Search

The fusion of multiple monitoring sensors is crucial to improve the accuracy and robustness of machinery fault diagnosis. However, existing fault diagnosis methods may underestimate the interference of noise in the multi-sensor fusion process, leading to unsatisfied performance. To handle this problem, this paper proposes a deep model based on the frequency adaptive wavelet pyramid. First, an adaptive frequency selection strategy is designed to prune the seriously polluted frequencies and only retain some key frequencies. Then, the self-attention mechanism is used to perform information fusion on the selected frequency bands of different sensors. Finally, a wavelet fusion pyramid is adopted by repeating the fusion process at multiple wavelet decomposition levels. In this way, different sensors can be fused in a more fine-grained manner. The experimental results on two multi-sensor-based fault diagnosis datasets demonstrate the anti-noise capability of our proposed method.

Bilateral filter has demonstrated its effectiveness in many traditional methods for image restoration tasks. In this letter, we incorporate the idea of bilateral grid processing in a CNN framework and propose a bilateral stereo super-resolution network (BSSRnet). Specifically, we use a parallax-attention module to incorporate information from left and right views to learn content-aware bilateral filters. Then, these bilateral filters are used to recover missing details at different spatial locations while preserving stereo consistency. Our network is fully differentiable and is robust to both content and disparity variations. Comparative results show that our BSSRnet achieves state-of-the-art performance on the Flickr1024, Middlebury, KITTI 2012 and KITTI 2015 datasets. Source code is available at.

In stereo image super-resolution (SR), exploiting both intra-view and cross-view information is significant but challenging. As existing single image SR (SISR) methods are powerful in intra-view information exploitation, in this letter, we propose a generic stereo attention module (SAM) to extend arbitrary SISR networks for stereo image SR. Specifically, we apply two identical pretrained SISR networks to stereo images. The extracted stereo features at different stages are fed to SAMs to interact cross-view information. Finally, the intra-view and cross-view information is incorporated by SISR networks for stereo image SR. Experiments on the KITTI2012 , KITTI2015 and Middlebury datasets have demonstrated the effectiveness of our scheme. Using SAM, we can exploit cross-view information while maintaining the superiority of intra-view information exploitation, resulting in notable performance gain to SISR networks. Moreover, SRResNet equipped with our SAM outperforms the state-of-the-art stereo SR methods. Source code is available at https://github.com/XinyiYing/SAM .

Excitonic states of a many-electron system in phosphorene quantum dots (PQDs) are investigated theoretically by using a configuration interaction approach. For a triangular PQD in various dielectric environments, its exciton is found to obey two distinct scaling rules. When there is a strong screening effect present in the nanodot, the exciton binding energy (Δex) is shown to be around -150 meV as the long-range Coulomb interactions are totally suppressed and it increases to about 100 meV when the effective dielectric constant (εr) decreases to 12.5. Over this range of εr, Δex is found to be well fitted into a quadratic form of εr-1, which scales neither linearly with εr-2 like the case of bulk three-dimensional semiconductors nor linearly with εr-1 like the case previously reported for graphene nanostructures. When εr is reduced below 10.0, however, Δex is shown to exhibit a perfect linear relationship with εr-1, which behaves just like that of a two-dimensional graphene sheet. On the other hand, with the reduced εr, the quasiparticle gap is found to decrease instead of increasing like in most of the semiconductor nanostructures. As a result, it is revealed that the relationship of Δex with the quasi-particle gap deviates largely from the linear one previously reported for graphene and many other two-dimensional materials.

Electronic structure of phosphorene nanoflakes which consist of hundreds of phosphorus atoms are studied in the framework of unrestricted Hartree–Fock approach. On the base of Pariser–Parr–Pople model for electron–electron interactions, a simplified Bethe–Salpeter formalism is established for the calculation of excitation states of the system. Taking into account the electron–hole interaction in various dielectric environments, the optical gap of a triangular phosphorene nanoflake is shown to increase as the screening effect becomes stronger while its graphene counterpart exhibits just the opposite dependence. After confirming an exponential dependence of the optical gap on the effective dielectric constant, the quasiparticle and optical gaps are also found to obey an exponential scaling rule against the total number of atoms in the nanoflakes, respectively. By extrapolating the dependence on the size of the system, one is able to estimate the exciton binding energy of a monolayer phosphorene sheet on a SiO2 substrate to be 0.894 eV. The result is found to agree well with the previous experimental result of 0.9 eV.

It is commonly believed that electron correlations would open up a quasiparticle gap in semiconductors. Contrary to this intuitive expectation, here we reveal that phosphorene quantum dots (PQDs) may exhibit just the opposite effect. By using a configuration interaction approach beyond the conventional double-excitation scheme, quasiparticle energies are calculated for hexagonal and rectangular PQDs in various dielectric environments. For the hexagonal PQD with a nominal gap of 2.26 eV, it is found that the quasiparticle shift decreases by more than 500 meV and eventually becomes negative when the effective dielectric constant is reduced from 20.0 to 5.0. For other trapezoidal, triangular, and rectangular PQDs, the quasiparticle shift exhibits a similar amount of decrement after the same change in the dielectric environment. Furthermore, the calculation by adopting the Rytova-Keldysh potential, which may be more suitable to describe two-dimensional screening, also shows a very similar result, although with smaller decrement of the quasiparticle shift. The origin of this anomalous quasiparticle shift is believed to be related to the long-range electron-electron interactions in the distinctive lattice structure of PQDs, as a similar phenomenon has never been found in graphene quantum dots.

Utilizing the exact diagonalization (ED) method, we find that excitons cannot form in π-conjugated molecules such as anthracene, phenanthrene, and pyrene when the electron–electron interaction is governed by the Rytova–Keldysh (RK) potential. Within the Pariser–Parr–Pople (PPP) model, however, the excitons may survive only in the presence of a weak screening effect. Either way, the optical gap is seen to be insensitive to the dielectric environment owing to the opposite contributions from the excitonic effect and quasiparticle correction. Furthermore, the latter two are shown to exhibit almost the same behavior in all three molecules when the screening parameter varies.

Aiming at the special background of continuous tracking of mid-range targets by space-based optical sensor networks, by analyzing the essential characteristics of mid-range target continuous tracking, we proposed a space-based optical sensor scheduling model to optimize target detection and tracking performance jointly. Specifically we employed the real-valued particle swarm optimization algorithm to find the optimal solution. To demonstrate the superiority of our model, we compared our results with GDOP model, which only considers the tracking accuracy. The comparative results show that the tracking accuracy of the proposed model is comparable to the GDOP model. However, the detection probability of our model is significantly improved, which demonstrates that our model can achieve superior detection accuracy while maintaining good tracking performance.

The spatio-temporal information among video sequences is significant for video super-resolution (SR). However, the spatio-temporal information cannot be fully used by existing video SR methods since spatial feature extraction and temporal motion compensation are usually performed sequentially. In this paper, we propose a deformable 3D convolution network (D3Dnet) to incorporate spatio-temporal information from both spatial and temporal dimensions for video SR. Specifically, we introduce deformable 3D convolution (D3D) to integrate deformable convolution with 3D convolution, obtaining both superior spatio-temporal modeling capability and motion-aware modeling flexibility. Extensive experiments have demonstrated the effectiveness of D3D in exploiting spatio-temporal information. Comparative results show that our network achieves state-of-the-art SR performance. Code is available at: https://github.com/XinyiYing/D3Dnet., Accepted by IEEE Signal Processing Letters

Satellite constellation design for space optical systems is essentially a multiple-objective optimization problem. In this work, to tackle this challenge, we first categorize the performance metrics of the space optical system by taking into account the system tasks (i.e., target detection and tracking). We then propose a new non-dominated sorting genetic algorithm (NSGA) to maximize the system surveillance performance. Pareto optimal sets are employed to deal with the conflicts due to the presence of multiple cost functions. Simulation results verify the validity and the improved performance of the proposed technique over benchmark methods.

In a conventional semiconductor, when the dielectric screening effect is suppressed, the exciton binding energy increases and the corresponding excitonic transition would exhibit a redshift in the spectrum. In this work, I study the optical properties of hexagonal graphene nanodots by using a configuration interaction approach and reveal that the edge of the absorption spectrum shows an abnormal blueshift as the environmental dielectric constant ϵr decreases. The two dominant many-body effects in the nanodot: the quasiparticle and excitonic effects are both found to scale almost linearly with ϵr-1. The former is shown to have a larger proportionality constant and thus accounts for the blueshift of the absorption edge. In contrast to the long-range Coulomb interaction, the on-site Coulomb energy is found to have a negative impact on the bright excitonic states. In the presence of a strong dielectric screening effect, a strong short-range Coulomb interaction is revealed to be responsible for the disintegration of the bright exciton.

Electronic structure and optical properties of triangular phosphorene quantum dots have been investigated theoretically. Based on systematic configuration interaction calculations, the ground and excited states of the interacting many-electron system together with its optical absorption spectrum are obtained. For the nanodot with 60 phosphorus atoms in various dielectric environments, it is found that the spin gap of the correlated system surprisingly overlaps its optical gap over a large range of the effective dielectric constant. The overlapping of the spin and optical gaps can be attributed to the fact that the extra correlation energy in the spin singlet almost compensates the exchange energy in the spin triplet in the presence of strong long-range electron-electron interactions. Moreover, both the spin and optical gaps are shown to be greatly suppressed as the screening effect becomes strong. When the dielectric constant decreases below 2.65, it is seen that the spin gap becomes negative and the quantum dot undergoes a phase transition from nonmagnetic to ferromagnetic. Our results are compared with the previous experimental and theoretical works.

By using a configuration interaction approach with up to the fifth excitations taken into account, we study the excitonic effect in the optical absorption in graphene nanodots. While the many-electron states are either singlet or doublet in a triangular nanodot system, all the excited singlet states are found to be optically dark in the absorption. These dark excitons are shown to originate mainly from the geometric symmetry of the system and would remain inactive even when the electron-hole or sublattice symmetry is broken. The first excited state in most of the cases is found to be a dark singlet; however, the order of dark and bright excitonic states is shown to be quite sensitive to the strength of electron-electron interactions such as the dielectric screening from the substrate. All the double degeneracies in the excitonic spectrum are found to be lifted when the rotational symmetry is absent such as in the case of a trapezoidal nanodot; however, the first excited state is shown to still remain a dark exciton when there is a strong screening effect. In order that the optical gap of a graphene nanodot can be efficiently tuned by its dielectric environment, the geometric symmetry is revealed to be a crucial factor.

In this paper an efficient method is proposed for blind separating and reconstructing signals from compressively sensed linear mixtures. We assume that all sources share a common sparse representation basis and nonzero indices of these signals are different mostly. Sparse Bayesian learning (SBL) is developed to estimate the product of mixture matrix and signal. Subsequently, we use clustering method to estimate the mixture matrix and use subspace algorithm to estimate all source signals. Numerical results show efficiency of the proposed algorithm compared to previous method.

At present, there are various methods to compute the infrared radiation characteristics of exhaust plume of the liquid rocket engine. Though they are different in computational complexity. Their ideas and methods are alike. This paper focuses on the computation methods of exhaust plume’s flow field, spectral parameters and radiation transfer equation. Comparison, analysis and conclusion of these methods are presented. Furthermore, existing problems and improvements of them are proposed as well.

Verification of gene expression analysis by quantitative realtime PCR (qRT-PCR). Note: Individual gene expression ratios were calculated using foldchange generated by RNA-seq and plotted against calculations done for the same gene using qRT-PCR. (DOCX 17Â kb)

Comparison of features of porcine lncRNAs and mRNAs. Note: Comparison of the lengths (a), ORF lengths (b) and exon numbers (c) of porcine lncRNAs and mRNAs. (DOCX 302Â kb)

Disintegration and formation of excitons in graphene nanodots are investigated within the framework of mean-field approximation to the Hubbard model. Optically active excitons are found to be disintegrated in the non-magnetic nanodots and are shown to form only when the system enters into the antiferromagnetic state. As the Hubbard parameter U / t exceeds a critical value, the nanodot is found to undergo a phase transition from the non-magnetic to an antiferromagnetic phase. Before the phase transition, both optical and quasiparticle gaps are found to be nearly independent of the interaction strength. After the phase transition, however, the quasiparticle gap is revealed to increase more rapidly than the optical gap, which eventually leads to a positive binding energy of the bright exciton in the antiferromagnetic nanodot. The understanding of this extraordinary disintegration and formation of excitons in different magnetic phases of graphene nanodots is believed to be of paramount importance for their potential optoelectronic applications.Disintegration and formation of excitons in graphene nanodots are investigated within the framework of mean-field approximation to the Hubbard model. Optically active excitons are found to be disintegrated in the non-magnetic nanodots and are shown to form only when the system enters into the antiferromagnetic state. As the Hubbard parameter U / t exceeds a critical value, the nanodot is found to undergo a phase transition from the non-magnetic to an antiferromagnetic phase. Before the phase transition, both optical and quasiparticle gaps are found to be nearly independent of the interaction strength. After the phase transition, however, the quasiparticle gap is revealed to increase more rapidly than the optical gap, which eventually leads to a positive binding energy of the bright exciton in the antiferromagnetic nanodot. The understanding of this extraordinary disintegration and formation of excitons in different magnetic phases of graphene nanodots is believed to be of paramount importance for their ...

The binding energy of an exciton in a semiconductor or an insulator is known to scale linearly with er−2, where er is its dielectric constant. In graphene however, since the kinetic energy scales linearly with the wave number instead of its square, the exciton binding energy is thus expected to scale with er−1. In this work we make use of the configuration interaction approach to study the properties of excitons in graphene nanodots embedded in various dielectric environments. With tens of million configurations taken into account in the calculation, we find that the exciton binding energy can be well described by a single scaling rule in which the scaling factor is found to vary with the dimension of the nanodots as well as with the on-site interaction parameter, which agrees well with a recent experiment. The linear relation of the exciton binding energy found with the quasi-particle gap also agrees with the previous work on bulk graphene and other two-dimensional materials.

This paper proposes a sensor dynamic scheduling algorithm of midcourse target tracking. This algorithm increases SNR constraints to optimization function which treats Geometric Dilution of Precision (GDOP) as the scheduling criterion. Firstly, by analyzing the different process technologies of midcourse target tracking handoff, this paper proposes a dual-stage of technical idea including recapture confirmation and stable tracking. Secondly, at the recapture confirmation stage, for the purpose of increasing the target recapture probability, this paper verifies that it’s necessary and reasonable to increase the target signal to noise ratio as a key constraint in considering the basic factors of the camera detection performance, observed background, stray light and so on. In stable tracking phase, this paper adopts a simplified geometric model which can get higher tracking accuracy to schedule the sensor. Next, The performance of the proposed model are analyzed by using the real particle swarm optimization algorithm that reduces the dimension of the particles and updates the position vector. The simulation results show that the model is more effective.

A simple yet effective state-estimation algorithm is presented and demonstrated to have advantages over previous standard clustering techniques used for the particle probability hypothesis density filter. The idea behind the proposed algorithm is that it uses the latest available information (i.e., the measurements) to direct particle clustering. The particle likelihood and target number estimation, computed during probability hypothesis density recursion, are both used to partition particles into clusters, and the center of each cluster gives the state estimation of an individual target. Simulation results indicate that the proposed algorithm outperforms the standard clustering approach using the k-means algorithm, achieving higher accuracy and shorter computational time.