To optimize the extraction process of drugs in Heye Jiangzhi Prescription by backpropagation(BP) artificial neural network combined with the orthogonal test to provide a reliable research basis for the standardized production and industrialization of drugs.Methods:The water reflux extraction method was used for extraction.The content of nuciferine was used as the evaluation index，and the extraction process was screened by orthogonal test design.The experimental data obtained by the orthogonal test analytic hierarchy process was used as the input layer of BP neural network，and the comprehensive score of the evaluation index was used as the output layer of the network.The main influencing factors were simulated and optimized to obtain the optimal extraction process.Results:The optimized extraction conditions were as follows：12 times the amount of water，extraction three times，and 0.5 h each time.Conclusion:BP artificial neural network combined with the orthogonal test can be used to optimize the extraction process of Heye Jiangzhi Prescription，which is scientific，reasonable，stable and feasible，and meets the needs of Chinese medicinal preparation research and development.