Abstract

The problem of evaluation of model parameters for supercritical fluid extraction of oil from ground plant materials with high initial oil contents was studied. The information content of laboratory experimental data was analyzed by the Markov chain Monte Carlo method. The accuracy of evaluation of the main effective parameters of the model was determined for the apparent diffusion coefficient, saturation concentration of oil in the solvent, and the total content of target compounds in the raw material. Various parameterizations of the inverse problem were considered. Given the initial oil content, the diffusion coefficient can be evaluated from the individual overall extraction curve corresponding to coarse grinding. At the same time, simultaneous evaluation of all the three parameters is possible only with combined analysis of at least two curves corresponding to fine and coarse grinding. The experiment with fine grinding allows identification of only the total content of the target compounds and the saturation concentration, but cannot be used to evaluate the apparent diffusion coefficient.

中文翻译：

---

用蒙特卡罗方法评估超临界流体萃取模型参数

摘要

研究了从具有高初始含油量的地面植物材料中超临界流体萃取油的模型参数评估问题。采用马尔可夫链蒙特卡罗方法分析实验室实验数据的信息内容。通过表观扩散系数、溶剂中油的饱和浓度、原料中目标化合物的总含量来确定模型主要有效参数评价的准确性。考虑了逆问题的各种参数化。给定初始油含量，扩散系数可以从对应于粗磨的单个整体提取曲线评估。同时，只有结合分析至少两条对应于细磨和粗磨的曲线，才能同时评估所有三个参数。细磨实验只能鉴定目标化合物的总含量和饱和浓度，但不能用于评估表观扩散系数。