Search

The manipulation of vector optical fields in three-dimensional (3D) space plays a vital role in both fundamental research and practical implementations of polarization optics. However, existing studies mostly focus on 3D vector optical fields with limited modes. Here, an approach of spin-decoupled spatial partitioning is proposed to generate complex 3D vector optical fields with a customizable number of modes on demand. The crosstalk among different modes is effectively suppressed by the decoupling capability of asymmetric photonic spin–orbit interactions (PSOIs) and the design of region displacement for opposite spin states. As a proof-of-concept demonstration, a metasurface is designed to generate longitudinally varying high-order cylindrical vector fields, ranging from the 2nd to the 10th order in even sequences along the propagation direction. The experimental results demonstrate the effectiveness and potential of our approach to enabling precise control of 3D vector optical fields with arbitrary mode combinations. This work holds promising applications in biophotonics, quantum optics, and communications.

To enhance the efficiency and sustainability, technical preparations were made for eliminating the Temperature, Sample, Oxygen test of basic oxygen furnace (BOF) steelmaking process in this work. Utilizing data from 13,528 heats and state-of-the-art (SOTA) machine learning (ML) and deep learning algorithms, data-driven models with different types of inputs were developed, marking the first use of time series data (off-gas profiles and blowing practice related curves) for BOF steelmaking’s endpoint prediction, and the tabular features were expanded to 45. The prediction targets are molten steel’s concentrations of phosphorus (Endpoint [P], %) and carbon (Endpoint [C], %), and temperature (Endpoint-Temp, °C). The optimal models for each target were implemented at a Hesteel Group’s BOF steelmaking facility. Initially, SOTA ML models (XGBoost, LightGBM, Catboost, TabNet) were employed to predict Endpoint [P]/[C]/Temp with tabular data. The best mean absolute errors (MAE) achieved were 2.276 × 10−3% (Catboost), 6.916 × 10−3% (Catboost), and 7.955°C (LightGBM), respectively, which surpassed the conventional models’ performance. The prediction MAEs of the conventional models with the same inputs for Endpoint [P]/[C]/Temp were 3.158 × 10−3%, 7.534 × 10−3%, and 9.150°C (Back Propagation neural network) and 2.710 × 10−3%, 7.316 × 10−3%, and 8.310°C (Support Vector Regression). Subsequently, predictions were explored to be made using SOTA time series analysis models (1D ResCNN, TCN, OmniScaleCNN, eXplainable Convolutional neural network (XCM), Time-Series Transformer, LSTM-FCN, D-linear) with the original time series data and SOTA image analysis models (Pre-activation ResNet, DenseNet, DLA, Dual path networks (DPN), GoogleNet, Vision Transformer) with resized time series data. Finally, the concat-model and the paral-model architectures were designed for making predictions with both tabular data and time series data. It was determined that the concat-Model with TCN and ResCNN as the backbone exhibited the highest accuracy. It’s MAE for predicting Endpoint [P]/[C]/Temp reaches 2.153 × 10−3%, 6.413 × 10−3%, and 5.780°C, respectively, with field test’s MAE at 2.394 × 10−3%, 6.231 × 10−3%, and 7.679°C. Detailed results of the importance analysis for tabular data and time series are provided.

Non-line-of-sight (NLOS) imaging aims at recovering hidden objects located beyond the traditional line of sight, with potential applications in areas such as security monitoring, search and rescue, and autonomous driving. Conventionally, NLOS imaging requires raster scanning of laser pulses and collecting the reflected photons from a relay wall. High-time-resolution detectors obtain the flight time of photons undergoing multiple scattering for image reconstruction. Expanding the scanning area while maintaining the sampling rate is an effective method to enhance the resolution of NLOS imaging, where an angle magnification system is commonly adopted. Compared to traditional optical components, planar optical elements such as liquid crystal, offer the advantages of high efficiency, lightweight, low cost, and ease of processing. By introducing liquid crystal with angle magnification capabilities into the NLOS imaging system, we successfully designed a large field-of-view high-resolution system for a wide scanning area and high-quality image reconstruction. Furthermore, in order to reduce the long data acquisition time, a sparse scanning method capitalizing on the correlation between measurement data to reduce the number of sampling points is thus proposed. Both the simulation and experiment results demonstrate a >20 % reduction in data acquisition time while maintaining the exact resolution.

Vibration protection is a reliable mean for safe operation of rotating machinery. In recent years, vibration protection of pumped storage unit gets more and more complicated, resulting in an increasing risk of vibration protection system failure. Focusing on vibration characteristics of pumped-storage units, the inverse-time vibration protection method for of pumped-storage unit vibration was proposed based on material fatigue damage curve and linear damage accumulation theory. Difficulties of determining damage level were provided on applying damage accumulation method to vibration protection. Simplified inverse-time protection method for vibration of pumped-storage unit was proposed, and research showed that the inverse-time vibration protection threshold should be set according to vibration accumulation that avoids all unit transient processes, the reliability coefficient can be determined from 1.5 to 2.0, the refresh periods of unit vibration characteristic values do not affect the cumulative calculation results, and the inverse-time vibration protection has characteristic of self-returning. The vibration protection method based on inverse-time has the advantages of reliable principle, simple logic, and easy implementation. It has the characteristics of reliability, selectivity, sensitivity, and rapidity as relay protection in electric engineering. It solves the problem of dual parameter coordination in the traditional protection strategy with fixed threshold and fixed time delay. The method can be extended to the vibration protection of other types of rotating machinery.

[Abstract]: [Objective]: To construct a Nomogram model for the prediction of essential hypertension (EH) risks with the use of traditional Chinese medicine (TCM) syndrome elements principles in conjunction with cutting-edge biochemical detection technologies. [Methods]: A case-control study was conducted, involving 301 patients with essential hypertension in the hypertensive group and 314 without in the control group. Comprehensive data, including the information on the four TCM diagnoses, general data, and blood biochemical indicators of participants in both groups, were collected separately for analysis. The differentiation principles of syndrome elements were used to discern the location and nature of hypertension. One-way analysis was carried out to screen for potential risk factors of the disease. Least absolute shrinkage and selection operator (LASSO) regression was used to identify factors that contribute significantly to the model, and eliminate possible collinearity problems. At last, multivariate logistic regression analysis was used to both screen and quantify independent risk factors essential for the prediction model. The “rms” package in the R Studio was used to construct the Nomogram model, creating line segments of varying lengths based on the contribution of each risk factor to aid in the prediction of risks of hypertension. For internal model validation, the Bootstrap program package was utilized to perform 1000 repetitions of sampling and generate calibration curves. [Results]: The results of the multivariate logistic regression analysis revealed that the risk factors of EH included age, heart rate (HR), waist-to-hip ratio (WHR), uric acid (UA) levels, family medical history, sleep patterns (early awakening and light sleep), water intake, and psychological traits (depression and anger). Additionally, TCM syndrome elements such as phlegm, Yin deficiency, and Yang hyperactivity contributed to the risk of EH onset as well. TCM syndrome elements liver, spleen, and kidney were also considered the risk factors of EH. Next, the Nomogram model was constructed using the aforementioned 14 risk predictors, with an area under the curve (AUC) of 0.868 and a 95% confidence interval (CI) ranging from 0.840 to 0.895. The diagnostic sensitivity and specificity were found to be 80.7% and 85.0%, respectively. Internal validation confirmed the model’s robust predictive performance, with a consistency index (C-index) of 0.879, underscoring the model’s strong predictive ability. [Conclusion]: By integrating TCM syndrome elements, the Nomogram model has realized the objective, qualitative, and quantitative selection of early warning factors for developing EH, resulting in the creation of a more comprehensive and precise prediction model for EH risks.

Objective: This study aimed to establish the methods for determining the content of sodium hyaluronate in solid drinks by reversed-phase high performance liquid chromatography. Methods: With sodium hyaluronate content as evaluation index, after the sample extract was precipitated by ethanol, hyaluronase was used to hydrolyze the precipitate. The optimal enzymolysis conditions obtained by single factor test and orthogonal test were as follows: enzymolysis, at 34 ℃ and pH 6.5, for 120 min. The determination was performed on Uliumate C18 (4.6 mm×250 mm,5 μm) column with mobile phase: acetonitrile 10 mmol/L sodium octane sulfonate solution (Vacetonitrile∶Voctane sulfonate =10∶90) at flow rate of 0.6 mL/min. The detection wavelength was 232 nm and quantified by external standard method. Results: The correlation coefficient of sodium hyaluronate was 0.999 9. The recoveries of samples were between 96.1%~101%, and the relative standard deviation was 3.16%. The determination of limit was 0.062 g/kg. Conclusion: Ethanol precipitation combined with enzymolysis-reversed phase chromatography has strong specificity, high accuracy, good repeatability and less interference, which is suitable for the determination of sodium hyaluronate in solid drinks.

Objective: To evaluate the uncertainty of the determination of thiocyanate in cheese by inhibition conductance ion chromatography. Methods: The content of thiocyanate in cheese was determined by inhibition conductance ion chromatography and a mathematical model was established. The source of uncertainty introduced in the process of the experiment was described, and the synthetic uncertainty was calculated for each uncertainty component. Results: The determination of thiocyanate was （10.2±0.56） mg/kg, k=2, p=95%. Conclusion: The uncertainty in the measurement process mainly comes from the concentration of thiocyanate in the test solution, followed by sample recovery and repeatability experiments.

Longitudinal optical field modulation is of critical importance in a wide range of applications, including optical imaging, spectroscopy, and optical manipulation. However, it remains a considerable challenge to realize a uniformly distributed light field with extended depth-of-focus. Here, a high-efficiency extended depth-of-focus metalens is proposed by adjoint-based topology-shape optimization approach, wherein the theoretical electric field intensity corresponding to a variable focal-length phase is utilized as the figure of merit. Using a dozen of metalens with random structure parameters as initial structures, the average focal depth of topology-shape optimized metalens is greatly improved up to 18.80 μm (about 29.7λ), which is 1.54 times higher than the diffraction-limited focal depth. Moreover, all the topology-shape optimized metalens exhibit high diffraction efficiency exceeding 0.7 over the whole focal depth range, which is approximately three times greater than that of the forward design. Our results offer a new insight into the design of extended depth-of-focus metalens and may find potential applications in imaging, holography, and optical fabrication.

Abstract The development of nitrogen fertilizer green and efficient application technology by exploring the mechanism of efficient sorghum N use is important for sustainable development of sorghum industry as well as barren marginal land development and utilization. This study was conducted in 2018, 2019, and 2020 at Shenyang, China, using the nitrogen-efficient sorghum variety Liaonian No. 3 as material. The correlation between soil microbial species, diversity, and metabolic pathways with photosynthetic parameters and yield traits was analyzed to elucidate the mechanisms of nitrogen utilization and photosynthetic material production in sorghum under four fertilizer application patterns. The results showed that 17 populations of soil inter-root microorganisms were active in the organic fertilizer + 0 kg per hm2 of nitrogen (N0Y) model, and the abundance of two key populations, Comamonadaceae and Ellin5301, was significantly increased. Soil microorganisms regulated sorghum growth mainly through 30 pathways, focus including ko00540, ko00471, ko00072 and ko00550, of which ko02030 (Bacterial chemotaxis) and ko00072 (Synthesis and degradation of ketone bodies) played the most critical role. The functional analysis of soil microbial populations revealed that N0Y fertilizer model significantly reduced the intracellular trafficking, secretion. In addition, vesicular transport of microorganisms, amino acid transport and metabolism and nucleotide transport and metabolism played a key role in the regulation of population function. Overall, the N0Y model of N-efficient sorghum can achieve high levels of photosynthetic material production and higher yield formation through regulation of population activities and metabolic pathways of loamy microorganisms, resulting in reduced chemical N application and efficient green production of sorghum.

Metasurfaces based on the Pancharatnam–Berry (PB) phase have attracted significant attention in the domains of subwavelength optics and electromagnetics. Conventional theory predicts that the PB phase is exactly twice the rotation angle of the anisotropic meta-atoms. Differently, a recent advance has demonstrated that the generalized PB phase representing multiple times of the rotation angle could be obtained with high-fold rotational symmetry meta-atoms, but it suffers from the low cross-polarization conversion efficiency (the theoretical upper limit of 25%) that impedes its further applications, especially for meta-atoms with rotational symmetry ≥3. Here, we verify that the chiral meta-atoms with high-fold rotational symmetries could produce the generalized PB phase. Besides, the all-metallic configuration is utilized to design C2, C3, and C5 chiral catenary meta-atoms to improve their efficiency and bandwidth. The equivalent air waveguide with low loss between two adjacent meta-atoms is formed to analyse the higher performances of the all-metallic scheme for the realization of the generalized PB phase compared with the metal–insulator–metal and all-dielectric C3 meta-atoms. As a proof of concept, four metadevices including two spin-Hall metadevices and two holograms are experimentally demonstrated and their maximum efficiency could exceed 83% in simulation. This work could provide a high-efficiency platform for the study of the generalized PB phase in linear and nonlinear optics.

Abstract Background Pseudomembranous necrotizing laryngotracheobronchitis refers to an acute diffuse necrotizing inflammation in the mucosa of the larynx, trachea, and bronchus. It often occurs in infants and children having viral infections secondary to bacterial infections. Mycoplasma pneumoniae (M. pneumoniae) is a common pathogen that causes pneumonia in children. In recent years, serious complications due to M. pneumoniae infection, including necrotizing pneumonia, pulmonary embolism, and pleural effusion, have been increasingly reported. Case presentation An 11-year-old girl was admitted to our unit with cough, fever, and hoarseness persistent for a week. The results of the M. pneumoniae serological test, PCR examination with bronchial aspirate and bronchoalveolar lavage fluid (BALF), next-generation sequencing (mNGS) for BALF, all suggested the presence of M. pneumoniae infection. High-resolution CT scanning of the chest showed inflammation of the middle and lower lobes of the right lung. By bronchoscopy, the necrosis of the vocal cords, trachea, and bronchial mucosa was observed; each bronchial lumen contained a large amount of white viscous sputum. Pathological findings for bronchial mucosa suggested inflammatory necrosis. After administration of azithromycin and glucocorticoids, the symptoms of the patients were ameliorated. After 2 weeks post-discharge, the X-ray scan of her chest indicated the pneumonia resolution in the right lung. Conclusions In patients with pneumonia due to M. pneumoniae infection, which causes obvious hoarseness, bronchoscopy is necessary even if the lung lesions are not massively consolidated. When necrotizing lesions of the larynx, trachea, and bronchi are detected by bronchoscopy, the necrotic tissues in the corresponding parts should be conducted tissue biopsy for pathological examination. Apart from macrolide antibiotics, the administration of small doses of glucocorticoids is necessary.

Optical imaging systems, like microscopes, cameras, and telescopes, continue to expand the scope of human observation of the world. As one of the key indicators of imaging systems, the field-of-view (FOV) is often limited by coma aberration. Expanding it generally relies on a combination of complex lenses, leading to a bulky and cumbersome system. Recently, the emergency of meta-optics provides an alternative to constructing compact and lightweight large-FOV metalens through elaborated phase modulation within a flat surface, showing great potential in surveillance, unmanned vehicles, onboard planes or satellites, medical science, and other new applications. In this article, we review recent advances of wide-angle metalenses, including operation principles, design strategies, and application demos. Firstly, basic principles of wide-angle imaging using a single metalens are interpreted. Secondly, some advanced methods for designing subwavelength structures with high angle robustness and high efficiency are discussed. Thirdly, some representative functional devices and applications are surveyed. Finally, we conclude with an outlook on future potentials and challenges that need to be overcome.

In order to study the mechanical behavior characteristics of dynamic crack initiation and propagation of surrounding rock of adjacent roadway under impact load, with the help of a new digital laser dynamic caustics experimental system, a rock-like medium material PMMA was used to make a double-section roadway model for impact tests, and the crack propagation simulation was carried out based on the ABAQUS numerical simulation platform. The results show that when the pre-crack deflection angle α of adjacent sections is 45°, 22. 5°, 0°, -22. 5°, -45°, the dynamic stress intensity factor are different when the cracks were initiated. The dynamic stress intensity factor change after crack initiation has a "platform section" with a duration of about 160 μs. When the right crack gradually moves away from the left cross-section, the crack propagation speed of left crack after initiation will increase. When the dynamic crack initiation and propagation of the double-section roadway, the adjacent sections will affect the dynamic stress intensity factor and propagation speed to a certain extent. When the tip of the right crack is close to the left crack, the latter's crack initiation toughness is reduced, that is, it is easier to crack.

The main cause of steroid-induced necrosis of femoral head (SNFH) is excessive glucocorticoid (GC) intake. The aim of this article was to investigate the role of lncRNA NEAT1 as a molecular sponge to adsorb miR-23b-3p and regulate CYP1A2 in SNFH. Fluorescence in situ hybridization was used to localize lncRNA NEAT1. Human bone marrow mesenchymal stem cells (hBMSCs) were collected from patients with SNFH. The expression of lncRNA NEAT1, miR-23b-3p and CYP1A2 in hBMSCs were intervened. Compared to the control group, the lncRNA NEAT1 and CYP1A2 expression in the SNFH group was increased, while miR-23b-3p expression was decreased. GCs could inhibit the osteogenic differentiation of hBMSCs and upregulate the expression of lncRNA NEAT1. Knockdown of lncRNA NEAT1 could promote the proliferation and osteogenic differentiation of hBMSCs in the SNFH group. Overexpression of miR-23b-3p could partially counteract the effect of lncRNA NEAT1 on hBMSCs. CYP1A2 was confirmed to be a target of miR-23b-3p. Overexpression of CYP1A2 could partially rescue the effect of miR-23b-3p overexpression on hBMSCs. In conclusion, lncRNA NEAT1 as a ceRNA can adsorb miR-23b-3p and promote the expression of CYP1A2, which then inhibits the osteogenic differentiation of hBMSCs and promotes the progress of SNFH.

Bladder cancer is a common malignancy in the world. It is reported that circular RNA VANGL1 (circ_VANGL1) was involved in bladder cancer progression. However, the functional role and molecular mechanism of circ_VANGL1 in bladder cancer were still unclear.

In this paper, a new method is developed to investigate the anti-penetration performance of the steel/nylon sandwich plate, which incorporates experimental method with numerical simulation. Constitutive parameters of the steel/nylon sandwich plate are measured by the Hopkinson Pressure Bar, and LS-DYNA is applied to simulate the penetration process of projectiles into the sandwich plate. The dynamic response, failure modes and energy absorption coefficients are obtained, and the penetration tests are conducted to verify the numerical method. Results show: results obtained by experimental and numerical methods have very high agreement, which demonstrates that the method proposed can be applicable for predicting the anti-penetration performance of the sandwich plate.

The relevant literature on environmental strategy has carried out extensive research on the internal and external factors that affect the sustainable development of enterprises, but some important horizontal factors, such as the influence from peers, are ignored. Based on benchmarking theory and expectation theory, this paper makes an empirical analysis on the panel data of Listed Companies in 24 industries in China from 2001 to 2019. The results show that the number of Listed Companies in the industry has a positive impact on the passive and active environmental strategies of the whole industry. The outstanding performance of peers promotes the whole industry to actively implement the active environmental strategy and reduce the passive environmental strategy. At the same time, the longer the listing time of the listed companies, the more able the companies to strengthen the impact of the number of listed companies on the implementation of active or passive environmental strategy in the whole industry, and the smaller the relative scale, the more able the companies to strengthen the impact of performance on the implementation of active environmental strategy in the whole industry.

Spatial differentiator is the key element for edge detection, which is indispensable in image processing, computer vision involving image recognition, image restoration, image compression, and so on. Spatial differentiators based on metasurfaces are simpler and more compact compared with traditional bulky optical analog differentiators. However, most of them still rely on complex optical systems, leading to the degraded compactness and efficiency of the edge detection systems. To further reduce the complexity of the edge detection system, a monolithic metasurface spatial differentiator is demonstrated based on asymmetric photonic spin-orbit interactions. Edge detection can be accomplished via such a monolithic metasurface using the polarization degree. Experimental results show that the designed monolithic spatial differentiator works in a broadband range. Moreover, 2D edge detection is experimentally demonstrated by the proposed monolithic metasurface. The proposed design can be applied at visible and near-infrared wavelengths by proper dielectric materials and designs. We envision this approach may find potential applications in optical analog computing on compact optical platforms.

Self-accelerating beams show considerable captivating phenomena and applications owing to their transverse acceleration, diffraction-free and self-healing properties in free space. Metasurfaces consisting of dielectric or metallic subwavelength structures attract enormous attention to acquire self-accelerating beams, owing to their extraordinary capabilities in the arbitrary control of electromagnetic waves. However, because the self-accelerating beam generator possesses a large phase gradient, traditional discrete metasurfaces suffer from insufficient phase sampling, leading to a low efficiency and narrow spectral band. To overcome this limitation, a versatile platform of catenary-inspired dielectric metasurfaces is proposed to endow arbitrary continuous wavefronts. A high diffraction efficiency approaching 100% is obtained in a wide spectral range from 9 to 13 μm. As a proof-of-concept demonstration, the broadband, high-efficiency and high-quality self-accelerating beam generation is experimentally verified in the infrared band. Furthermore, the chiral response of the proposed metasurfaces enables the spin-controlled beam acceleration. Considering these superior performances, this design methodology may find wide applications in particle manipulation, high-resolution imaging, optical vortex generation, and so forth.

[Objectives] In order to improve the anti-explosion performance of ship structures subjected to multiple underwater explosion loading, and obtain a numerical method for simulating multiple underwater explosions is proposed.[Methods] The dynamic response of the 45# clamped air-back steel plate with diameter of 600 mm and thickness of 5 mm subjected to multiple underwater explosion loading was studied by a method of combining experiment and numerical simulation. The plastic deformation history of the steel plate was obtained and the basic law of plastic deformation was analyzed.[Results] The results show that the morphology produced by plastic deformation of the 45# clamped air-back steel plate subjected to multiple underwater explosion loading is spherical-like,and the deformation mode is mainly bending deformation and biaxial tensile deformation. When the amount of explosive and the explosion distance are fixed,the deflection increment of the center of the steel plate will be decreased gradually as the number of underwater explosion loading increased,and after the second explosion loading,the deflection increment will be third-order of the first explosion loading. When the total amount of explosive is constant,compared with the results of multiple uniform underwater explosion loading,both the deflection and thickness reduction rate of the steel plate after single underwater explosion loading are larger,and the maximum difference of the center deflection is 20.25%.[Conclusions] The experimental results agree well with the simulation results,which can provide certain reference for ship protection design.

Geometric metasurfaces have shown superior phase control capacity owing to the geometric nature of their phase profile. The existing geometric metasurfaces are generally composed of metal-dielectric composites or all-dielectric subwavelength structures. Here, a novel configuration, all-metallic structure, is proposed to achieve broadband and high-performance electromagnetic wavefront manipulation based on the geometric phase. A catenary model is built to describe the optical dispersion and guide the design of metasurfaces. Two metadevices including a beam deflector and a hologram are designed and experimentally demonstrated in the infrared regime, with the measured optical efficiency up to 84% (the simulated efficiency reaches 93%). Compared to previous metal-insulator-metal structures, this approach can realize higher efficiency and broader operating bandwidth owing to its lower ohmic loss. This design strategy is universal and can be easily scaled to any other spectra without complex optimization. Moreover, since metals have excellent mechanical and physical properties, such as good thermal and electrical conductivity, this all-metallic structure may provide a new thinking on interdisciplinary research.

Terahertz (THz) absorbers have attracted considerable attention due to their potential applications in high-resolution imaging systems, sensing, and imaging. However, the limited bandwidth of THz absorbers limits their further applications. Recently, the dispersion management of metasurfaces has become a simple strategy for the bandwidth extension of THz devices. In this paper, we used the capability of dispersion management to extend the bandwidth of THz absorbers. As a proof-of-concept, a dual metasurface-based reflective device was proposed for broadband near-unity THz absorber, which was composed of two polarization-independent metasurfaces separated from a metallic ground by dielectric layers with different thickness. Benefiting from the fully released dispersion management ability in adjusting the dimensions of the metasurfaces, we obtained an absorbance above 90% in the frequency range from 0.52 to 4.4 THz and the total thickness for the bandwidth approaching the theoretical Rozanov limit. The experimental results verified the ability of dispersion management in designing broadband absorbers and the performance of the designed absorber. The underlying physical mechanism of dispersion management was interpreted in the general equivalent circuit theory and transmission line model. In addition, the catenary optical model was used to further interpret the physics behind this dual metasurface. Moreover, we found that the alignment deviations between the dual metasurface had little impact on the performance of the designed absorber, which indicates that the dual-metasurface does not require center alignment and is easy to be fabricated. The results of this work could broaden the application areas of THz absorbers.

Melt density is a crucial quality indicator for polymer composites, yet real-time measurement remains challenging due to processing complexities. While existing machine learning methods offer solutions, they often fall short in complex compounding scenarios. This study presents a novel multi-source data-driven approach for measuring melt density in polycarbonate/acrylonitrile butadiene styrene blends. By incorporating ultrasonic, near-infrared, and Raman spectra data acquired during melt processing, a deep separable convolutional neural network model is developed to predict melt density accurately. The model effectively fuses multi-source data to establish the mapping relationship between input data and melt density output. Results demonstrate the model's ability to monitor melt density in real-time, achieving a prediction accuracy with RMSE and [R.sup.2] indexes of 0.[infinity]5 g/[cm.sup.3] and 0.9841, respectively. The proposed approach outperforms existing methods, showcasing its effectiveness and superiority in melt density prediction for polymer compounding processes. Highlights * Establishment of the real-time monitoring system for polymer extrusion processes. * Conversion of multi-sensor signals into time-frequency images using wavelet decomposition. * Fusion of sensor data into a three-channel tensor-image. * Development of a data-driven DSCNN model for predicting melt density. * Implementation for online monitoring and prediction in PC/ABS compounding system. KEYWORDS data fusion, data-driven, density measurement, DSCNN, extrusion processing, in-line monitoring, polymer melt, 1 | INTRODUCTION Extrusion processing involves the use of specialized machinery to transform polymer materials into a variety of shapes and forms. This stage of fabrication is critical for producing [...]

Along with the development of advanced high-performance aero-engines to the higher thrust-weight ratio, further improvement of stage load, the adoption of new materials and new lightweight structures, the aeroelasticity of blade structure is becoming more and more prominent. The high cycle fatigue failure of blades significantly reduces the structural reliability during the process of development and using. At the same time, a large number of failure forms of aero-engine experimental and server can be attributed to aeroelastic problems. Therefore, it is urgent to improve the aeroelastic stability of the blade. One of the most important factors is to suppress the airflow separation, but its mechanism is still unclear. Based on this, this paper combines the aerodynamic damping analysis of energy method with the plasma excitation simulation and references low-speed wind tunnel plasma expansion test to consider the effects of different exciter distributions and intensities on flutter. The results show that stall flutter is related to the flow separation, but the flow separation is not a key factor that determinates whether the flutters occurs or not. Flutter suppression is strongly correlated with the shock wave intensity, amplitude of first harmonic aerodynamic force, low-speed separation and aerodynamic work density. In addition, the relative distribution of the excitation field and the positive work zone also has a direct effect on the suppression of flutter.