China Academy of Space Technology

Visual domain adaptation aims to learn robust classifiers for the target domain by leveraging knowledge from a source domain. Existing methods either attempt to align the cross-domain distributions, or perform manifold subspace learning. However, there are two significant challenges: (1) degenerated feature transformation, which means that distribution alignment is often performed in the original feature space, where feature distortions are hard to overcome. On the other hand, subspace learning is not sufficient to reduce the distribution divergence. (2) unevaluated distribution alignment, which means that existing distribution alignment methods only align the marginal and conditional distributions with equal importance, while they fail to evaluate the different importance of these two distributions in real applications. In this paper, we propose a Manifold Embedded Distribution Alignment (MEDA) approach to address these challenges. MEDA learns a domain-invariant classifier in Grassmann manifold with structural risk minimization, while performing dynamic distribution alignment to quantitatively account for the relative importance of marginal and conditional distributions. To the best of our knowledge, MEDA is the first attempt to perform dynamic distribution alignment for manifold domain adaptation. Extensive experiments demonstrate that MEDA shows significant improvements in classification accuracy compared to state-of-the-art traditional and deep methods.

With the increasingly excellent performance of magnesium alloy materials, magnesium alloys are increasingly widely used under the urgent need for weight reduction in aerospace applications. However, due to the severe aviation environment, the strength, corrosion resistance and electrical conductivity of magnesium alloy materials need to be further improved. Many scholars are committed to studying higher comprehensive mechanical properties. Besides, they have studied surface treatment processes with space application characteristics, such as high emissivity oxidation and high anti-corrosion electroplating. To further improve the safety and reliability of magnesium alloys and expand their applications, this paper discusses several kinds of magnesium alloys and summarizes their research progress. The whole manuscript should be revised by an expert who has more experience on English writing. At the same time, the surface treatments of magnesium alloy materials for aerospace are analyzed. Besides, the application of magnesium alloy in aerospace field is summarized. With the in-depth research of many scholars, the improvement of material properties and the development of surface protection and functional technology, it is believed that magnesium alloys will be used in more and more aerospace applications and make more contributions to the aerospace field.

ABSTRACT This paper investigates the determinants of leveraged buyout (LBO) activity by comparing firms that have implemented LBOs to those that have not. Consistent with the free cash flow theory, we find that firms that initiate LBOs can be characterized as having a combination of unfavorable investment opportunities (low Tobin's q ) and relatively high cash flow. LBO firms also tend to be more diversified than firms which do not undertake LBOs. In addition, firms with high expected costs of financial distress (e.g., those with high research and development expenditures) are less likely to do LBOs.

Abstract Forty-five years after the Apollo and Luna missions returned lunar samples, China's Chang’E-5 (CE-5) mission collected new samples from the mid-latitude region in the northeastern Oceanus Procellarum of the Moon. Our study shows that 95% of CE-5 lunar soil sizes are found to be within the range of 1.40–9.35 μm, while 95% of the soils by mass are within the size range of 4.84–432.27 μm. The bulk density, true density and specific surface area of CE-5 soils are 1.2387 g/cm3, 3.1952 g/cm3 and 0.56 m2/g, respectively. Fragments from the CE-5 regolith are classified into igneous clasts (mostly basalt), agglutinate and glass. A few breccias were also found. The minerals and compositions of CE-5 soils are consistent with mare basalts and can be classified as low-Ti/low-Al/low-K type with lower rare-earth-element contents than materials rich in potassium, rare earth element and phosphorus. CE-5 soils have high FeO and low Mg index, which could represent a new class of basalt.

Accurate electrical load forecasting is of great significance to help power companies in better scheduling and efficient management. Since high levels of uncertainties exist in the load time series, it is a challenging task to make accurate short-term load forecast (STLF). In recent years, deep learning approaches provide better performance to predict electrical load in real world cases. The convolutional neural network (CNN) can extract the local trend and capture the same pattern, and the long short-term memory (LSTM) is proposed to learn the relationship in time steps. In this paper, a new deep neural network framework that integrates the hidden feature of the CNN model and the LSTM model is proposed to improve the forecasting accuracy. The proposed model was tested in a real-world case, and detailed experiments were conducted to validate its practicality and stability. The forecasting performance of the proposed model was compared with the LSTM model and the CNN model. The Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE) and Root Mean Square Error (RMSE) were used as the evaluation indexes. The experimental results demonstrate that the proposed model can achieve better and stable performance in STLF.

The recent advances in deep transfer learning reveal that adversarial learning can be embedded into deep networks to learn more transferable features to reduce the distribution discrepancy between two domains. Existing adversarial domain adaptation methods either learn a single domain discriminator to align the global source and target distributions, or pay attention to align subdomains based on multiple discriminators. However, in real applications, the marginal (global) and conditional (local) distributions between domains are often contributing differently to the adaptation. There is currently no method to dynamically and quantitatively evaluate the relative importance of these two distributions for adversarial learning. In this paper, we propose a novel Dynamic Adversarial Adaptation Network (DAAN) to dynamically learn domain-invariant representations while quantitatively evaluate the relative importance of global and local domain distributions. To the best of our knowledge, DAAN is the first attempt to perform dynamic adversarial distribution adaptation for deep adversarial learning. DAAN is extremely easy to implement and train in real applications. We theoretically analyze the effectiveness of DAAN, and it can also be explained in an attention strategy. Extensive experiments demonstrate that DAAN achieves better classification accuracy compared to state-of-the-art deep and adversarial methods. Results also imply the necessity and effectiveness of the dynamic distribution adaptation in adversarial transfer learning.

Abstract Origami‐based designs refer to the application of the ancient art of origami to solve engineering problems of different nature. Despite being implemented at dimensions that range from the nano to the meter scale, origami‐based designs are always defined by the laws that govern their geometrical properties at any scale. It is thus not surprising to notice that the study of their applications has become of cross‐disciplinary interest. This article aims to review recent origami‐based applications in engineering, design methods and tools, with a focus on research outcomes from 2015 to 2020. First, an introduction to origami history, mathematical background and terminology is given. Origami‐based applications in engineering are reviewed largely in the following fields: biomedical engineering, architecture, robotics, space structures, biomimetic engineering, fold‐cores, and metamaterials. Second, design methods, design tools, and related manufacturing constraints are discussed. Finally, the article concludes with open questions and future challenges.

Abstract TianQin is a planned space-based gravitational wave (GW) observatory consisting of three Earth-orbiting satellites with an orbital radius of about $10^5 \, {\rm km}$. The satellites will form an equilateral triangle constellation the plane of which is nearly perpendicular to the ecliptic plane. TianQin aims to detect GWs between $10^{-4} \, {\rm Hz}$ and $1 \, {\rm Hz}$ that can be generated by a wide variety of important astrophysical and cosmological sources, including the inspiral of Galactic ultra-compact binaries, the inspiral of stellar-mass black hole binaries, extreme mass ratio inspirals, the merger of massive black hole binaries, and possibly the energetic processes in the very early universe and exotic sources such as cosmic strings. In order to start science operations around 2035, a roadmap called the 0123 plan is being used to bring the key technologies of TianQin to maturity, supported by the construction of a series of research facilities on the ground. Two major projects of the 0123 plan are being carried out. In this process, the team has created a new-generation $17 \, {\rm cm}$ single-body hollow corner-cube retro-reflector which was launched with the QueQiao satellite on 21 May 2018; a new laser-ranging station equipped with a $1.2 \, {\rm m}$ telescope has been constructed and the station has successfully ranged to all five retro-reflectors on the Moon; and the TianQin-1 experimental satellite was launched on 20 December 2019—the first-round result shows that the satellite has exceeded all of its mission requirements.

Adaptive camouflage in thermal imaging, a form of cloaking technology capable of blending naturally into the surrounding environment, has been a great challenge in the past decades. Emissivity engineering for thermal camouflage is regarded as a more promising way compared to merely temperature controlling that has to dissipate a large amount of excessive heat. However, practical devices with an active modulation of emissivity have yet to be well explored. In this letter we demonstrate an active cloaking device capable of efficient thermal radiance control, which consists of a vanadium dioxide (VO2) layer, with a negative differential thermal emissivity, coated on a graphene/carbon nanotube (CNT) thin film. A slight joule heating drastically changes the emissivity of the device, achieving rapid switchable thermal camouflage with a low power consumption and excellent reliability. It is believed that this device will find wide applications not only in artificial systems for infrared camouflage or cloaking but also in energy-saving smart windows and thermo-optical modulators.

The influence of inherent imperfections should be systematically investigated to ensure the safety and utilization of additive manufacturing-fabricated multi-scale parts and structures. Herein, two different types of multi-layer lattice sandwich panels, BCC and BCCZ, are prepared by selective laser melting (SLM) using the AlSi10Mg material. X-ray micro-computed tomography (μ-CT) is employed to capture the realistic geometrical information of lattice struts. Based on the statistical characteristics, a novel finite element model is established, which considers the specific non-uniform distribution of geometrical imperfection. Uniaxial compressive tests are performed to evaluate the influence of defects and number of layers on the overall mechanical performance and energy absorption capability. The results reveal that the diameter deviation of struts is changed with the change of strut location and built angle. In terms of compressive modulus and initial crushing strength, the prediction results of the reconstruction model are consistent with experimental results as compared to the as-designed and statistical average models. The layer-by-layer crushing behavior is the main failure mode for the multi-layer lattice panels. With the increase of layers number, the densification strain and crash load efficiency increased, whereas the specific energy absorption gradually decreased due to the impact of boundary conditions and failure modes.

The Internet of Things is a novel paradigm with access to wireless communication systems and artificial intelligence technologies, which is considered to be applicable to a variety of promising fields and applications. Meanwhile, the development of the fifth-generation cellular network technologies creates the possibility to deploy enormous sensors in the framework of the IoT and to process massive data, challenging the technologies of communications and data mining. In this article, we propose a novel paradigm, 5G Intelligent Internet of Things (5G I-IoT), to process big data intelligently and optimize communication channels. First, we articulate the concept of the 5G I-IoT and introduce three major components of the 5G I-IoT. Then we expound the interaction among these components and introduce the key methods and techniques based on our proposed paradigm, including big data mining, deep learning, and reinforcement learning. In addition, an experimental result evaluates the performance of 5G I-IoT, and the effective utilization of channels and QoS have been greatly improved. Finally, several application fields and open issues are discussed.

Abstract The exploitation of highly efficient carbon dioxide reduction (CO 2 RR) electrocatalyst for methane (CH 4 ) electrosynthesis has attracted great attention for the intermittent renewable electricity storage but remains challenging. Here, N‐heterocyclic carbene (NHC)‐ligated copper single atom site (Cu SAS) embedded in metal–organic framework is reported (2Bn‐Cu@UiO‐67), which can achieve an outstanding Faradaic efficiency (FE) of 81 % for the CO 2 reduction to CH 4 at −1.5 V vs. RHE with a current density of 420 mA cm −2 . The CH 4 FE of our catalyst remains above 70 % within a wide potential range and achieves an unprecedented turnover frequency (TOF) of 16.3 s −1 . The σ donation of NHC enriches the surface electron density of Cu SAS and promotes the preferential adsorption of CHO* intermediates. The porosity of the catalyst facilitates the diffusion of CO 2 to 2Bn‐Cu, significantly increasing the availability of each catalytic center.

Synthetic aperture radar (SAR) image classification is a hot topic in the interpretation of SAR images. However, the absence of effective feature representation and the presence of speckle noise in SAR images make classification difficult to handle. In order to overcome these problems, a deep convolutional autoencoder (DCAE) is proposed to extract features and conduct classification automatically. The deep network is composed of eight layers: a convolutional layer to extract texture features, a scale transformation layer to aggregate neighbor information, four layers based on sparse autoencoders to optimize features and classify, and last two layers for postprocessing. Compared with hand-crafted features, the DCAE network provides an automatic method to learn discriminative features from the image. A series of filters is designed as convolutional units to comprise the gray-level cooccurrence matrix and Gabor features together. Scale transformation is conducted to reduce the influence of the noise, which integrates the correlated neighbor pixels. Sparse autoencoders seek better representation of features to match the classifier, since training labels are added to fine-tune the parameters of the networks. Morphological smoothing removes the isolated points of the classification map. The whole network is designed ingeniously, and each part has a contribution to the classification accuracy. The experiments of TerraSAR-X image demonstrate that the DCAE network can extract efficient features and perform better classification result compared with some related algorithms.

Two-dimensional (2D) semiconductors are very promising channel materials in next-generation field effect transistors (FETs) due to the enhanced gate electrostatics and smooth surface. Two new 2D materials, arsenene and antimonene (As and Sb analogues of graphene), have been fabricated very recently. Here, we provide the first investigation of the many-body effect, carrier mobility, and device performance of monolayer (ML) hexagonal arsenene and antimonene based on accurate ab initio methods. The quasi-particle band gaps of ML arsenene and antimonene by using the GW approximation are 2.47 and 2.38 eV, respectively. The optical band gaps of ML arsenene and antimonene from the GW-Bethe–Salpeter equation are 1.6 and 1.5 eV, with exciton binding energies of 0.9 and 0.8 eV, respectively. The carrier mobility is found to be considerably low in ML arsenene (21/66 cm2/V·s for electron/hole) and moderate in ML antimonene (150/510 cm2/V·s for electron/hole). In terms of the ab initio quantum transport simulations, the optimized sub-10 nm arsenene and antimonene FETs can satisfy both the low power and high performance requirements in the International Technology Roadmap for Semiconductors in the next decade. Together with the observed high stability under ambient condition, ML arsenene and antimonene are very attractive for nanoscale optoelectronic and electronic devices.

Recently, large quantities of oily wastewater discharged from our daily life and industries have caused serious environmental problems. In addition, frequent oil spill accidents occurred all over the world have also lead to a waste of precious resources. Oil/water separation has become a worldwide challenge for us to overcome. Nowadays, superwetting materials have attracted considerable attention. Among them, porous materials with special wettability are more popular since this kind of materials is easy to fabricate, cost saving and time saving. Moreover, by combining the design of special wettability with the proper pore size, the porous materials could achieve the separation of sundry oil/water mixtures including immiscible oil/water mixtures and stabilized emulsions. In this review, we summarized two types of superwetting porous materials for immiscible oil/water mixtures separation and emulsion separation: water blocking porous materials with superhydrophobic/superoleophilic wettability and oil blocking porous materials with superhydrophilic/underwater superoleophobic wettability. In each type, we introduce the mechanism, fabricating process, effects of oily wastewater treatment and the representative works in detail. Moreover, the smart controllable superwetting porous materials and the wastewater treatment of other pollutants are also introduced briefly.

2D semiconductors are promising channel materials for field‐effect transistors (FETs) with potentially strong immunity to short‐channel effects (SCEs). In this paper, a grain boundary widening technique is developed to fabricate graphene electrodes for contacting monolayer MoS 2 . FETs with channel lengths scaling down to ≈4 nm can be realized reliably. These graphene‐contacted ultrashort channel MoS 2 FETs exhibit superior performances including the nearly Ohmic contacts and excellent immunity to SCEs. This work provides a facile route toward the fabrication of various 2D material‐based devices for ultrascaled electronics.

Graphene-based strain sensors have attracted much attention recently. Usually, there is a trade-off between the sensitivity and resistance of such devices, while larger resistance devices have higher energy consumption. In this paper, we report a tuning of both sensitivity and resistance of graphene strain sensing devices by tailoring graphene nanostructures. For a typical piezoresistive nanographene film with a sheet resistance of ∼100 KΩ/□, a gauge factor of more than 600 can be achieved, which is 50× larger than those in previous studies. These films with high sensitivity and low resistivity were also transferred on flexible substrates for device integration for force mapping. Each device shows a high gauge factor of more than 500, a long lifetime of more than 10(4) cycles, and a fast response time of less than 4 ms, suggesting a great potential in electronic skin applications.

Epitaxial growth of A–A and A–B stacking MoS2 on WS2 via a two-step chemical vapor deposition method is reported. These epitaxial heterostructures show an atomic clean interface and a strong interlayer coupling, as evidenced by systematic characterization. Low-frequency Raman breathing and shear modes are observed in commensurate stacking bilayers for the first time; these can serve as persuasive fingerprints for interfacial quality and stacking configurations. As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.

The production of H2O2 has been taken for a crucial reason for antimicrobial activity of ZnO without light irradiation. However, how the H2O2 generates in ZnO suspension is not clear. In the present work, the comparatively detections on three kinds of ZnO, tetrapod-like ZnO whiskers (t-ZnO), nanosized ZnO particles (n-ZnO), and microsized ZnO particles (m-ZnO), showed that the antimicrobial activity of ZnO was correlated with its production of H2O2. Oxygen vacancy (V(O)) in the surface layer of ZnO crystals determined by XPS indicated that it was quite probably involved in the production of H2O2. To validate the role of V(O), the concentration of VO in t-ZnO was adjusted by heat-treatment under the atmospheres of H2, vacuum, and O2, respectively, and the H2O2 production and antimicrobial effect were detected. Consistently, the t-ZnO treated in H2, which possessed the most V(O) in its crystal, produced the most H2O2 and displayed the best antimicrobial activity. These results provide the basis for developing a more detailed mechanism for H2O2 generation catalyzed by ZnO and for taking greater advantage of this type of antimicrobial agent.

A successful two-dimensional (2D) semiconductor successor of silicon for high-performance logic in the post-silicon era should have both excellent performance and air stability. However, air-stable 2D semiconductors with high performance were quite elusive until the air-stable Bi2O2Se with high electron mobility was fabricated very recently (J. Wu, H. Yuan, M. Meng, C. Chen, Y. Sun, Z. Chen, W. Dang, C. Tan, Y. Liu, J. Yin, Y. Zhou, S. Huang, H. Q. Xu, Y. Cui, H. Y. Hwang, Z. Liu, Y. Chen, B. Yan and H. Peng, Nat. Nanotechnol., 2017, 12, 530). Herein, we predict the performance limit of the monolayer (ML) Bi2O2Se metal oxide semiconductor field-effect transistors (MOSFETs) by using ab initio quantum transport simulation at the sub-10 nm gate length. The on-current, delay time, and power-delay product of the optimized n- and p-type ML Bi2O2Se MOSFETs can reach or nearly reach the high performance requirements of the International Technology Roadmap for Semiconductors (ITRS) until the gate lengths are scaled down to 2 and 3 nm, respectively. The large on-currents of the n- and p-type ML Bi2O2Se MOSFETs are attributed to either the large effective carrier velocity (n-type) or the large density of states near the valence band maximum and special shape of the band structure (p-type). A new avenue is thus opened for the continuation of Moore's law down to 2-3 nm by utilizing ML Bi2O2Se as the channel.