33 results about "Partial least squares analysis" patented technology

The invention provide a magnetic resonance imaging decoupling hydrogen spectrum method for identifying true and adulterated honey and relates to nuclear magnetic resonance spectroscopy. The method comprises the following steps: adding an identified honey sample into a buffer solution, centrifuging the prepared sample, removing supernatant liquid, and transferring the sample into a magnetic resonance resonance tube; importing a precompiled decoupling NMR spectrum pulse sequence into a nuclear magnetic resonance spectrometer; setting pulse sequence parameters and sampling parameters, and collecting data; after the sampling is finished, carrying out relevant data processing so as to obtain a decoupling spectrogram of all samples, and preprocessing corresponding experimental data, so as to obtain subsection integral data of the decoupling spectrogram; and importing the decoupling the subsection integral data into statistical analysis software, carrying out principal component analysis and orthogonal partial least square analysis so as to judge the sizes of inter-group and in-group differences of two groups, analyzing a compound which causes the inter-group differences, and judging an adulteration manner of the adulterated honey. The pulse sequence is homonuclear wide-band decoupling pure chemical displacement spectrum method, and a 1HNMR spectrum with relatively high resolution for eliminating coupled splitting can be obtained.

The invention discloses a radix scutellariae identification method. According to the identification method, radix scutellariae is used for preparing a solution of a test product, solutions in which abaicalin reference substance, a wogonoside reference substance, a baicalein reference substance and a wogonin reference substance are dissolved are used as reference substance solutions, and through high performance liquid chromatograph measurement, liquid chromatograms of the solution of the test product and the reference substance solutions are obtained respectively; the obtained liquid chromatograms are guided into a traditional Chinese medicine chromatograph fingerprint similarity evaluation system for analysis, through analysis, a radix scutellariae fingerprint is obtained, and hierarchical cluster analysis or principal component analysis or partial least square analysis is adopted for processing data of the radix scutellariae fingerprint, so that classification identification is conducted on the radix scutellariae or processed products of the radix scutellariae. According to the radix scutellariae identification method, on the basis of building the radix scutellariae fingerprint,the radix scutellariae index component content is measured, and by combining different chemical mode identification methods, the origin places, processing and appearance characters of the radix scutellariae can be effectively judged and distinguished.

The invention provides a method for identifying organic milk and conventional milk based on metabonomics, and relates to the technical field of milk identification and detection. The method comprises the following steps: (1) carrying out sample pretreatment on organic milk and conventional milk under the same condition; (2) carrying out data acquisition on the pretreated sample through high performance liquid chromatography-tandem high-resolution mass spectrometry to obtain a metabolite detection multimodal map; (3) performing peak extraction and correction on the chromatographic peak to obtain the relative peak area and retention time of all metabolites; (4) carrying out principal component analysis and orthogonal partial least square analysis, screening differential markers, and obtaining an identification model to distinguish organic milk and conventional milk; and (5) comparing the difference marker with an online database for identification. The method disclosed by the invention can be used for rapidly and comprehensively detecting and identifying the milk products, the detection result is accurate, and the operation method is simple.

The invention belongs to the field of spectral analysis, and provides a method for rapidly selecting multiple characteristic spectral lines related with specific element content from a lot of spectrum data, which can greatly increase the selection efficiency, effectively reduces the characteristic spectral line quantity which participates in modeling, simplifies a calculating process, and increases the stability and accuracy of a quantitative prediction model. The rapid selection method comprises the following steps: 1) performing spectrum pretreatment on spectral original data; 2) dividing the all spectral variable spectral lines X into several groups of Xi at intervals; 3) performing integral partial least squares on all variables X' corrected by the step 2); 4) performing partial least squares on the current all variables X'; and 5) comparing the importance of the current variables with threshold one by one; and 6) if quantitative analysis requirement is satisfied, stopping an iteration process, otherwise returning to the step 4). The method is mainly used for spectral analysis occasion.

The invention discloses a cortex phellodendron chinese identification method. According to the identification method, cortex phellodendron chinese is used for preparing a solution of a test product, solutions in which a berberine hydrochloride reference substance, a phellodendrine chloride reference substance and a magnoflorine reference substance are dissolved are used as reference substance solutions, and through high performance liquid chromatograph measurement, liquid chromatograms of the solution of the test product and the reference substance solutions are obtained respectively; the obtained liquid chromatograms are guided into a traditional Chinese medicine chromatograph fingerprint similarity evaluation system for analysis, through analysis, a cortex phellodendron chinese fingerprint is obtained, and hierarchical cluster analysis or principal component analysis or partial least square analysis is adopted for processing data of the cortex phellodendron chinese fingerprint, so that classification identification is conducted on the cortex phellodendron chinese or processed products of the cortex phellodendron chinese. According to the cortex phellodendron chinese identification method, on the basis of building the cortex phellodendron chinese fingerprint, the cortex phellodendron chinese index component content is measured, and by combining different chemical mode identification methods, the cortex phellodendron chinese, the processed products of the cortex phellodendron chinese and fake cortex phellodendri amurensis can be effectively identified and distinguished.

The scheme discloses the technical field of Maotai-flavor Baijiu identification, and a method for identifying a Maotai-flavor Baijiu of a specific brand by combining a nuclear magnetic resonance hydrogen spectrum (<1>H NMR) technology with a partial least square analysis method based on non-volatile substances. The method comprises:respectively establishing a data set of a Maotai-flavor Baijiu of a specific brand and a data set of Maotai-flavor Baijius of other brands, constructing an identification model, carrying out reliability verification on the identification model, and finally identifying an unknown Baijiu sample to be detected. The method is used for discriminating Maotai-flavor Baijiu of a specific brand and other brands, whether the detected Maotai-flavor Baijiu is the Maotai-flavor Baijiu of the specific brand or not is highly truly judged, risks are avoided to the maximum extent, and economic losses of enterprises with the Baijiu of the specific brand are reduced.

The invention relates to a method for identifying tea leaf growth area differences through pure chemical shift NMR spectrums. A PSYCHE pure chemical shift spectrum process and a multivariate statistical analysis process are combined and used for composition analysis of low-concentration natural food tea leaves. An obtained pure chemical shift spectrum is subjected to principal component analysis and an orthogonal partial least square analysis, growth area identification for tea leaves is achieved, and differential substances in the tea leaves of different growth areas are found out. The PSYCHEpure chemical shift spectrum process adopted by the method can remove broad-band homonuclear 1H-1H coupling, and only preserve chemical shift information so that a high-quality pure chemical shift hydrogen spectrum is obtained by applying the method to a low-concentration sample and the spectrum resolution is increased. A PSYCHE pulse sequence is simple, operation is easy, and conventional spectrometer conditions can meet experiment requirements. Compared with present tea leaf growth area identification methods, the NMR method provided is repeatable, non-invasive, high in spectrum resolution,and the like.

The invention discloses a method for identifying local wild astragalus membranaceus. The method comprises a step of analyzing characteristic elements in an astragalus membranaceus main root slice by adopting inductively coupled plasma mass spectrometry (ICPMS) to obtain characteristic element fingerprints of astragalus membranaceus and characterizing the repeatability and accuracy of an analysis method by using an astragalus membranaceus standard substance to ensure that obtained data is real and reliable, a step of establishing a multivariate statistical discrimination model of the wild astragalus membranaceus in Shanxi province and other astragalus membranaceus by using partial least squares analysis, a step of analyzing an unknown type of the astragalus mongholicus sample by using the established model, and a step of performing judgment, wherein the accuracy of model prediction is high, and the accuracy of model prediction of the astragalus mongholicus in Shanxi province can reach 100%. The method is simple and rapid to operate, a result is accurate and reliable, the method is suitable for identification of the wild astragalus mongholicus in Shanxi province in astragalus membranaceus with different growth years in Inner Mongolia, Jilin, Gansu and Shanxi production places, and the method can provide a reference for characteristic inorganic element analysis and source researchof the astragalus membranaceus.