DPAL-BERT:A Faster and Lighter Question Answering Model

Recent advancements in natural language processing have given rise to numerous pre-training language models in question-answering systems.However,with the constant evolution of algorithms,data,and computing power,the increasing size and complexity of these models have led to increased training costs and reduced efficiency.This study aims to minimize the inference time of such models while maintaining computational performance.It also proposes a novel Distillation model for PAL-BERT(DPAL-BERT),specifically,employs knowledge distillation,using the PAL-BERT model as the teacher model to train two student models:DPAL-BERT-Bi and DPAL-BERTC.This research enhances the dataset through techniques such as masking,replacement,and n-gram sampling to optimize knowledge transfer.The experimental results showed that the distilled models greatly outperform models trained from scratch.In addition,although the distilled models exhibit a slight decrease in performance compared to PAL-BERT,they significantly reduce inference time to just 0.25%of the original.This demonstrates the effectiveness of the proposed approach in balancing model performance and efficiency.

Simulated Reflectance of Apple Trees in Canopy Level Based on the PROSAIL Model and HJ-1A-HSI Data

Using the PROSAIL radiation transfer model and HJ-1A-HSI data to simulate the canopy reflectivity of apple trees, this study lays the foundation for the inversion of canopy parameters. Taking Qixia City of Yantai City, Shandong Province as the research area, the apple tree was taken as the research object, and the hyperspectral reflectance, LAI and sample GPS of apple canopy were measured in the field. The parameters required for the PROSAIL model were obtained by experimental methods. The model simulates the reflectivity;the HSI image data is preprocessed, and the canopy reflectivity is extracted by GPS coordinates. The PROSAIL model and the HSI image simulated reflectance were fitted to the measured apple canopy reflectivity. The decisive factor (R2) of the simulated reflectance and the measured reflectance of the PROSAIL model was 0.9944, and the relative error (RE%)was 0.1845. The HSI data simulated reflectance and measured reflectance. The coefficient of determination is 0.9714 and the relative error is 0.6202. Both have achieved good fitting effects and can be used for inversion studies of apple canopy parameters.

Quantitative Relationship between Leaf Area Index and Canopy Reflectance Spectra of Rice under Different Nitrogen Levels

Monitoring rice growth by spectral remote sensing technology can provide scientific basis for the high yield and efficient production of rice. Field experiments with different nitrogen application amounts using Tianyouhuazhan rice as test sam- ples were set up to study the relationship between rice leaf area index （LAI） and canopy reflectance spectral. The results showed that： the LAI increased with the amount of applied nitrogen; the canopy reflectance spectral showed significant re- sponse characteristics to groups with different nitrogen application levels; the corre- lation coefficient of LAI and canopy spectral reflectance reached the maximum at 720 nm red edge region. The mathematical model was constructed to predict the LAI according to the canopy reflectance spectra of rice.

Model of Reflection Spectra of Rock Surface in 2π Space

This paper deals with reflection spectra and polarized reflection spectra of 20 sorts of rock in 2π space, and then creates a model of reflection spectra of rock surface in 2π space. We measured the change of reflection and polarized reflection spectra as altering the incidence angle, vertex angle, azimuth angle, band and polarization. The results show that influence of the incidence angle on spectral curves is very strong. And when the vertex angle is constant, the horizontal azimuth polarizes rock spectra, and distorts the circular spectrum to become elliptic. The polarization influences the reflection intensity of rock spectra, but has no evident influence on the characteristics of wave forms of rock in 2π space. Therefore, we can describe the whole reflection spectral characteristics, including polarization, of rock surface in 2π space by measuring and calculating the e and p values in several key directions.

Estimation of As and Cu Contamination in Agricultural Soils Around a Mining Area by Reflectance Spectroscopy:A Case Study

Concentrations of Iron (Fe), As, and Cu in soil samples from the fields near the Baoshan Mine in Hunan Province, China, were analyzed and soil spectral reflectance was measured with an ASD FieldSpec FR spectroradiometer (Analytical Spectral Devices, Inc., USA) under laboratory condition. Partial least square regression (PLSR) models were constructed for predicting soil metal concentrations. The data pre-processing methods, first and second derivatives (FD and SD), baseline correction (BC), standard normal variate (SNV), multiplicative scatter correction (MSC), and continuum removal (CR), were used for the spectral reflectance data pretreatments. Then, the prediction results were evaluated by relative root mean square error (RRMSE) and coefficients of determination (R 2 ). According to the criteria of minimal RRMSE and maximal R 2 , the PLSR models with the FD pretreatment (RRMSE = 0.24, R 2 = 0.61), SNV pretreatment (RRMSE = 0.08, R 2 = 0.78), and BC-pretreatment (RRMSE = 0.20, R 2 = 0.41) were considered as the final models for predicting As, Fe, and Cu, respectively. Wavebands at around 460, 1 400, 1 900, and 2 200 nm were selected as important spectral variables to construct final models. In conclusion, concentrations of heavy metals in contaminated soils could be indirectly assessed by soil spectra according to the correlation between the spectrally featureless components and Fe; therefore, spectral reflectance would be an alternative tool for monitoring soil heavy metals contamination.

A Multi-slack Bus Model for Bi-directional Energy Flow Analysis of Integrated Power-gas Systems

The bi-directional energy conversion components such as gas-fired generators(GfG)and power-to-gas(P2G)have enhanced the interactions between power and gas systems.This paper focuses on the steady-state energy flow analysis of an integrated power-gas system(IPGS)with bi-directional energy conversion components.Considering the shortcomings of adjusting active power balance only by single GfG unit and the capacity limitation of slack bus,a multi-slack bus(MSB)model is proposed for integrated power-gas systems,by combining the advantages of bi-directional energy conversion components in adjusting active power.The components are modeled as participating units through iterative participation factors solved by the power sensitivity method,which embeds the effect of system conditions.On this basis,the impact of the mixed problem of multi-type gas supply sources(such as hydrogen and methane generated by P2G)on integrated system is considered,and the gas characteristics-specific gravity(SG)and gross calorific value(GCV)are modeled as state variables to obtain a more accurate operational results.Finally,a bi-directional energy flow solver with iterative SG,GCV and participation factors is developed to assess the steady-state equilibrium point of IPGS based on Newton-Raphson method.The applicability of proposed methodology is demonstrated by analyzing an integrated IEEE 14-bus power system and a Belgian 20-node gas system.

Determination of Protein and Starch Content in Whole Maize Kernel by Near Infrared Reflectance Spectroscopy

Using 128 bulk-kernel samples of inbred lines and hybrids, a study was conducted toinvestigate the feasibility and method of measuring protein and starch contents inintact seeds of maize by near infrared reflectance spectroscopy (NIRS). The chemometricalgorithms of partial least square (PLS) regression was used. The results indicated thatthe calibration models developed by the spectral data pretreatment of firstderivative+multivariate scattering correction within the spectral region of 10000-4000cm-1, and first derivative + straight line subtraction in 9000-4000cm-1 were thebest for protein and starch, respectively. All these models yielded coefficients ofdetermination of calibration (R2cal) above 0.97, while R2cv and R2val of cross and externalvalidation ranged from 0.92 to 0.95, respectively; however, the root of mean squareerrors of calibration, cross and external validation (RMSEE, RMSECV and RMSEP) werebelow 1(ranged 0.3-0.7),respectively. This study demonstrated that it is feasible touse NIRS as a rapid, accurate, and none-destructive technique to predict protein andstarch contents of whole kernel in the maize quality improvement program.

Reflection of PlaneWaves from Electro-magneto-thermoelastic Half-space with a Dual-Phase-Lag Model

The aim of this paper is to study the reflection of plane harmonic waves from a semi-infinite elastic solid under the effect of magnetic field in a vacuum.The expressions for the reflection coefficients,which are the relations of the amplitudes of the reflected waves to the amplitude of the incident waves,are obtained.Similarly,the reflection coefficient ratio variations with the angle of incident under different conditions are shown graphically.Comparisons are made with the results predicted by the dual-phase-lag model and Lord-Shulman theory in the presence and absence of magnetic field.

Analyzing Bandwidth on the Microwave Absorber by the Interface Reflection Model

Metallic flaky sendust particles are prepared for use as fillers in electromagnetic attenuation composites. We report the interface reflection model to divide the broad bandwidth into electromagnetic loss and quarter-wavelength （λ/4） cancelation. Combining with the face reflection calculation, we identify the electromagnetic loss originated from skin effect, which is used to explain over half of the absorbed energy in high frequency band. Most impor- tantly, the unique electromagnetic loss cannot generate the reflection loss （RL） peak. Using the phase relation of face reflection, we show evidence that the λ/4 cancelation is vital to generate the RL peak. The calculated energy loss agrees well with the experimental data and lays the foundation for further research.

Three-Component Model for Bidirectional Reflection Distribution Function of Thermal Coating Surfaces

We present a bidirectional reflection distribution function （BRDF） model for thermal coating surfaces based on a three-component reflection assumption, in which the specular reflection is given according to the microfacet theory and Snell＇s law, the multiple reflection is considered Nth cosine distributed, and the volume scattering is uniformly distributed in reflection angles according to the experimental results. This model describes the reflection characteristics of thermal coating surfaces more completely and reasonably. Simulation and measurement results of two thermal coating samples SR107 and S781 are given to validate that this three-component model significantly improves the modeling accuracy for thermal coating surfaces compared with the existing BRDF models.

Reflectance features of water stressed Larix gmelinii needles

In recent years, Xing＇an larch （Larix gmelinii） has been seriously infected by pests and drought. In order to improve the accuracy of monitoring the damage to larch by remote sensing （RS） and to predict the health of the larch, We studied fire reflectance features of larch needles under different water conditions at the needle level by using the LIBERTY （Leaf Incorporating Biochemistry Exhibiting Reflectance and Transmittance Yields） model. Before applying the LIBERTY model, we recalibrated it for the needles of L. gmelinii based on ten field-measured spectral curves. After recalibration, LIBERTY can accurately model the needle reflectance spectra of L. gmelinii. Based on the recalibrated LIBERTY model, we extracted and analyzed the sensitive bands to needle water content by simulating the needle reflectance spectra under different drought conditions. Then, we established mathematical equations between the spectral indices （MSI, NDWI, and GVMI） and needle water content. Results show that the variations of larch needle water content can significantly change the needle spectra at the near-infrared and short-wave infrared bands. The higher the water content is, the higher the absorption peak is. We believe that our study will provide the theoretical basis and an optional method to investigate the forest water stress using multi-spectral or hyper-spectral remote sensing data.

Analysis of reflected signal of quad rotor UAV based on model fitting in mobile communication system

In view of the many scenes of unmanned aerial vehicle(UAV)detection,a third-party signal source is used to design a receiver to monitor the UAV.It is of great significance to understand the reflection of the signal illuminating the UAV.Taking the communication base station(BS)signal as the third-party signal source,and considering the complete transmission link,reflection changes and loss fading of the communication signal,this study conducts model fitting for irregular UAV targets,simplifying complex targets into a combination of simple targets.Furthermore,the influence of the dielectric constant of the target surface and the signal irradiation angle on the signal reflection is analyzed.The analysis shows that the simulation results of this model fitting method are consistent with the results of other literature,which provides theoretical support for the detection of low and slow small targets such as UAVs.

Geographic Classification of Chinese Grape Wines by Near-Infrared Reflectance Spectroscopy

Near-infrared reflectance spectroscopy (NIRS) was applied to classify grape wines of different geographical origins (Changli, Huailai, and Yantai, China). Near infrared (NIR) spectra were collected in transmission mode in the wavelength range of 800-2500 nm. Wines (n=90) were randomly split into two sets, calibration set (n=54) and validation set (n=36). Discriminant analysis models were developed using BP neural network and discriminant partial least-squares discriminant analysis (PLS-DA). The prediction performance of calibration models in different wavelength range was also investigated. BP neural network models and PLS-DA models correctly classified 100% of the wines in calibration set. When used to predict wines in validation set, BP neural network models correctly classified 100%, 81.8%, and 90.9% of the wines from Changli, Huailai, and Yantai respectively, and PLS-DA models correctly classified 100% of all samples. The results demonstrated that NIRS could be used to discriminate Chinese grape wines as a rapid and reliable method.

Accuracy of the staggered-grid finite-difference method of the acoustic wave equation for marine seismic reflection modeling

Seismic wave modeling is a cornerstone of geophysical data acquisition, processing, and interpretation, for which finite-difference methods are often applied. In this paper, we extend the velocity- pressure formulation of the acoustic wave equation to marine seismic modeling using the staggered-grid finite-difference method. The scheme is developed using a fourth-order spatial and a second-order temporal operator. Then, we define a stability coefficient (SC) and calculate its maximum value under the stability condition. Based on the dispersion relationship, we conduct a detailed dispersion analysis for submarine sediments in terms of the phase and group velocity over a range of angles, stability coefficients, and orders. We also compare the numerical solution with the exact solution for a P-wave line source in a homogeneous submarine model. Additionally, the numerical results determined by a Marmousi2 model with a rugged seafloor indicate that this method is sufficient for modeling complex submarine structures.

Rapid, Non-Destructive, Textile Classification Using SIMCA on Diffuse Near-Infrared Reflectance Spectra

Soft independent modeling of class analogy (SIMCA) was successful in classifying a large library of 758 commercially available, non-blended samples of acetate, cotton, polyester, rayon, silk and wool 89% - 98% of the time at the 95% confidence level (p = 0.05 significance level). In the present study, cotton and silk had a 62% and 24% chance, respectively, of being classified with their own group and also with rayon. SIMCA correctly identified a counterfeit “silk” sample as polyester. When coupled with diffuse NIR reflectance spectroscopy and a large sample library, SIMCA shows considerable promise as a quick, non-destructive, multivariate method for fiber identification. A major advantage is simplicity. No sample pretreatment of any kind was required, and no adjust-ments were made for fiber origin, manufacturing process residues, topical finishes, weave pattern, or dye content. Increasing the sample library should make the models more robust and improve identification rates over those reported in this paper.

General Solutions of Reflection Equations for Bariev Model

The integrable general open-boundary conditions for the one-dimensional Bariev chain are considered. All kinds of solutions to the reflection equation （RE） and its dual are obtained.

3D model of Svecofennian Accretionary Orogen and Karelia Craton based on geology,reflection seismics,magnetotellurics and density modelling:Geodynamic speculations

A 3D model of deep crustal structure of the Archaean Karelia Craton and late Palaeoproterozoic Svecofennian Accretionary Orogen including the boundary zone is presented.The model is based on the combination of data from geological mapping and reflection seismic studies,along profiles 1-EU,4 B,FIRE-1-2 a-2 and FIRE-3-3 a,and uses results of magnetotelluric soundings in southern Finland and northern Karelia.A seismogeological model of the crust and crust-mantle boundary is compared with a model of subhorizontal velocity-density layering of the crust.The TTG-type crust of the Palaeoarchaean and Mesoarchaean microcontinents within the Karelia Craton and the Belomorian Province are separated by gently dipping greenstone belts,at least some of which are palaeosutures.The structure of the crust was determined mainly by Palaeoproterozoic tectonism in the intracontinental settings modified by a strong collisional compression at the end of the Palaeoproterozoic.New insights into structure,origin and evolution of the Svecofennian Orogen are provided.The accretionary complex is characterized by inclined tectonic layering:the tectonic sheets,~15 km thick,are composed of volcanic-sedimentary rocks,including electro-conductive graphite-bearing sedimentary rocks,and electro-resistive granitoids,which plunge monotonously and consecutively eastward.Upon reaching the level of the lower crust,the tectonic sheets of the accretionary complex lose their distinct outlines.In the seismic reflection pattern they are replaced by a uniform acoustically translucent medium,where separate sheets can only be traced fragmentarily.The crust-mantle boundary bears a diffuse character:the transition from crust to mantle is recorded by the disappearance of the vaguely drawn boundaries of the tectonic sheets and in the gradual transition of acoustically homogeneous and translucent lower crust into transparent mantle.Under the effect of endogenic heat flow,the accretionary complex underwent high-temperature metamorphism and partial melting.Blurring of the rock contacts,which in the initial state created contrasts of acoustic impedance,was caused by partial melting and mixing of melts.The 3 D model is used as a starting point for the evolutionary model of the Svecofennian Accretionary Orogen and for determination of its place in the history of the Palaeoproterozoic Lauro-Russian intracontinental orogeny,which encompassed a predominant part of the territory of Lauroscandia,a palaeocontinent combining North American and East European cratons.The model includes three stages in the evolution of the Lauro-Russian Orogen(-2.5,2.2-2.1 and 1.95-1.87 Ga).The main feature of the Palaeoproterozoic evolution of the accretionary Svecofennian Orogen and Lauroscandia as a whole lay in the causal link with evolution of a superplume,which initiated plate-tectonic events.The Svecofennian-Pre-Labradorian palaeo-ocean originated in the superplume axial zone;the accretionary orogens were formed along both continental margins due to closure of the palaeo-ocean.

The Effect of a Proposed Training Program Based on the Reflective Model on the Teachers＇ Reading Practices in Mafraq

This study aimed at investigating the effect of a proposed EFL （English as a Foreign Language） training program on the teachers＇ reading practices in Mafraq. This program was developed by the researcher according to the ninth-grade EFL teachers＇ needs. The participants of the study were 20 ninth-grade females EFL teachers who represented the experimental and control groups （10 teachers in each group）. They were randomly chosen from 10 schools in the Northwestern Badia Directorate of Education. To achieve the purposes of the study, the researcher designed a needs analysis questionnaire, a training program, a pre and post observation checklist, and a pre and post received knowledge test for teachers. Proper statistical analysis was used to analyze the results of the research questions. The results of the study showed that teachers were in need to be trained on 20 reading comprehension practices to be able to teach reading comprehension effectively. Besides, there were statistically significant differences at the level of （a = 0.05） between the means scores of the ninth-grade teachers＇ performance of the control and experimental groups on the post observation checklist and on the post received knowledge test in favor of the experimental group due to the training program.

Simplified Reflection Models of Representative Ground Metal Target in Millimeter Wave Passive Surveillance

For designing camouflage of ground metal target against millimeter wave(MMW) passive detections,based on the radiation and reflection theories related to MMW passive detection and an assumption of high frequency approximation,simplified models of two kinds of ground metal targets are established.Analysis indicates that the reflection from ground is the leading factor on a ground metal target's radiant temperature.In an 8-mm imaging experiment,the uneven distribution of radiant temperatures in horizontal and vertical metal plates is directly perceived.In a radiometer experiment,two positions of a radiometer device are utilized to obtain the radiant temperatures of a horizontal metal in two different grounds(made of asphalt and metal net respectively),which are 295.41 K and 301.55 K,respectively.The experimental results are close to the analytical results calculated from the proposed simplified models,which are 306 K and 296 K,respectively.According to the simplified models,different parts of a vertical plate receive incident MMW of different angles from the asphalt background.The radiant temperature of a vertical metal plate obtained from a radiometer experiment at night is 307 K,and it is in coherent with 304.22 K,the result from calculation using the simplified model of vertical plate.Both the analytic results and the experimental results agree with the uneven distribution of radiant temperature.

Reflections on Improvement of Utility Model System in China

Even since 1985 when the Patent Law was launched in China, the utility model patent has been playing a very important role. Over the two decades, the utility model system has played an active part in encouraging invention-creations, and promoting the progress and development of science and technology.