained by orthogonal design were as follows: the extraction temperature of 80 ℃, adding 10 times of 75% ethanol, extracting for 30 minutes, with a comprehensive score of 1.003 7. The comprehensive score of the process obtained from the BAS-BPNN model was slightly better than that from the orthogonal test, indicating that the optimized process from BAS-BPNN model was more ideal, so it was finally determined as the best extraction process for Fengyin Decoction. The process of Fengyin Decoction obtained from BAS-GA-BPNN has high extraction efficiency and good stability, which provides reference for the subsequent development and quality control.