Study On The Online Expert System Of Traditional Chinese Medicine Hpce Fingerprints

On the basis of systematical summary of present literatures and experiment methods, the online expert system of traditional Chinese medicine HPCE fingerprints was constructed in this thesis. Under the guidance of this expert system, the HPCE fingerprints of Zhibai Dihuang Pill, Pedicellus Melo and Radix et Rhizoma Notoginseng were developed. Moreover, the fingerprints were evaluated by the software of Digitized Evaluation System for Super-information Characteristics of the Traditional Chinese Medicine Fingerprint 3.0, separately showing the super-information characteristics, the dual qualitative and quantitative similarities and unified characteristics of the fingerprints. This study preliminary validated the practicability of the constructed expert system and provided new methods for quality control of Zhibai Dihuang Pill, Pedicellus Melo and Radix et Rhizoma Notoginseng.1. In this chapter, the online expert system of traditional Chinese medicine HPCE fingerprints was constructed. The overall frame design and inferential flow of knowledge database and inference engine were designed in detail from three aspects, including pretreatment methods of sample, recommendation and optimization of operation conditions of HPCE.2. The experiment was conducted by using the methods recommended by the expert system. Then the extraction method and background electrolyte were optimized separately by uniform design and triangle optimization method. The detection wavelength and applied voltage were also optimized taking index of chromatographic fingerprints F as the assessment index. And the HPCE fingerprint of Zhibai Dihuang Pill was finally developed. The fingerprint was evaluated by the software of Digitized Evaluation System for Super-information Characteristics of the Traditional Chinese Medicine Fingerprint 3.0 to comprehensively exploit the potential qualitative and quantitative information and to display the digitized characteristics.3. Under the guidance of the expert system constructed, the experiment was carried out with the background electrolyte, extraction method, detection wavelength and applied voltage optimized by the methods recommended to develop the HPCE fingerprint of Pedicellus Melo. The variance characteristics of qualitative and quantitative similarities of the developed fingerprint were studied by the evaluation software. Moreover, the fingerprints were assessed by the dual qualitative and quantitative similarities (S_F and S_F', C and P), which provided a novel way for the quality control of Pedicellus Melo.4. The experiment was conducted on the basis of electrophoresis conditions recommended by the expert system. And the background electrolyte, extraction method and applied voltage were optimized according to the optimization steps recommended to develop the HPCE fingerprint of Radix et Rhizoma Notoginseng. The HPCE unified fingerprint of Radix et Rhizoma Notoginseng was also generated, of which the relative characteristics was revealed by a series of parameters. The developed fingerprints provided a novel way for the quality control of Radix et Rhizoma Notoginseng. The unified fingerprint realized drawing all the chromatographic fingerprints in the same coordinate system, which was beneficial to the comparison of the characteristics of fingerprint and the application of fuzzy recognition method of chemometrics.