Tartaric acid, as an important organic acid, is widely present in wine, fruit juice, carbonated beverages, and certain gonfectionery products. Its concentration directly influences the balance between sweetness and acidity as well as the stability of lavor. During food production, the tartaric acid concentration may fluctuate due to variations in raw materials and formulation Adjustments. Therefore, establishing a method for real-time online monitoring of tartaric acid concentration is crucial for insuring product quality and production consistency. However, conventional detection methods (e. g. titration, HPLC) suffer from response delays and are unsuitable for real-time monitoring. Considering the multivariate, nonlinear, and dynamic haracteristics of industrial processes, more accurate concentration prediction models are required. To address this, we integrate one-dimensional convolutional neural network (1D-CNN) with a feature-space attention (FSA) mechanism, resulting in a NN-FSA hybrid model. By conducting near-infrared (NIR) spectroscopy driven experiments to detect tartaric acid Concentration, this study explores the potential of CNN-FSA to improve prediction speed and model robustness, thereby Broviding an innovative approach for real-time online monitoring of solution-phase chemical processes, Spectral data were first processed using principal component analysis (PCA) combined with Mahalanobis distance to remove outliers, followed by standard normal variate (SNV) transformation to eliminate scattering and baseline drift. Subsequently, the proposed CNN-FSA hodel and the traditional partial least squares regression (PLSR) model were trained and evaluated. Model performance was Comprehensively assessed using the coefficient of determination (R), root mean square error (RMSE), and mean absolute error MAF). Six rounds of experiments were designed, with each round starting with 500 g of a mixed solution (water, ethanol, glucose, malic acid, and citric acid) as the initial substrate, supplemented with 500 g of a solution (475 g water 25 g tartaric Acid). Data from the first four rounds were randomly split into training and test sets at a 7: 3 ratio, In comparison, data from the last two rounds were used as independent test sets to evaluate the model's generalization ability rigorously. On the independent prediction sets, the CNN-FSA model achieved outstanding performance: R <^> 2 = 0 ,989 6 , RMSE 0.000 702, and MAE 0.000 580, In contrast, the PLSR model yielded R <^> 2 = 0 , RMSE-0.001 214, and MAE 0.001 059. Compared With PLSR. CNN-FSA reduced RMSE by 42. 17% and MAE by 45. 23% on the independent prediction sets. The CNN-FSA model significantly outperforms PLSR in tartaric acid concentration prediction, demonstrating superior generalization and Fobustness on independent prediction datasets.
Soil dry density directly influences the mechanical properties of compacted soil. To explore a hyperspectral-based detection method for compacted soil dry density, this study thoroughly analyzed the influence patterns of soil moisture content and dry density on soil spectra. A spectral dewatering method was proposed to improve the accuracy of soil dry density estimation. This study obtained relevant parameters and corresponding spectral data through integrated soil moisture gradient experiments, soil static compaction tests, and soil spectral measurements. By combining spectral processing analysis methods. and correlation analysis algorithms, the spectral response characteristics of soil moisture content and dry density were analyzed, leading to the proposal of the spectral dewatering method. Subsequently, optimal feature hands were screened and extracted using correlation analysis algorithms and optimal subset construction algorithms. A soil dry density estimation model was constructed using partial least squares regression. Key findings include: (1) Soil moisture content is the primary factor influencing compacted soil spectra, with dry density being a secondary factor, both significantly affect the overall compacted soil spectral signature, (2) Compared to raw spectra and soil compaction coefficients, spectra corrected by the Spectral Dewatering method exhibit significantly higher sensitivity to soil dry density. The maximum correlation coefficient R reached 0. 858, with an average correlation coefficient improvement of 33.7% (wavelet transform). This indicates that the spectral dewatering method - employed in this study effectively mitigates moisture's influence on soil spectra and enhances spectral sensitivity to soil dry density, (3) Compared to the soil compaction coefficient, the model constructed based on SD showed an average increase of 36% in R-2 and an average decrease of 9.985% in RMSE. The optimal model (7 scales) constructed using the Spectral Dewatering method achieved R-2 = 0.79Z and RMSE 0.184. demonstrating that the proposed Spectral Dewatering technique further enhances the ability of spectral data to estimate soil dry density. The conclusions drawn in this study provide fundamental. heoretical and methodological support for rapid, non-destructive monitoring of soil dry density in engineering foundations,
Powdery mildew (PM) is a common foliar disease that negatively impacts the health of rubber trees and the yield of natural rubber. Rapid and accurate disease diagnosis is essential for implementing precise control measures and ensuring optimal rubber production. This study employed hyperspectral imaging technology to analyze infected leaves in Hainan rubber plantations. Samples of rubber leaves at various infection levels were collected, and hyperspectral reflectance data ranging from 965.4 to 1 668.0 nm were obtained using hyperspectral imaging equipment. The hyperspectral data contained noise and redundant information. Three traditional models, namely Support Vector Machine (SVM), Random Forest (RF), and Multi-Layer Perceptron (MLP), as well as the Tabular Prior Data Fitting Network (TabPFN), which incorporates automatic feature weighting, were used to model and analyze both the raw spectral data and the full-band data preprocessed by Savitzky-Golay smoothing, Standard Normal Variate (SNV), and Fractional Order Differentiation (FOD). A multi-model evaluation identified the optimal spectral preprocessing method. To assess the feature weighting capability of TabPFN, Principal Component Analysis (PCA), ReliefF, Maximum Relevance Minimum Redundancy (mRMR), and HSICLasso algorithms were employed for feature selection, extracting sensitive bands associated with powdery mildew grading. The performance of the whole band and feature subsets was compared across different classifiers to determine the optimal model architecture, Finally, Shapley Additive Explanationswas used to analyze the key features and their influence on disease grading. The results showed that TabPFN butperformed all other models, demonstrating superior robustness and effective feature weighting selection, FOD preprocessing effectively reduced spectral noise and enhanced the extraction of essential detail features, resulting in the highest data quality improvement. The full-band TabPFN model with FOD preprocessing achieved a classification accuracy of 95.27%, surpassing Braditional methods by 3.24% similar to 13.24% After applying HSICLasso to select 20 critical features, the accuracy remained at 94.31%, while reducing model complexity by nearly 90% and only decreasing accuracy by 1.01%. SHAP analysis identified the 1 160 nm and 1400 nm regions as key discriminatory bands, linked to C-H and O-H chemical bond vibrations. These bands correspond to the leaf's carbohydrate, lignin, and water content, indicating the model's ability to capture spectral responses related to physicochemical changes caused by powdery mildew. This study validates the integration of FOD and TabPFN for PM Hetection, providing an accurate model for assessing disease severity, which can aid in precise pesticide application and promote the health of rubber trees, ultimately improving rubber production.
With the increasing number of document forgery and economic dispute cases, the accurate identification of homochromy inks is of great significance in judicial expertise. Traditional methods (such as thin-layer chromatography and Raman spectroscopy) have limitations including sample destruction and time-consuming procedures. At the same time, hyperspectral imaging (HSI) has emerged as a promising alternative due to its advantages of image-spectrum integration and non-destructive detection, However, the existing ink classification methods based on "dimension reduction and clustering" are difficult to fully explore the nonlinear characteristics of high-dimensional data, and shallow machine learning models have limited expressive ability and are susceptible to information loss and error accumulation. Therefore, a deep learning model H-CNN integrating multi-scale convolution and channel attention mechanism is proposed in this paper, and it is combined with hyperspectral imaging for the identification of homochromy inks. The model employs multi-branch parallel convolution to extract spectral features at different scales, comprehensively capturing spectral information across hands. And a channel attention mechanism dynamically enhances discriminative bands, focusing on key spectral information. Residual connection optimization gradient propagation is adopted to avoid gradient explosion and gradient vanishing, thereby reducing error accumulation and improving training efficiency, Experiments were conducted on the UWA Writing Ink Hyperspectral Image (WIHSI) dataset to determine the optimal training data partitioning and parameter settings. Ablation studies were designed to validate the effectiveness of the multi-branch parallel convolution structure, channel attention mechanism, and residual connections in improving model performance. Finally, the performance of the model proposed in this paper was compared with that of other model architectures on the current dataset. The experimental results show that the multi-branch structure and the channel attention mechanism improved the accuracy rates by 4.6% and 1.0% respectively, and the training cycle was shortened by 34% through the residual network connection, For the most challenging identification of the black ink, HI-CNN achieved an accuracy rate of 98.07% (an improvement of 5.3% compared to the optimal model CAE-LR), In comparison, for the identification of blue ink the accuracy rate reached 99.06%, which was generally superior to the existing methods. This study provided an accurate and efficient solution for identifying homochromy inks, thereby reducing reliance on professional expertise in forensic document examination. It had significant application value in the field of judicial expertise and promoted the leapfrog development of homochromy ink identification technology, transitioning from reliance on experience to scientific quantification,
Geological hazard prevention and control is a crucial national policy in China. Understanding the intrinsic driving factors of geological hazards is fundamental for disaster mitigation, As a province prone to geological hazards, previous studies in Guangdong have primarily focused on the physical and mechanical properties of rock and soil masses. At the same time. insufficient attention has been paid to the underlying causes of mechanical property deterioration chemical weathering. This study conducted statistical analyses on 6 841 slope hazard sites (collapses and landslides) in typical igneous and sedimentary rock regions of northern and eastern Guangdong. From the perspective of "chemical water-rock interaction (CWRI)", major chemical elements (including SiO2, Al2O, FeO, CaO, MgO, K2O, Na2O, and loss on ignition (LOI)) in 40 samples (weathered rocks, residual soils, and slope-residual soils) from 11 hazard sites were analyzed using X-ray luorescence spectrometry (XRF). Classic chemical weathering indices silica-alumina ratio (Si/Al), silica-sesquioxide ratio (Si/R2O), chemical index of alteration (CIA), and a newly proposed CIA-rate by the author were calculated to explore the driving mechanisms of chemical weathering intensity on slope hazards. Here, the CIA-rate is equal to the CIA value of the weathered product minus the CIA value of the fresh parent rock, then divided by the CIA value of the fresh parent rock. XRF results revealed that the average CIA-rate of igneous rock weathering products (71.28) was significantly higher than that of sedimentary rock products (21.26), consistent with the 1.4 times higher hazard density in igneous rock arcas. In igneous rock regions, slope hazards predominantly occur in fully weathered layers and residual soils with CIA values of 75 similar to 85 and CIA-rates of 50 similar to 70, particularly at sites with spheroidal weathering bodies or heterogeneous interlayers. In sedimentary rock regions, hazards are often observed in lower fully weathered layers and residual soils with CIA values of 75 similar to 85 and CIA-rates of 15 similar to 25, where bedding structures and "soft-hard interlayers" facilitate multi-stage or deep-seated landslides. This study demonstrates that the hot-humid climate and intense rainfall in South China enhance surface water and groundwater erosion and corrosion, accelerating chemical weathering. During desilication and enrichment of aluminum/iron, elements such as K, Na, Ca, and Mg are leached. In contrast, clay newbornminerals and iron hydroxides accumulate, weakening the mechanical properties and structural integrity of rock and soil masses. These processes serve as intrinsic drivers of geological hazards. The findings provide a scientific basis for early identification and warning of geological hazards.
Photothermal therapy (PTT) has garnered significant interest as a promising alternative to conventional cancer treatments, owing to its low systemic toxicity, high targeting precision, and minimal invasiveness. Room-temperature liquid petals (LMS) - a class of functional materials exhibiting both metallic conductivity and fluidic processability have emerged as Ittractive candidates for PTT applications. Their unique attributes, including fluidity, dispersibility, high thermal conductivity, efficient photothermal response, and biocompatibility, underscore their potential in this field. However, the practical application of LMs is hampered by their high surface tension and reflectivity, which limit photothermal conversion efficiency. Notably, the highly dynamic and dispersible nature of LM surface atoms offers a pathway for modulation via surface chemical modification, In This study, we employ sodium alginate (SA), a nontoxic and biocompatible natural polymer, to functionally tailor the surface of Ms. Through ultrasonication, a uniform SA coating was formed on LM nanoparticles, yielding well-dispersed core-shell panostructures, termedLiquid metal (@sodium alginate (LM@SA). The SA-modified nanoparticles exhibited remarkable photothermal performance; under 808 nm near-infrared laser irradiation at 1.5 W cm(-2), the heating rate reached 5.4 degrees C. min(-1), with a temperature plateau of 63 degrees C attained within 4 minutes. The photothermal conversion efficiency was calculated to be 41.9%. Furthermore, the SA coating significantly enhanced colloidal and thermal stability, as evidenced by consistent heating Performance over eight consecutive laser on-off cycles without noticeable decay. In summary, this work demonstrates a biocompatible polymer-based strategy for effectively regulating the surface optical properties of LMs. The resulting LM@SA Aanomedicine exhibits efficient, stable photothermal behavior, offering a promising platform for precise, effective tumor photothermal therapy.
Against the background of new agricultural construction, improving the practical innovation ability of undergraduates in agricultural colleges and universities has become the core goal of experimental curriculum reform. The traditional experimental course content is obsolete, and the technical method is single, which makes it difficult to meet the needs of new agricultural science for compound talents. In this study, Fourier transform infrared spectroscopy (FTIR) was used as a starting point to systematically compare the operational procedures, data characteristics, and teaching applicability of attenuated total reflection (ATR) and transmission (FTIR) for the first time in the determination of soil functional groups under different fertilization treatments. Based on the soil analysis experiment, a three-stage teaching mode of "basic cognition-method comparison comprehensive application" was developed, with teaching conducted through group experiments and comparative analysis, The results showed that: The average peak height of the OH functional group (3 400 cm(-1)) detected by the FTIR method was 0.32 +/- 0.04(#3), and the ratio of the CH vibration peak area at 1 440 cm(-1) was 3-4 times that of the ATR method, which was more suitable for accurate quantitative analysis. After the reform, the score rate of students in the final "ATR/ FTIR method selection application question" increased from 62% to 89%, the success rate of FTIR tableting increased from 65% to 88%, and the error points of the self-recognition experiment increased from 1.8 to 4. 2. Under different fertilization treatments, SNPK treatment significantly reduced the content of soil aromatic compounds (ATR method, 3.16% +/- 0.14% vs. CK: 4.27%+/- 0.01%), and 50% NPK 50% M treatment significantly increased the content of phenols and alcohols (FTIR --method: 15.35% 1.9326_CK 13-12261222) Through the logical closed loop of "technical comparison-practical verification-innovative application", this teaching mode significantly improves students' scientific research cognitive depth, operational innovation ability and scientific research thinking, more effectively connects the principle of FTIR technology with the needs of new agricultural soil research, optimizes the structure of agricultural resources and environment experiment course, and provides a replicable practical path for the cultivation of new agricultural cross-disciplinary talents.
With the rapid development of new energy vehicles and energy storage devices, the number of waste lithium batteries Las surged. Black mass, as the most critical material in the battery recycling process, has a complex and diverse composition. which is very likely to cause resource waste and environmental pollution if it cannot be effectively identified and categorized, Traditional detection methods are time-consuming and costly, making it difficult to meet the demand for real-time classification of plack mass in industrialized scenarios, Laser-induced breakdown spectroscopy (LIBS) offers a new approach for rapid Hentification of black mass. leveraging its advantages of simultaneous multi-element detection, rapidity, and high efficiency. In this study, a handheld LIBS spectrometer is combined with machine learning algorithms to achieve accurate identification and fficient classification of black mass from used lithium batteries. The experiment firstly purchased nine common lithium battery black mass samples from Ganzhou Haohai New Material Co. Ltd, and collected the spectra of the black mass samples by a Kandheld LIBS instrument; In order to improve the quality of spectral data and the accuracy of the subsequent modeling, maximum and minimum normalization (MMN) and Savutzky-Golay smoothing filter (SG) were used to optimize the preprocessing of LIBS spectral datas In the feature extraction stage, the pre processed spectral data were subjected to dimensionality reduction by introducing two data dimensionality reduction methods, Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA), respectively; Finally, three types of classification models, namely, Random Forest (RF), Partial Least Squares Discriminant Analysis (PLS-DA) and Back Propagation Neural Network (BPNN), were established based in the dimensionality reduced spectral data, The optimal black mass classification model is selected by comparing four aspects: @lassification accuracy, precision, recall and F1 score of the test set. The experimental results show that the classification model Constructed using a combination of Linear Discriminant Analysis (LDA) and a Backpropagation Neural Network (BPNN) achieves the best recognition performance, with an overall accuracy of 99. 70% on the rest set. The results validate the feasibility and effectiveness of LIBS technology combined with machine learning methods for identifying lithium battery black mass, providing a theoretical basis and practical value for the efficient classification and reuse of waste lithium battery black mass.
Detecting trace organic compounds in deep-space minor celestial bodies is crucial for understanding the origins of life. However, conventional spectroscopic techniques often struggle to simultaneously excite and release all the organic compounds in the sample for comprehensive detection. This is particularly challenging for organic compounds that are diffusely distributed, as their signals are often difficult to capture effectively, leading to limitations in detection within complex matrices. To address this Challenge, this study proposes a novel analytical approach that combines laser pyrolysis-Fourier transform infrared spectroscopy LP-FTIR) with machine learning, establishing a high-precision method for both qualitative and quantitative detection of organic compounds in space dust. This work aims to provide a new technical solution for identifying potential biosignatures in deep-space Exploration. First, simulated space dust samples containing six typical life-related organic molecules glycine, stearic acid, cytidine nucleoside, ribose, deoxyribose, and soybean lecithin were prepared. Infrared spectral data of pyrolysis gases were obtained using a miniaturized LP-FTIR detection platform. For qualitative analysis, a multi-model ensemble classification Algorithm was developed, integrating SVM, RF, XGBoost, RNN, and BPNN, with hyper parameters tuned using Bayesian ptimization, Predictions were integrated through a majority voting mechanism, For quantitative analysis, a novel regression model was crafted by integrating a one-dimensional CNN with a multi-head attention mechanism, employing segmented pooling to pinpoint critical spectral regions and improve feature extraction efficiency. The study results indicate that the multi-model insemble classification method achieved a 90% accuracy rate in identifying six types of organic compounds, representing a Significant improvement over the best single model (BPNN at 87%). The improved attention-based CNN achieved a coefficient of determination (R-2) of 0.979 and a root mean square error (RMSE) of 0. 21 mg in predicting glycine content, showing significant performance enhancement over traditional PLSR (R-2 = 0.969) and the basic CNN model (R- 2 = 0.891)
Breast cancer and ovarian cancer are common malignant tumors in women, and the differences in their metabolic activity and protein structures reveal unique pathological mechanisms. However, due to the overlap in symptoms and molecular haracteristics between the two, clinical diagnosis and differentiation remain challenging. A systematic study of the metabolic processes and protein conformational changes in breast cancer and ovarian cancer provides scientific evidence and guidance for disease diagnosis and personalized treatment. This study, based on attenuated total reflection Fourier transform infrared (ATR-TIR) spectroscopy combined with machine learning methods, explores the differential spectral markers of breast cancer and varian cancer and evaluates their diagnostic and discriminative potential. A total of 157 female participants were included in the Study, including 67 breast cancer patients, 41 ovarian cancer patients, and 49 healthy controls, with serum samples collected for Spectral analysis. The results show that at the 1 450 cm(-1) band, the absorbance of the breast cancer group was significantly bigher than that of the ovarian cancer group (p < 0.05) , accompanied by a blue shift in the wavenumber, suggesting lipid metabolism and cell membrane synthesis abnormalities, Peak fitting analysis of the Amide I region revealed that the a helix proportion in the breast cancer group was significantly lower than that in the ovarian cancer group (p < 0.05) In comparisor the beta-sheet proportion in the ovarian cancer group was significantly higher than that in the breast cancer group (p < 0.05) , revealing & iexcl;pecific differences in protein conformation changes between the two cancers. The Linear Discriminant Analysis (LDA) model onstructed using the relative intensity ratio of 1 450/1 650 cm(-1) and Amide | spectral data showed a reasonable differentiation performance (AUC 0.851, Specificity 73.2%, Sensitivity 80.3%). The results of this study indicate that ATR-FTIR spectroscopy combined with spectral feature analysis and classification models can provide effective support for the diagnosis and differentiation of breast cancer and ovarian cancer, laying the foundation for future cancer subtype diagnostic research.
The Multi-wavelength pyrometer is a crucial non-contact temperature measurement instruments that simultaneously detect radiant energy from targets at different wavelengths, enabling real temperature retrieval through data processing. This measurement methodology eliminates physical contact with the measured target, thereby preserving its original thermal Characteristics and temperature distribution, making it particularly valuable in high-temperature and ultra-high-temperature applications. The temperature inversion process fundamentally relies on establishing correlations between emissivity and wavelength or radiance temperature. Emissivity, a critical parameter quantifying a radiator's emission capacity relative to blackbody radiation, serves as the bridge connecting real-world radiators with blackbody radiation laws. By determining the Emissivity and radiance temperature at specific wavelengths, the real temperature can be computationally derived. Despite decades of international research advancements, two persistent challenges remain: (1) Time-dependent variations in emissivity an lead to significant temperature calculation errors whenever applied models differ from actual conditions: (2) Conventional missivity temperature-wavelength relationship models, developed typically through rigorous experimentation and empirical Validation, demonstrate limited generalizability and often fail to perform effectively whenever the objects or conditions being preasured are altered. The study proposes an innovative emissivity model-independent methodology for rapid real temperature inversion in multi-wavelength pyrometry. By analyzing intrinsic constraints within multi-wavelength measurement theory and tegrating them with multi-constraint optimization principles, we developed a novel temperature inversion framework. The approach eliminates traditional dependence on predefined emissivity models while maintaining measurement accuracy. Through Figorous theoretical derivation and experimental validation, we demonstrated the feasibility and universality of this multi-onstraint optimization-based method, establishing a new paradigm for multi-spectral pyrometric temperature solutions. This Advancement provides enhanced adaptability for diverse measurement scenarios and evolving target characteristics.
Meat is an important source of protein and nutrients in the human diet, and the adulteration of meat has become a major problem in the field of food safety in China. To address the problems of complex operations, time-consuming procedures, And high equipment costs in traditional meat detection methods, this paper proposes using Laser-induced Breakdown Spectroscopy (LIBS) combined with a deep learning network to detect and classify multi-variety meat tissues quickly. Spectral data for beef, mutton, and pork were collected using LIBS, Nine spectral lines of five elements were selected as the analysis pectral lines and model input for modeling and recognition. A ResNet18 backbone network was designed, and three machine learning models were designed to model and recognize the spectral data. The results show that the deep learning network achieves the best recognition performance, with an accuracy of 98.1%. Among the 120 groups of spectral data. 117, 119, and 117 roups of beef, mutton, and pork spectral data were identified correctly, respectively. In the horizontal comparison using the Jame deep learning model, the ResNet18 model was superior to the three deep learning models, GoogLeNet, Vgg16, and ResNet50, in the recognition of meat spectral data. On this basis, the model's generalization was verified using re-collected data, and the accuracy reached 98.9%, indicating that the model maintains strong cross-data-set recognition ability. It has good generalization and consistency. The above research shows that the combination of LIBS and a convolutional neural network can provide objective, quantitative information on differences between meat varieties in multi-variety meat classification and recognition tasks and has the potential to quickly and in situ diagnose different types of meat tissue.
A control algorithm for realizing closed-loop fiber positioning was proposed with the structure of the special-shaped micro-lens aimer (SMART) that can achieve real-time fiber positioning, to fit the requirement of the dual-rotary positioning device on the focal plane of the Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST). During the closed-loop qontrol process, if the fiber and the star image were misaligned, the starlight would deviate from the center plate of the SMART And enter the special-shaped micro-lens, which would deflect part of the starlight into the corresponding feedback fiber. Based on the light intensity signals received by the six feedback optical fibers, the light intensity contrast relative to the feedback optical fibers is calculated to obtain the azimuth information of the misalignment, and a control signal is sent back to the optical fiber Positioning unit to drive the dual-rotation positioning device to adjust the angle, thereby achieving high-precision alignment Aynamically. Its design logic closely aligns with the optical fiber alignment requirements of the dual-rotation positioning device, as the device relies on two independent rotation axes (a central axis and an eccentric axis) to achieve two-dimensional positioning. The closed-loop control dynamically corrects the dual-axis angle using real-time feedback signals to compensate for mechanical Irrors and environmental disturbances, ultimately meeting the requirements of large-scale optical fiber arrays for high-precision. high-efficiency alignment. The correspondence between the rotation angle and the movement distance of the dual-rotary positioning device was analyzed, and a two-round, multi-step convergence method was developed. The calculation formulas for the alignment path between the fiber and the star image under different dual-axis expansion conditions were analyzed. In the first Hound of positioning, the single-step length was set to 30 & micro;m, and the contrast threshold was set to 0.9, in the second round, the single-step length was set to 10 & micro;m, and the contrast threshold was set according to the calibration. The fiber needs to undergo multiple movements, from misalignment to alignment. After each movement of the positioning unit, the control system would again obtain the real-time feedback signal from the detection system to determine whether the fiber had reached the Threshold, A simulation system of LAMOST was built in the laboratory, and the lengths of the dual axes were calibrated. The positioning accuracy and positioning time of the fiber closed-loop positioning system were tested. The results showed that the positioning system could make any initially misaligned fiber return to the aligned state. The average correction time was 27. 6 s. There were 72.5% fibers of 10 & micro;m correction accuracy, and 97.5% fibers of 30 & micro;m correction accuracy
Pyroxene megacrysts are common minerals in Cenozoic alkaline basalt bodies in the eastern region of China. These pyroxenes and their internal inclusions provide important evidence for studying the composition of deep-mantle materials and deep-seated processes in eastern China, This study employs electron probe microanalysis (EPMA) and laser Raman spectroscopy ko investigate the chemical composition and Raman spectral characteristics of pyroxene megacrysts and their microscopic Inclusions in basalt bodies from the Mingxi area. Fujian Province. The results indicate that the Mingxi pyroxenes are sugites. characterized by well-developed fractures and relatively high Ca and Mg contents. Their calculated chemical formula is (Ca-0.66 similar to 0.68 Na-0.04 similar to 0.05 Mg-o. 27 similar to 0.30) (Al-0.25 similar to 0.26 Fe2+ (0.18 similar to 0.21) Fe3+ (0.04) Ti-0.02 similar to 0.03 Mg-o. 48 similar to 0.50 ) [ ( Si-1.86 similar to 1.92 Al-0.08 similar to 0.14) O-6]with Raman Characteristic peaks located at 1 008, 670, 545. 396. 338. and 232 cm(-1). The pyroxenes contain abundant linear, dotted, or dashed metallic mineral inclusions, primarily composed of pyrrhotite and hematite mixtures, while fractures exhibit goethite impregnation, Among these, pyrrhotite is a mantle-derived inclusion with a chemical formula of Feo S, whereas hematite And goethite are its later oxidation and hydration products. The fluid inclusions in Mingxi pyroxenes are mainly CO2, with a Fermi resonance doublet peak separation (Delta sigma) of 104. 643 cm(-1) in Raman spectroscopy. Based on the linear relationship between the Fermi resonance doublet and gas density, as well as the ideal gas equation of state, the formation depth is estimated to be greater than 72 km. The findings of this study provide valuable data for further research on the formation of basalt bodies and deep-seated processes in southeastern China.
Using ordinary ground calcium carbonate (D-50 approximate to 38 mu m) as the core phase, a CaCO3/ Al(OH)(3) core-shell composite material was constructed in a supersaturated sodium aluminate solution via heterogeneous nucleation. By multiple spectroscopic techniques, including X-ray diffraction (XRD), scanning electron microscopy-energy dispersive spectroscopy (SEM-EDS), Fourier transform infrared spectroscopy (FTIR). Raman spectroscopy, and X-ray photoelectron spectroscopy (XPS), the dynamic evolution of phase composition, interfacial chemical states, and bonding structures during the coating process was A systematically revealed. The coating pathway was regulated by reaction time gradients (2 h, 12 h, 36 h). Combined with SEM EDS cross-sectional morphology analysis and XRD phase identification, a three-layer model of the core-shell structure was stablished: a calcite-type CaCO3 core ( approximate to 38 mu m ).a C(3)AH(6) intermediate layer (300 - 500nm) and a dense Al (OH)(3) outer hell, following a stepwise reaction path of "CaCO3 -> C(3)AH(6)-> Al(OH)(3)". FTIR and Raman spectra further corroborated the interfacechemical bonding path; FTIR detected the characteristic OH stretching vibration peak of C(3)AH(6) at 3 664 cm(- 1) while Raman identified additional lattice vibration modes of C(3)AH(6) in the 3390 similar to 3 640 cm(- 1) range, collectively confirming chemical bonding (via Ca-O-Al bonds) rather than physical mixing between calcium and aluminum, XPS quantitative analysis demonstrated that the surface Ca/Al molar ratio decreased from 1.45 at 2 h (close to the theoretical value of 1.5 for C(3)AH(6)) to 4.08 at 36 h. Combined with the Ca(2p) binding energy shift and the Al(2p) chemical state transition (Al-O -> AI-OH ), this directly revealed the dynamic formation of the C(3)AH(6) intermediate phase and the coverage mechanism of the Al (OH)(3) shell, The innovation of the multispectral coupling strategy lies in, achieving phase-morphology collaborative characterization through XRD/SEM-EDS, analyzing bonding structure evolution with FTIR/Raman, and quantitatively tracking surface chemical state migration with XPS. This integrated approach overcomes the limitations of single-technique characterization and systematically glucidates the pivotal role of the C(3)AH(6) intermediate phase in interfacial bonding. The study provides a theoretical foundation for the controllable synthesis of core-shell composites. For instance, optimizing the Lambda Al (OH)(3) shell thickness (300 similar to 500 nm) and Aniformity by adjusting the sodium aluminate concentration and reaction time can enhance the material's weathering resistance and interfacial compatibility. Through multidimensional synergy of spectroscopic techniques, this work establishes a critical theoretical basis for the interfacial chemical regulation and industrial applications of CaCO3/Al(OH)(3) composites.
Bauxite is one of the strategic minerals in China, Accurate and efficient analysis of the contents of major, associated Beneficial, and harmful elements is of great significance for evaluating and comprehensively utilizing bauxite. At present, the ystematic analysis of bauxite is still dominated by traditional chemical methods and atomic absorption spectrometry, with relatively low analytical efficiency. It is necessary to develop a simple, efficient, and simultaneous multi-element analysis method for bauxite. The key points of simultaneously determining the multi-element content in bauxite by inductively coupled plasma Optical emission spectrometry (ICP-OES) include selecting efficient decomposition methods for refractory bauxite samples, improving the precision and accuracy of determination under high dilution conditions, and eliminating matrix and coexisting lement interferences. This study proposes a method for 10 elements (Al, Si, Fe, Ti. Sr, Li, Cr, V, Zr, Sc) in bauxite using Aixed-flux fusion-ICP-OES with an offline internal standard. The effects of mixed flux dosage and melting temperature on Sample decomposition efficiency were systematically investigated. 0.1 g sample was added to 0.7 g mixed flux and melted at 1000 C-omicron for 20 min to ensure complete decomposition of the sample: the extraction conditions of the fused samples were Optimized, Ultrasonic extraction with 15% hydrochloric acid was selected, as it offers high extraction efficiency without silicic Acid precipitation, Based on the content of various elements in bauxite, the sensitivity and interference of spectral lines, the analytical spectral lines of the determined elements were selected. By comparing the correction effects of Cd 214.438 nm. Cd 128.802 nm. Co 228. 616 nm, and Co 345. 351 nm on the 10 elements, Co 228.616 nm and Co 345.351 nm were selected as ternal standard lines to correct different elements, respectively, which significantly improved the precision of the determined lements. The interference of aluminum at different concentrations on other elements was systematically investigated, and it was found that the degree of interference varied among elements. The concentration of aluminum in the sample solution (dilution of 500 times) was generally not more than 200 mu g ml(-1), and the interference on other elements could he ignored. The detection Amit for each element in this method ranged from 0.5 mu g(-1) to 0.1%. According to the verification of certified reference materials, the relative standard deviation (RSD, n = 12 ) of all elements except lithium was less than 5%. The relative error (RE) of major elements (Al2O, SiO2, Fe2O3, TiO2) ranged from 1.81% to 1.61% (e. g., Al2O3 - 0.65% similar to 0.28%) , while the relative error of other elements ranged from 10.53% to 12.78%. The determination values were hasically consistent with the certified values. This method is convenient and fast, with a low detection limit, high accuracy, and good precision, and is suitable for the simultaneous analysis of multiple elements in hauxite with different contents.
To address the issues of peak overlap caused by complex matrices in agricultural product terahertz (THz) spectral signals and the dynamic, nonlinear interference induced by environmental and system noise, this study explores the feasibility of adaptive-signal-decomposition-based denoising methods to improve THz spectral quality, THz time-domain spectroscopy (THz TDS) combined with an attenuated total reflection KATR) accessory was used to collect THz absorbance spectra from 48 peanut samples. Taking the quantitative prediction model of peanut moisture content based on THz-ATR as an example, wavelet transform (WT), empirical mode decomposition (EMD), local mean decomposition (LMID), and its improved methods segmented local mean decomposition (SLMD) and piecewise mirror extension local mean decomposition (PME-ILMD) were employed for spectral denoising. The applicability of different denoising methods was evaluated using a support vector regression (SVR) model Experimental results show that the peanut moisture-prediction model constructed after PME-LMD denoising achieved the best performance, with a root mean square error (RMSE), coefficient of determination (R-2), and mean absolute percentage error (MAPE) of 0.010, 0.912, and 0.040, respectively. Compared with traditional methods, PME-LMD significantly improved spectral quality and model prediction performance. The PME-LMD denoising strategy proposed in this study effectively suppresses non-uniform noise interference in THz spectral signals, providing an efficient and accurate preprocessing method for THz spectral analysis of agricultural products. This research provides theoretical support and technical guidance for the application of THz technology for detecting agricultural product quality.
Sudan Red I is an illegal food colorant that can enhance the color intensity of egg yolks. Rapid detection of Sudan Red in egg yolks is of great significance. In this study, a near-infrared spectrometer was used to collect spectral data from 60 unadulterated egg yolk samples and 102 adulterated samples containing Sudan Red I at concentrations ranging from 0.5 to 10 mg (100 g). After spectral analysis and data preprocessing, the sample dataset was divided into training and test subsets Ita 3:1 ratio, Qualitative and quantitative models were then built to detect Sudan Red I in egg yolks. The models were Evaluated using prediction accuracy, calibration, and prediction R? coefficients (R-c(2)/R-p(2)), and root mean square errors (RMSEC/RMSEP). For qualitative analysis, the Partial Least Squares Discriminant Analysis (PLS-DA) algorithm was used to classify igg samples as adulterated with Sudan Red 1. After data preprocessing using the Standard Normal Variate (SNV) Fransformation, the model achieved optimal performance, with accuracy rates of 98.3% for the training set and 97.6% for the fest set. For quantitative analysis, the Competitive Adaptive Reweighted Sampling (CARS) method was first used to select characteristic wavelengths from the spectral data. Then, regression models were established using the linear Partial Least Squares Regression (PLSR) and the nonlinear Back-Propagation Artificial Neural Network (BP-ANN) algorithms to predict Sudan Red I content. The PLSR model showed better performance, with R-c (2) of 0.98. R-p(2) of 0.98. RMSEC of 0.79, and in egg yolks.
The extraction and identification of bloodstains left at crime scenes provide an important basis for case investigation,< br /> 1 But their rapid and non-destructive development and examination remain a research hotspot in the field of forensic science, To Enhance the development efficiency and detection accuracy of bloodstains, hyperspectral technology has gradually been applied to the non-destructive identification of bloodstains. However, existing hyperspectral imaging techniques have limitations, including ow recognition accuracy and insufficient efficiency in identifying bloodstains and blood-like substances, particularly when dealing with bloodstains on complex objects, Therefore, a bloodstain identification model based on hyperspectral imaging technology by htegrating the SENet channel attention mechanism with a one-dimensional residual network (ResNet18-1D) was proposed in this paper, aiming to improve the accuracy and efficiency of bloodstain recognition by hyperspectral imaging technology. The SENet ghannel attention mechanism automatically acquired the importance weight of each feature channel through learning, thereby Inhancing effective features and suppressing irrelevant ones. In view of the complexity of the trace-bearing object, this paper improved the traditional SENet module and adopts a dual-branch bottleneck module to enhance the applicability of the model. To address the complex and dynamic nature of forensic practice, this paper conducted two sets of experiments on the public blood detection dataset, which contains multiple traceable objects. (1) Hyperspectral Transductive Classification Scenario. Both raining and test sets were derived from the same HSI image, This experiment focused on analyzing substrate interference with bloodstain spectral features, Results show the model achieved an overall accuracy (OA) of 96.8% and an average accuracy (AA) of 97.6% in the complex simulated scenario, representing improvements of 1.3% and 1.9%, respectively, compared to the tate-of-the-art Hybrid CNN model. (2) Hyperspectral Inductive Classification Scenario, The model trained on the baseline Acenario was directly transferred to test on a different image. This experiment focused on the pre-identification capability for bloodstains and blood-like substances, presenting greater challenges but better reflecting real-world application needs. Experiments showed that the overall accuracy and average accuracy of the model were 63.3% and 65% respectively, which were 1.2% and 1.6% higher than the current optimal RNN model. Through error source analysis, it was found that tomato juice, due to its absorption peak near 470 nm being similar to the characteristic absorption peak of blood at 415 nm, has become the main interference source, In addition to making horizontal comparisons of different algorithms, this paper also verified the impact of the SENet channel attention mechanism module on model performance through ablation experiments. The results showed that The improved SENet channel attention mechanism module, compared with the original SENet channel attention mechanism module, had enhanced the overall and average accuracy of the model in both classification scenarios. Meanwhile, the efficiency jest showed that, despite having a large number of parameters, the collaborative design of the residual structure and the dual-branch SENet significantly reduced the computational cost. The training time is only 45 ms epoch, meeting the efficiency Jequirements of practical combat.
Soil organic matter content in the plow layer is a key indicator for evaluating soil quality. It not only provides crops with abundant nutrients but also improves the soil environment in the plow layer, making it an essential component of the plow layer. This study proposes a spectral data mining algorithm to enhance the sensitivity of spectral data to soil organic matter Content and improve its estimation capability. The study first employed discrete wavelet algorithms to sequentially perform Jeparation, correlation analysis, and model construction on soil spectral data, thereby establishing a model for estimating soil organic matter content, Subsequently, coupled algorithms were used to sequentially perform data mining, correlation analysis, and model construction on soil spectral data, with evaluation metrics used to assess the accuracy of the resulting model. Finally. 1 the sensitivity and estimation capability of spectral data toward soil organic matter content were compared before and after Coupling. The research results indicate: (1) The spectral information mining algorithm proposed in this study can integrate the advantages of various wavelet bases, significantly enhancing the sensitivity of the spectra to soil organic matter content, with the gorrelation coefficient increasing by an average of 15.33%. (2) A comparison of model accuracy before and after coupling Indicates that the spectral information mining algorithm proposed in this study can significantly enhance the estimation capability. If spectra for soil organic matter content and reduce estimation errors. The conclusions of this study can support the mining and analysis of spectral data across different locations and serve as a reference for the development of related algorithms.
