Search

Prediction of Total Electron Content (TEC) in the ionosphere is vital to improve the accuracy of satellite positioning, navigation and remote sensing systems. Most existing TEC prediction methods ignored the local variation patterns between various positions within the TEC sequence, resulting in limited prediction accuracy. To address this issue, this paper combined attention techniques, convolutional neural networks (CNN) and bidirectional long short-term memory networks (BiLSTM) to propose the Att-CNN-BiLSTM model. In the proposed model, CNN is used to extract positional features, BiLSTM is used to extract bidirectional temporal features, and attention technique is used to adaptively weight the features. This paper selected six locations in China, each with a six-year TEC sequence, including three years of high solar activity and three years of low solar activity. The paper first conducted ablation experiments, and the results showed that adding CNN and Attention can effectively improve prediction performance. Then, the proposed model was compared with LSTM and GRU. The experimental results show that compared with LSTM and GRU, the average RMSE of Att-CNN-BiLSTM in six regions decreased by 10.28% and 16.92% in high solar activity years, and by 11.82% and 8.92% in low solar activity years, respectively. The paper also conducted comparative experiments within one week of the magnetic storms, and the results showed that during the magnetic storm period, the RMSE of the proposed model decreased by 35.50% and 37.35% compared to LSTM and GRU. The R2 s of the proposed model are also higher than those of the comparison models in all cases.

Sag sections are freeway sections where the gradient varies significantly from top to bottom. Due to their special geometric characteristics, sag sections are often bottlenecks in motorway networks. However, we are still unclear about the underlying mechanisms of traffic congestion in the two-lane sag sections. Based on the actual lane change experience, a two-lane lattice traffic flow model at the sag sections considering the empirical lane change rate is proposed. The model is developed to reveal the evolutionary characteristics of traffic congestion at the sag section, and then to alleviate the traffic congestion problem at the sag section. In this paper, the empirical lane change rate is described as a function that depends on the road traffic density. First, the condition of traffic flow stability was obtained by analyzing the linear stability at sag sections when lane changes were allowed. The stability condition of traffic system show that lane change parameter and slope have important effects on the stability of traffic system. Then, the mKdV equation describing the phase transition of the traffic flow and its solution are obtained using nonlinear analysis. Finally, numerical simulations were performed. The results show that the increase of slope increases the traffic flow stability in uphill sections and the opposite in downhill sections. Moreover, the lane-change rate parameter can effectively increase the stability region and suppress traffic congestion.

Marine oil spills accidents has caused a large amount of crude oil to leak into the marine environment and threaten marine ecological environment. Hyperspectral remote sensing images (HRSI) technology can quickly and accurately identify oil film of different thickness on marine surface. In order to overcome the traditional spectrum analysis method and space extraction method of long time sampling, calculation, analysis and other shortcomings. On account of the advantages of the spectral and spatial information in the field of HRSI classification, a spectral-spatial features extraction (SSFE) method based convolutional neural networks (CNNs) was proposed to analyse oil spills. In this way, one and two dimensional models based on convolutional neural networks (1D-CNN,2D-CNN) have been introduced as the spectral and spatial features extractor. When extracting spatial features, double-two convolution layers are connected to increasing nonlinearity and reduce the number of parameters. Furthermore, in order to address overfitting and imbalance samples, L2 regularization, class_weight and drouput is added to classes data modeling. More importantly, principal component analysis (PCA) is applied to data dimension reduction, 1D-CNN and 2D-CNN is combined into a unified model further extract the joint spatial-spectral features. To evaluate the effectiveness of the proposed approach, three hyperspectral datasets were utilized, including: University of Pavia dataset, oil spill area 1, oil spill area 2. Experimental results reveal that the proposed method have a very satisfactory performance and better distinguish oil spills.

Triangular planar ring resonant cavities are widely used in cavity ring-down spectroscopes (CRDS) because of the unique optical path advantage. Hermite-Gaussian theory is the classical method used to describe the beam in the resonant cavity. Since this theory can express the transverse electromagnetic mode (TEM) in the meridian plane and sagittal surface separately in an astigmatic triangular cavity, it has a unique advantage in the analysis of triangular cavities. However, among the many analyses of triangular ring-down cavities, existing theories of Hermite-Gaussian modes have not been able to reasonably explain the many phenomena observed in alignment and ring-down experiments. In this paper, the mechanisms of the effects of first-order Hermite-Gaussian mode on mounting alignment in CRDS instruments are investigated in depth. The frequency separation characteristics of the lower-order mode in the triangular cavity are verified in the mounting alignment stage. The separation characteristics of first-order mode can guide the misalignment of the two orthogonal planes separately. The resonant frequency separation property of the Hermite-Gaussian mode in a triangular ring-down cavity can also be applied to more complex odd-numbered cavity mirror resonant systems.

In the design of two-degree-of-freedom (2-DOF) fast steering mirror (FSM) system, in order to improve the control bandwidth of the system, the low-order natural frequency in the working direction should be reduced as much as possible, and the high-order natural frequency in the non-working direction should be increased. In this paper, a deep-cut flexure hinge mirror system was studied. Firstly, the motion direction of the first to third order natural vibration mode of the system was analyzed. Next, the working stiffness in the third vibration mode direction was deduced to solve the problem that the traditional stiffness calculation method is not suitable for the third vibration mode direction. Then, the energy method and the Castigliano's second theorem were used to analyze the working stiffness of the deep-cut flexure hinge. Next, the formula for calculating the relationship between the thickness of the mirror and the moment of inertia in the direction of the vibration mode was derived. Combined with the calculation formula of working stiffness in the vibration mode direction, the formula of the first to third order natural frequencies was derived, and the finite element verification and sensitivity analysis of structural parameters were carried out. Finally, by using NSGA-II, a multi-objective optimization design was carried out on the first to third natural frequencies of the system with hinge structure parameters and reflector thickness as independent variables, and the optimization results were analyzed by theoretical calculation and finite element simulation verification. The results show that the first and second-order natural frequencies of fast steering mirror system are reduced by 7.8% and 7.11%, and the third-order natural frequencies are increased by 139.8%. It proves that the optimized structure is much better than the original structure, and the optimal calculation can effectively increase the control bandwidth of the system.

The ever-evolving mesoscopic scale optical imaging systems facilitate the new discoveries in the neuroscience research. They excel at providing a complete view of the long projections of a single neuron across the whole brain. Although the preparation protocols and the optical imaging systems are rapidly evolving, there are still some noise and artifacts interfering downstream data processing and analysis due to the defects in the imaging system or during the sample preparation. A specific denoising procedure is usually developed for one optical imaging system as a post-processing measure. However, the development of the optical imaging systems usually follows in an incremental manner. It would be better to adapt the denoising model to the new optical imaging system than training a denoising model from scratch. In this paper, the proposed learning schema and practice learn a new denoising model for a new optical imaging system based on the existing denoise models and bootstrap a denoising model without the ground-truth denoising labels. We achieve this through a CycleGAN based model and the fact that the optical imaging systems usually produce images both from the signal area and the fixture area. The experiments show that our proposed procedure can provide a comparable denoising performance against other state-of-the-art denoising methods for several optical neuroimaging datasets.

The steel cord conveyor belt surface is prone to damage in mining. The worn belt surface has acceleration characteristics, so timely and rapid repair is very necessary. To quickly and automatically repair the worn belt surface is a core design objective of the gluing robot (GR). Based on this objective, a new variant Traveling Salesman Problem (TSP) is put forward: after the worn segments are divided according to the worn information and GR’s workspace, path optimization of the gluing robot (POGR) problem is presented at a certain worn segment; then the POGR is simplified into a “double vertices” TSP problem by Hamilton graph, and the mathematical model is built. An improved genetic algorithm (IGA) is proposed to handle the POGR problem, which is called IGA-POGR. The main benefit of the proposed IGA-POGR is the ability to solve POGR of different scales in different ways. The performance of the IGA-POGR is illustrated on four well-known TSP problems. Numerical results show that IGA-POGR does not give any deviation (0%) from the optimal solution. Compared with discrete particle swarm optimization (DPSO), IGA-POGR has better performance in terms of the solving quality and time consumption when solving four idealized POGR problems.

Indoor mobile robots normally cannot capture the whole information of a scene by a single frame of perceptive data due to the limited sensor scope. The category of the current scene may be misjudged by robotics due to incomplete scene information, which leads to operation error. To address this problem, we propose an approach that leverages conditional random fields (CRFs) to fuse multiframe RGB and depth (RGB-D) visual data corresponding to the same scene. This method takes full advantage of prior knowledge that object categories significantly relate to the scene attributes. As a new image arrives, we incrementally integrate the current object detection results to update scene understanding by identifying duplicate objects between images, ranking available objects in terms of their relevance to the scene, and fusing new information with the existing CRF. With this approach, scene classification can be solved with higher precision based on multiview images than on single image frames sampled in the same places. Additionally, a configuration map of a scene is incrementally built into the above framework. The map includes identities of the recognized objects and various relations between them. This kind of map would not only benefit common robotic tasks but also offer a novel channel for human-robot interaction. We test the efficiency of our method on image sequences extracted from the NYU v2 dataset. The results show that our approach achieves the best performance against state-of-the-art baselines.

Performance degradation assessment (PDA) is of great significance to ensure safety and availability of mechanical equipment. As an important issue of PDA, the robustness of the trained model directly affects the assessment efficiency and restricts its application in practice. This paper proposes a robust modeling approach based on Student's t-hidden Markov model (Student's t-HMM) and nuisance attribute projection (NAP). NAP can remove nuisance attributes caused by individual differences from the feature space. Student's t-HMM utilizes the finite Student's t-mixture models (SMMs) to describe the observation emission densities associated with each hidden state, which can be more tolerant towards outliers than conventional HMMs. Based on these two techniques, the proposed method is supposed to be more robust and can assess the performance degradation process of new objects based on data of tested objects. The superiority of the proposed approach is evaluated using the vibration data from the accelerated life tests of bearings and the public XJTU-SY Bearing Datasets. The results demonstrate the robustness and effectiveness of the proposed approach for PDA modeling of rolling element bearings.

Considering that large quantities of soil hyperspectral data include the redundancy and overlap of spectral information, the technology of selecting feature wavelength can effectively solve these problems, and improve the accuracy and stability of the soil organic matter (SOM) content retrieval model. Traditional methods of wavelength selection mainly attempt to establish the empirical relationship between reflectance and SOM contents, and the performance is directly related to the quality and representativeness of the “training data”. This study first distinguished the sensitive wavelength interval of SOM through the sensitivity analysis (SA) of the SOM to soil reflectance in radiative transfer model. Then sensitive wavelength points of SOM were ascertained using the successive projection algorithm (SPA): 468, 476, 496, 599, 775 and 900nm. Results show that SOM content can be estimated with high accuracy (root-mean-square error of prediction (RMSEP) 2) > 82.9%) by adopting the selected six wavelengths. Especially at 599nm, the accuracy of SOM content estimation is the highest (RMSEP: 0.176%, R2: 90.4%). Compared with traditional empirical wavelength selection methods, the wavelength selection based on the SA-SPA with the SOM radiative transfer model improves the generalizability and accuracy of the result. The research provides theoretical basis and technical support for the remote sensing retrieval of SOM, the development of rapid spectral instruments, and the bands setting of sensor instrument.

In order to recognize different materials from passive polarimetric measurements specifically, a model that can describe the distribution of polarization energy scattered from the target surface is needed for estimating the parameters of interest (complex refractive index n, k and surface roughness σ) of targets. This paper presents a new method for targets parameters estimation. The presented method modifies the degree of linear polarization (DOLP) model by introducing the Lorentz-Drude dispersion constants which do not vary with wavelength as well as the diffuse component of reflection related to the surface roughness to the DOLP equation. This solves the problem of difficulty and inaccuracy caused by multi-angle passive polarimetric measurements used by previous works and helps to improving the parameters estimation accuracy for rough targets. The Monte Carlo simulations result illustrate that the presented method is robust to Gaussian noise. Moreover, the experimental results of rough aluminium plate and rough copper plate validate the effectiveness of the presented model. The errors of estimated n are about 8% for aluminium plate and 9% for copper plate. The errors of estimated k are less than 4% for aluminium plate and no more than 5% for copper plate. The errors of estimated surface roughness σ are about 6.8% and about 6% for aluminium plate and copper plate respectively. The high estimation accuracy and less detection angle requirement achieved by the presented method are significant for material characterization in passive remote sensing application.

In tunable diode laser absorption spectroscopy (TDLAS), the wavelength modulation spectroscopy (WMS) is widely employed. However, when measuring the gas concentration with the second harmonic signal, errors will occur and the inversion accuracy will be reduced, because the change of the modulation depth will lead to the change of the second harmonic peak value, and the value of modulation depth cannot be calculated directly in the absence of gas component. To improve the inversion accuracy of WMS technology, we calibrate the modulation depth and take it as a parameter to the concentration inversion model. In this article, the relationship between the modulation depth and the valley spacing of the second harmonic is first derived, and used to calibrate the modulation depth in the measurement; then, an improved concentration inversion model is established by adding the calibrated modulation depth to the traditional model. Through simulations in Simulink and gas concentration measurement experiments with carbon monoxide (CO), the effectiveness of modulation depth calibration using valley spacing and the accuracy of the improved model are verified. The gas concentration measurement error of the improved model (relative error -5) is reduced by nearly one order of magnitude compared to the traditional model without calibration of the modulation depth (relative error -4). The correction method of modulation depth in this article does not need the priori information of the gas component, and the improved model produces a better determination of gas concentration immune to the variations of modulation depth, which is important for field measurements especially those with large environmental variations.

Considering that industrial data exhibit nonlinearity, high dimensionality and inherent multiscale characteristics, this paper proposes an intelligent industrial process monitoring and fault diagnosis method based on the discrete wavelet transform and deep learning. First, the discrete wavelet transform is used to present the multiscale representation of the raw data. Second, a multiscale convolutional neural network is used to extract the features at each scale, and then the extracted multiscale features are fused by the long short-term memory network to further reduce useless information and retain useful information. Finally, softmax classification is performed. The proposed method has two advantages: 1) the hierarchical learning structure with multiple pairs of convolutional and pooling layers can effectively learn nonlinear, high-dimensional fault features; and 2) the multiscale feature learning scheme can capture complementary diagnosis information at different scales. Detailed comparative studies between the proposed method and conventional methods have been carried out through the Tennessee Eastman benchmark process and the p-xylene oxidation reaction process.

This paper mainly introduces a design circuit and control method of pulse power supply for high-power semiconductor laser diode arrays. According to the high voltage and large current operating characteristics of semiconductor laser, the control strategy of multi-module cascade topology combined with multiple MOSFET connected in parallel is adopted. The topology of multi-module cascade is presented to meet the requirement of wide output voltage range. The multiple MOSFET connected in parallel in the grounding module is proposed to increase output current of the pulse power supply, and each MOSFET has exclusive closed-loop regulator. The pulse current is controlled by the gate driving voltage of the MOSFET which operates in the linear zone, which can ensure the pulse current to be smooth and stable without overshoot. Digital control and analog control are combined to make the control of the pulse power supply more flexible. An experimental platform with a pulse power of 132.3 kW is built. The parameters of pulse power supply are exhibited: pulse current 210 A, pulse voltage 630 V, pulse width 200μs, and repetition frequency 100 Hz.

Regular maintenance inspection and testing is essential in extending cable life and reducing failure probability. This can be achieved by improving the conduit conditions and taking corrective actions on faulty cable components and accessories. Regulators and corporate governance among power utilities require the implementation of risk-based approaches to asset management. However, practitioners lack sufficient historical event data and knowledge that allow them to determine the failure probability of individual cable components, which is an essential component for risk assessment, due to that the high voltage (HV) cable population are relatively young, and many have not yet reached the end of their design life. This paper presents a novel holistic approach to allow the risk based maintenance strategy to be conveniently implemented for the cable conduit, cable terminations, joints, main bodies and the earthing systems separately for each cable circuit. Contributions include: (i) a failure frequency model which accounts for every past failure record of individual cable circuit components to approximate the probability of failure. This, when multiplying with the cable importance or failure consequence, yields the risk level of an individual cable component or a cable circuit; and (ii) a method of optimally scheduling the maintenance activities by setting the objective functions as the minimal cable system risk. The benefit of the simple failure frequency model has the advantage of not having to depend on human intervention and it does not need a large sample to generate valid results, as is the case with other statistical methods. Results of applying the proposed maintenance scheduling model to 21 selected High Voltage (HV) cable circuits show that the average risk can be significantly reduced while continuing with the same number of inspections and test operations.

The meaningful data-based fault diagnosis is beforehand revealing the potential faults to reduce the costly breakdowns, one challenging of which is extracting the weak features from the complicated signals. Ensemble noise-reconstructed EMD (ENEMD) is an intelligent method by the nice integration of adaptively decomposing and naturally denoising. However, ENEMD still suffers from such issues as the false possible noise-only IMFs and the universal minimax threshold, reducing the precision of the critical noise estimation for the weak feature extraction. Thus, the dual-mode noise-reconstructed EMD method is proposed for weak feature extraction and fault diagnosis of rotating machinery. First, the possible noise-only IMF selection rule is redesigned according to the noise characteristic and the correlation evaluation, to eliminate the redundant slowly oscillating IMFs mistakenly chosen for noise estimation. Second, the adaptive local minimax threshold is proposed in the noise estimation technique for the low SNR signal, to overcome the drawback of additionally keeping some critical but weak fault features into the estimation noise. Hereinto, the local threshold is respectively performed in each sliding window defined by the demodulated rotating-related feature frequency. Third, the proposed method is addressed with the flowchart. Finally, two engineering case studies are implemented to demonstrate the feasibility and effectiveness of the method. The analytic results show that the method could effectively extract the periodic impulses generating by the early local damage in the gearbox of a hot strip finishing mill. Meanwhile, the method could successfully reveal the weak rubbing-impact faults along with alleviating the mode mixing phenomenon in the refined results for fault diagnosis of a heavy oil catalytic cracking unit. Hence, the method could provide a promising tool for weak feature extraction and fault diagnosis of rotating machinery.

Rapid and accurate monitoring of soil organic matter (SOM) content is of great significance for precision fertilization of farmland. However, the SOM retrieval models are mainly established by statistical methods, which have limited application scope and incomplete theoretical foundation. Moreover, the accuracy of the SOM retrieval models remains raised. In this paper, for the first time, a semi-empirical SOM content retrieval model is constructed, which has certain theoretical basis, strong applicability and higher accuracy than before. Based on the Kubelka-Munk (KM) theory, the SOM retrieval model with the absorption coefficient k and scattering coefficient s related to SOM (r = k/s) is derived. The validity and reliability of the model are confirmed with validation set (n = 26) including three sorts of soils. Results show that the model can estimate SOM content in different sorts of soils with high prediction accuracy and good prediction ability (root mean square errors of prediction (RMSEP), coefficients of determination (R2) and relative percentage deviation (RPD) values of 0.18%, 89.9% and 3.2, respectively) in the range of 552-950nm. The model provides an innovative method for predicting SOM content.

Broadband absorber in the terahertz region (0.1-10 THz) has attracted considerable attentions due to its important applications in detecting, imaging, and electromagnetic stealth. Recently, terahertz absorber with broadband features has been widely investigated, however, the achievement of ultrabroad bandwidth is still challenging due to the limitations of complex structural design and fabrication processes. In this paper, an ultrabroadband terahertz absorber covering the entire terahertz regime (0.1-10 THz) based on the heavily doped silicon has been designed and fabricated, which is composed of double-layer binary gratings filled with the SU-8 photoresist. Antireflection techniques (SU-8 layer) were utilized to further promote the performance of the terahertz absorber at high frequencies through matching the impedance between free space and doped-silicon substrate. The measured absorption exceeding 87% within the frequency range of 0.3-10 THz has verified the proposed approach in designing the ultrabroadband terahertz absorber. Furthermore, the designed absorber remains high performance in the case of wide-angle incidence even up to 60°. Benefiting from the simple structure, the absorber is easy to be fabricated by common optical lithography. We believe that the results of this paper could broaden the application areas of terahertz absorbers.

A reconstruction approach is presented with a random laser plane and a cylindrical reference. A semi-cylinder reference with the known diameter is designed as the cylindrical-coordinate-system. The optimization function is provided by the parameterized re-projection errors of the feature points on the cylindrical reference, the laser intersection points on the cylindrical reference and the object. The reconstruction approach is tested and analyzed by contrasting the difference between the reconstructed value and the true value of the vernier caliper. The means of the root mean squares of the optimization and initialization are 1.36 and 1.44 mm, respectively. The experiments indicate that the approach with the random laser plane and the cylindrical reference is a reliable method for the 3-D reconstruction in the active vision system with the planar structured light.

With the increasing development of information technology and emerging of modern mobile devices, researchers have been shedding light on the integration of ubiquitous learning and online games in the construction of smart campus, aiming to enhance the online users' learning performance and perceptions. However, challenges of such blend learning model, for example, how to avoid the learners' distraction from online games, triggered many concerns from software engineers and educators. This paper investigated the factors and driving mechanism to construct a learner's technology engagement (TE) model towards ubiquitous game-based learning (UGL) in the smart campus context. Based on the technology acceptance model and the media system dependency theory, this paper identified the relationships among individual expectation (IE), facility environment, social environment (SE), and TE. A total of 168 participants, who were randomly assigned in the UGL settings, took part in the investigation. SPSS 19.0 was employed to analyze factorial validity, internal consistency reliability, correlation, and multiple regression analyses. The result shows that learners' TE toward UGL was influenced directly by IE (including fragmented learning and infotainment learning) and SE (including peer coaching and parental supports). In addition, the ease of use significantly moderated the fragmented learning and TE in the context of smart campus, while it was an insignificant moderator of infotainment. Besides, the ease of use significantly moderated the coaching of peers and TE, while it failed to moderate the parental supports and TEs. The theoretical and practical implications are finally discussed in terms of the development tendency of the construction of smart campus with UGL model.

This paper mainly introduces a pulse power supply for high-power semiconductor laser diode arrays with a novel micro-current pre-start control method. According to the operating characteristics of the semiconductor laser, the output current of the pulse power supply is required to be smooth, stable, and without overshoot. In the method of normal direct pulse start control, semiconductor laser is damaged easily by the overshoot and oscillation of pulse current caused by the switching threshold voltage of MOSFET. And the time delay between the current reference signal and the output pulse current makes the design of closed loop more complicated. Based on the above problems, the method of micro-current pre-start control is proposed. The method is simple to operate and it can solve the nonlinear problem of MOSFET caused by the switching threshold voltage. Therefore, it solve the problems of time delay between the current reference signal and the output pulse current and the overshoot and oscillation of output pulse current in the rising process. A 25.6-kW prototype is intended for verifying the effectiveness of the micro-current pre-start control.

An inversion-mode In0.53Ga0.47As FinFET exhibiting a transition frequency (fT) of 271 GHz and a maximum oscillation frequency (fmax) of 78 GHz at Vds = 0.5 V is reported in this work. Simultaneously, the transconductance (gm) of 2056 (μS/μm) and drain induced barrier lowering (DIBL) of 121 (mV/V) obtained were also exhibited. Moreover, a subthreshold swing (SS) of 86 and 97 (mV/dec) was obtained at Vds = 0.05 and 0.5 V, respectively. These exceptional properties are the result of employing N2 remote plasma treatment, enhancing the quality of high-k/III-V interface and carrier mobility. The fin structure contributes to good suppression of short channel effects (SCEs), and the superior transfer characteristics are due to the improved gate controllability. We also extracted an effective electron velocity (νeff) of 2.18 × 107 (cm/s) by delay time analysis, providing a probable explanation for the obtained fT performance. This work explicitly exhibits the potential of high-performance inversion-mode InGaAs FinFETs, positioning them as promising candidates for future high-frequency system applications. [ABSTRACT FROM AUTHOR]