Study On Set Value Parameter Optimization Method Of Supercritical Fluid Extraction Process

Supercritical extraction technology is a representative new separation technology with clean chemical and green chemistry concept. It has been widely used in chemical、pharmaceutical、light industry、food、coal and other fields. The parameter optimization method of supercritical fluid extraction process is an important topic in the study of supercritical extraction technology. The extraction temperature, extraction pressure,coolant flow and other factors in the supercritical extraction process will affect the extraction rate, and these factors not a single action, and there is a certain coupling relationship between them. This article is from the combination of theoretical study of grape seed oil supercritical CO₂ extraction experiments. The supercritical extraction technology principles and processes were studied separately in accordance with the principle of single factor analysis of the influence of various factors CO₂ supercritical extraction process on the extraction rate, then selecting RBF neural network algorithm and genetic algorithm and K-means clustering algorithm supercritical extraction process models, and the setting of parameter extraction process is optimized to obtain the optimal process parameters.The thesis includes three parts:（1） The experiment is carried out in HA221-40-11 supercritical extraction equipment,with CO₂ as extracting agent, grape seed extract as a raw material, supercritical extraction experiments based on the principle of single factor analysis of the impact of various factors on the supercritical extraction rate.（2） According to preliminary experimental data, choosing RBF neural network as a modeling tool that adopts K- means clustering algorithm to determine the neural network hidden layer nodes, and using genetic algorithm to optimize neural network weights and enhance neural network generalization ability, then finally improve accuracy of the model.（3） With the use of genetic algorithms to optimize the model to obtain the input and output of the optimal solution, then the optimal parameters was obtained as: CO₂ flow 33L/ h, extraction pressure 43 MPa, extraction temperature 39 ℃,at this time of extraction was19.08%. After obtaining the optimum parameters, the optimal experimental results and error analysis verified the feasibility of the method..