Based On The "gas" Bionic Olfactory System In The Methodology Of Chinese Traditional Medicine Quality Evaluation Of The Overall Research

Chinese Materia Medica (CMM), one of the most important parts of Chinese culture heritage, is getting more and more international attention since its effectiveness is self-evident based on thousands of years of clinical application by traditional Chinese practitioners. Authentication and quality evaluation are peculiarly important issues as the global usage of CMM increasing rapidly.Macroscopic identification is an important method for authentication of CMM. In current quality standards, Pharmacopoeia of the People’s Republic of China taken as an example, macroscopic identification occupies the important position with the unique superiority.However, the traditional linguistic descriptions of CMM’s odor are mostly vague and imprecise because its implement depends heavily on individual ability and personal experience. Furthermore, it is difficult to distinguish samples with similar odor, for example, Jianghuang (dried rhizoma of Curcuma longa L.) and Ganjiang (dried rhizoma of Zingiber officinale Rosc). Therefore, the traditional statements with regard to CMMs’ quality need to be validated and standardized by modern scientific approaches.E-nose technology, whose application in many fields has gained considerable momentum in the past decades, is a very promising opportunity for rapid detection of CMM’s odor. Two of the considerable topics in the application of E-nose are sensor array and discriminative model. However, the MOS (Metal Oxide Semiconductor) sensor array and the modeling method, which have been used in the fields of agriculture, industry, public security, etc, are unsuitable for TCM owing to their particularities. Therefore, the problem that how to make the odour of TCM digitalized and standardized using an electronic nose with optimized and suitable MOS sensor array and pattern recognition algorithm is most concerned to further study.In this paper, the research is planned to be implemented based on traditional characters identification and chromatography experiment with the bulk medicinal herb Fructus Amomi as the object. After that, the screening and optimization of MOS sensor arrays and pattern recognition algorithms are the next key items in order to achieve the research targets. The optimization of MOS sensor array based on the results of correlation and variance analysis which is much more sensitive and suitable for the extraction and identification of CMM’s odour is the first target. Secondly, the best pattern recognition algorithms according to the results of PC A, Fisher LDA, NBN, RBF and RF is planned to be established. Finally, an intelligent identification system of Fructus Amomi is due to be built. Then a new method that assesses the quality of Fructus Amomi more scientifically will be offered and a rapid, sensitive and accurate technology for the identification of TCM will be provided.In this paper E-nose was successfully employed for qualitative distinction of aroma in different CMM samples. The results give insights to further studies to reduce the number of sensors, which could potentially shorten the data processing time, and to construct the cascade classifier, which could improve the accuracy of discriminative model.