Guangzhou University
  • Guangzhou, Guangdong, China
Recent publications
Magnetic nanometer combined with microwave thawing (MN-MT) could become a novel solution to challenges uneven and overheating of microwave thawing (MT), while retaining high thawing efficiency, compared to conventional water immersion thawing (WT). In this study, MN-MT was applied to thaw fruit (lychee as an example) for the first time, and was evaluated by comparison with WT, MT and water immersion combined with microwave thawing (WI-MT). Results showed that MN-MT could significantly shorten the thawing time of frozen lychee by 80.67%, 25.86% and 18.83% compared to WT, MT and WI-MT, respectively. Compared to WT, MN-MT was the only thawing treatment which significantly enhanced the release of quercetin-3-O-rutinose-7-O-α-l-rhamnoside, according to HPLC-DAD. Meanwhile, thermal-sensitive procyanidin B2, phenylpropionic acid and protocatechuic acid were found to be protected from degradations only by MN-MT based on UPLC-ESI-QTOF-MS/MS results. In summary, MN-MT is a potential novel treatment for rapid thawing and quality maintenance of frozen fruits.
In this study, it is confirmed that without addition of organic solvent and embedding polymer hydrogel into glass nanopore, bare glass nanopore can faithfully measure various lengths of DNA duplexes from 200 to 3000 base pairs with 200 base pairs resolution, showing well-separated peak amplitudes of blockage currents. Furthermore, motivated by this readout capability of duplex DNA, amplicons from Polymerase Chain Reaction (PCR) amplification are straightforwardly discriminated by bare glassy nanopore without fluorescent labeling. Except simultaneous discrimination of up to 7 different segments of the same lambda genome, various pathogenic bacteria and viruses including SARS-CoV-2 and its mutants in clinical samples can be discriminated at high resolution. Moreover, quantitative measurement of PCR amplicons is obtained with detection range spanning from 0.75 aM to 7.5 pM and detection limit of 7.5 aM, which reveals that bare glass nanopore can faithfully disclose PCR results without any extra labeling.
A beam-column connection with satisfactory seismic performance and assembly efficiency is preferred in prefabricated structures. An evolved detachable configuration for the precast beam-column connection using metallic damper as connector (PCF-MDC-nuts) is proposed. The detachable connection significantly reduces the labor cost of building demolition, repair, and recycling. A specimen PCF-MDC-nuts was designed and constructed based on a precast “wet” connection (specimen PC)). A dog-bone damper was adopted in PCF-MDC-nuts as a sacrificial element. First, a laboratory test was performed on the damper, followed by the beam-column connection. The dominant failure pattern of the dog-bone damper, both in the damper and connection tests, was mainly out-of-plane flexural-torsional buckling. The test results show that PCF-MDC-nuts has a similar flexural strength as PC. The maximum in-elastic drift ratio of PCF-MDC-nuts has been improved to 5.48% compared to that of PC (3.27%). The degradation of PC-MDC-nuts was alleviated compared with PC. The cumulated absorbed energy and maximum equivalent viscous damping ratio of the PCF-MDC-nuts were 41.04 kN·m and 44.17%, respectively, while those of PC were 11.06 kN·m and 19.48%. The seismic performance of PCF-MDC-nuts was superior to PC. The concept of the proposed detachable configuration was justified.
Previous researches showed that the replaceable coupling beam damper (RCBD) can effectively enhance the seismic performance of shear wall structures. However, the extensively studied metallic-type RCBD has poor fatigue performance, while the viscoelastic-type RCBD is sensitive to the loading frequency. To overcome the drawbacks of the RCBD mentioned above, this paper proposes a novel RCBD, i.e., the lead-viscoelastic coupling beam damper (LVCBD), which is not only independent of the loading frequency, but also has better fatigue performance with good energy dissipation capacity. The proposed LVCBD mainly consists of three components, i.e., the lead rods, composite viscoelastic layers and steel plates. Taking advantages of the hyper-elasticity property of rubber and the dynamic recrystallization capability of lead, the combined effects from the rubber and lead make the LVCBD reusable after a severe earthquake, and facilitate the post-earthquake recovery with less or without replacement of the RCBD. To examine the performance of the proposed novel damper, three specimens were manufactured and cyclic loading tests were carried out. The influences of the loading amplitude, loading frequency and fatigue loading on the mechanical properties of the specimen were systematically investigated and discussed. Experimental results indicated that the LVCBD showed a favorable deformation capability, frequency-independent performance, fatigue resistance and stable energy dissipation capacity under cyclic loading.
Influence maximization (IM) is a widely investigated issue in the study of social networks because of its potential commercial and social value. The purpose of IM is to identify a group of influential nodes that will spread information to other nodes in a network while simultaneously maximizing the number of nodes that are ultimately influenced. Traditional IM methods have different limitations, such as limited scalability to address large-scale networks and the neglect of community structural information. Here, we propose a novel influence maximization approach, i.e., the layered gravity bridge algorithm (LGB), to address the IM problem, which emphasizes the local structural information of networks and combines community detection algorithms with an improved gravity model. With the proposed LGB, a community detection method is used to derive the communities, and then the bridge nodes are found, which can be regarded as possible candidate seeds. Later, communities are merged into larger communities according to our proposed algorithm, and new bridge nodes are determined. Finally, all candidate seed nodes are sorted through an improved gravity model to determine the final seed nodes. The algorithm fully explores the network structural information provided by the communities, thereby making it superior to the current algorithms in terms of the number of ultimately infected nodes. Furthermore, our proposed algorithm possesses the potential to alleviate the influence overlap effect of seed nodes. To verify the effect of our approach, the classical SIR model is adopted to propagate information with the selected seed nodes, while experiments are performed on several practical datasets. As indicated by the obtained results, the performance of our proposed algorithm outperforms existing ones.
Co-contamination of soil from microplastics (MP) and arsenic (As) is becoming more prevalent, posing a severe threat to agricultural productivity. However, how this joint pollution affects crop growth needs to be better understood. To assess this, we investigated the transcriptomic and phenotypic patterns of rice (Oryza sativa) to MP, As, and their mixtures. The results revealed that, compared to As, MP had much less impact on rice growth, while the MP-As mixture decreased rice's aboveground biomass and altered As's biodistribution in rice tissues. Transcriptome further corroborated this pattern: 13 (294), 4195 (1842), and 3112 (2063) genes differentially regulated in response to MP, As, and their mixtures were observed in root (leaf) tissues, respectively. The joint application of MP and As produced a synergistic effect on crucial metabolic processes, such as carbohydrate, carboxylic acid, oxoacid, organic acid, amino acid, and tetrapyrrole metabolism. Moreover, we found that the joint stress reprogrammed the expression of hub genes encoding photosynthetic enzymes, protein kinases, and transcription factors, which likely reflect a transcript-driven tradeoff strategy between rice growth and defense. Together, these results strongly indicate that MP aggravated the As-induced toxicity in rice plants, which may impact the crop's acclimation to other abiotic field environments.
Nanoscale zero-valent iron (nZVI) shows high effectiveness in the catalyzed removal of contaminants in wastewater treatment. However, the uncontrolled interfacial electron transfer behavior and formation of surface iron oxide (FeOx) layer led to severe electron wasting and occasionally form highly toxic intermediates. Here, we constructed magnetic mesoporous SiO2 shell on surface of nZVI to stimulate a magnetic spatial confinement effect and regulate the electron transfer pattern. Therein, Fe atom facilely spread out from the nZVI core, orderly release electron to surface adsorbed H2O molecule, which is efficiently transformed into active hydrogen (H*). Meanwhile, in-situ Raman revealed that Fe atoms were involved in the formation of penetrable γ-FeOOH rather than FeOx layer, enabling the continuous inward diffusion of H2O and outward diffusion of H* . Employing the catalyzed removal of halogenated phenols as demo reaction, the presence of magnetic mesoporous SiO2 shell utilized the reaction between electrons and H2O to switch the reaction pathway from the reduction/oxidation hybrid process to hydrodehalogantion, and increased the conversion of halogenated phenols-to-phenols by 5.53 times. This study shows the forehand of improving the decontamination performance of nZVI through sophisticated designed surface coating, as well as fine regulating the environmental behavior of magnetic material via micro-magnetic field.
In order to solve the uncapacitated facility location problem (UFLP) quickly and effectively, an enhanced group theory-based optimization algorithm (EGTOA) is proposed in this paper. Firstly, a new local search operator, One Direction Mutation Operator, is proposed, which is suitable for solving UFLP. Secondly, a Redundant Checking Strategy is presented to further optimize the quality of feasible solutions. To verify the performance of EGTOA, 15 benchmark instances of UFLP is selected in OR-Library, the comparison results with the 16 existing algorithms show that the solution obtained by EGTOA is better than other algorithms, moreover its speed is much faster than state-of-the-art algorithms. These demonstrates that EGTOA is a fast and effective algorithm for solving UFLP.
This paper is concerned with the dynamics of a predator-stage structured model with cannibalism, degenerate diffusion and free boundaries. The existence and uniqueness of the global solution are discussed firstly. Next, we investigate long time behavior of the predators and the prey. Then a spreading–vanishing dichotomy and the criteria are obtained for two cases. The theoretical analyses indicate that: (i) the propagation profile of the free boundary problem is significantly different from that of the corresponding ordinary differential equations; (ii) the transition ratio of two developing forms (cannibalization and natural maturation) from juvenile to adult stage should be sufficiently small so that predators can spread successfully.
Inflammation is a major adverse outcome induced by inhaled particulate matter with a diameter of ≤ 2.5 μm (PM2.5), and a critical trigger of most PM2.5 exposure-associated diseases. However, the key molecular events regulating the PM2.5-induced airway inflammation are yet to be elucidated. Considering the critical role of circular RNAs (circRNAs) in regulating inflammation, we predicted 11 circRNAs that may be involved in the PM2.5-induced airway inflammation using three previously reported miRNAs through the starBase website. A novel circRNA circ_0008553 was identified to be responsible for the PM2.5-activated inflammatory response in human bronchial epithelial cells (16HBE) via inducing oxidative stress. Using a combinatorial model PM2.5 library, we found that the synergistic effect of the insoluble core and loaded Zn²⁺ ions at environmentally relevant concentrations was the major contributor to the upregulation of circ_0008553 and subsequent induction of oxidative stress and inflammation in response to PM2.5 exposures. Our findings provided new insight into the intervention of PM2.5-induced adverse outcomes.
Fundus image retrieval can help ophthalmologists make evidence-based medico-decision by providing similar cases. Its basic task is to learn highly discriminative visual descriptors from image space, in which lesion features are the main differentiating clue. Lesions in fundus images appear small in size, similar in textures, and scatter around vessels, such as microaneurysms and hemorrhages. Hence, although a single small lesion has a saliently visual manifestation, its discriminative information is hard to reserve in the last image descriptors. For fundus images, the optic disc of the left and right eyes are symmetric, and the macular area lies in the central axis from the vertical view. Based on such spatial structure and lesion characteristics, we present a novel deep metric learning framework equipped with mirror attention to enhance the discriminative features of small and scattering lesions and encode them into image descriptors. The mirror attention can give lesions high attention scores by capturing spatial dependency of vertical and horizontal views, especially the relations between lesions and vessels. Based on the mirror attention, we further propose a new fine triplet loss to confine distances of positive pairs by exploiting the learned relevant degrees of positive pairs in a self-supervised manner. The fine triplet loss can help detect the subtle differences of positive pairs to improve the ranking performance of hit items. To demonstrate the effectiveness of improving retrieval performance, we conduct comprehensive experiments on the largest fundus dataset of diabetic retinopathy (DR) detection and achieve the best precision compared to counterparts. The experiments show that our method produces significant performance improvements for fundus image retrieval, especially the ranking quality of DR grades containing microaneurysms and hemorrhages. Our proposed mirror attention can be applied to off-the-shelf backbones and trained efficiently in an end-to-end manner for other medical images to obtain highly discriminative image descriptors.
Poplar, a crucial player in lignocellulosic biofuel production worldwide, can be divided into five types. The feedstock variability would significantly impact the scale-up and commercialization of biofuel technologies from poplar. To date, few studies were found comparing ethanol and biogas production from different types of poplar. This study compared the conversion efficiency of NaOH pretreated five types of poplar residues to ethanol and biogas under three biological conversion processes (namely ethanol fermentation process, anaerobic digestion (AD) process, and cascading ethanol fermentation and AD process); the flow of mass, carbon, nitrogen and energy were also analyzed by material flow analysis. 48.4–60.1 % of theoretical ethanol yield, and the specific methane yield of 201 ± 11.0–270 ± 3.40 mL/g volatile solids were obtained under the co-production process, with the highest yield from Populus trichocarpa (N2). These results showed that the co-production process (cascading ethanol fermentation and AD process) of poplar N2 outperformed ethanol fermentation and AD in terms of energy conversion efficiency, which was 121 % and 42.9 % higher than the ethanol fermentation process and AD process, respectively. Mass and energy balance analysis showed that under the co-production process, 158 g ethanol and 103 g methane could be obtained from 1 kg pretreated poplar N2, corresponding to a total energy output of 9.5 MJ, with an energy conversion efficiency of 54.3 %. This study may provide new insights into the breeding of new poplar species for the current biorefinery process.
Drawing on self-representation theory, we explore how trait competitiveness affects innovative behavior and career satisfaction via perceived insider status and the boundary role of perceived leader competitiveness. Data were collected in two waves among 316 employees in China. The results showed that employee trait competitiveness was positively related to innovative behavior and career satisfaction via perceived insider status. These indirect effects were significantly stronger when working with competitive leaders and were insignificant otherwise. Our findings make theoretical contributions and practical implications regarding how, and under what circumstances, competitive employees will innovate and satisfy with their career development.
Magnetic resonance and nuclear medicine images are the two categories of multimodal medical images. Magnetic resonance images reveal physiological anatomical information of patients, and nuclear medicine images accurately show tissue lesion information. Through medical image fusion algorithms, these fusion images containing both tissue lesion information and physiological anatomical information are obtained to provide sufficient information for clinical medical technologies. However, most existing fusion algorithms are based on mathematical transform domains, and these fusion results have the weaknesses of blurred edges, color distortion and detail loss. To address these problems, a multiscale dense residual attention network (MDRANet) is proposed and applied to magnetic resonance and nuclear medicine image fusion. MDRANet combines multiscale dense network and multiscale residual attention network to extract and enhance deep features. Moreover, four different loss functions are used to optimize MDRANet and improve the fusion quality. The experimental results show that the fusion results of our proposed algorithm have richer details and better objective metrics compared with the reference algorithms.
Dynamic properties of widely used butt welds and fillet welds would affect the behavior of welded joints in steel structures under dynamic loading. In the open literature, there are very limited studies on the behavior of butt welds and fillet welds under dynamic loads, such as impact and blast. In this study, a compression-tension conversion experimental facility was specifically designed for dynamic testing of butt welds and fillet welds at intermediate strain rates. The steel plate (SP), butt weld (BW), tensile fillet weld (TFW), and shear fillet weld (SFW) at different strain rates were tested. The corresponding quasi-static tests were also performed for comparison. The strength, failure mode, and fracture strain of the BW, TFW and SFW under static and dynamic loads were compared and analyzed. The effects of weld strength matching or overmatching on butt welds and fillet welds were also investigated. The results showed that with the increase in strain rate, the strength of the BW, TFW, and SFW increase; the fracture strain of the BW also increases but those of the TFW and SFW decrease. The failure modes of the BW and TFW were significantly influenced by the weld strength matching, whilst that of the SFW was less affected. Two empirical formulae were proposed to predict the dynamic strength of different steel materials and butt welds as well as fillet welds.
This study aims to improve the accuracy of risk prediction and credibility detection of network public opinion (NPO) and optimize the network environment. Using blockchain technology (BT) network system for optimization, a smart contract is used to build a risk management system of NPO, so that public opinion can be traced through the smart ledger using risk association tree technology. First, BT, NPO, and related theories of risk management are introduced. Second, the network situation of public opinion dissemination in the blockchain environment is discussed. Finally, a BT-based risk management model of the NPO is implemented, and experiments are conducted to verify the effect of the model. Furthermore, the theoretical framework of NPO credibility model detection is optimized according to the public opinion risk prediction model. The research results reveal that the three experimental schemes reached the peak density of the number of public opinion disseminators in the third step, and the highest values were 0.1335, 0.1318, and 0.1296, respectively. Moreover, the density of public opinion dissemination in the three experimental schemes reached a stable state of 0.087, 0.088, and 0.0898, respectively after 48 steps. Under BT, the designed three experimental schemes can reasonably conduct the risk prediction and credibility detection of NPO. This work contributes to optimizing the control measures of the network environment.
Automatic identifying target multi-class objects in tunnel scenes from 3D point clouds is widely thought to be critical for maintaining the healthy condition of the tunnel using deep learning methods. However, those methods require extensive data with labels, which is time-consuming and labor-intensive. Targeting effective multi-class tunnel point cloud segmentation for practical applications, this research proposes a deep learning method named semi-supervised learning-based point cloud network (SPCNet) to boost segmentation by alleviating labeling tasks. It contains a supervised learning module, a self-training module, a mean teacher-based learning module, loss functions, and evaluation metrics. To validate the effectiveness and reliability of the proposed method SPCNet, a point cloud collected from a real tunnel is implemented. The results indicate that the proposed method SPCNet performs excellently with MIoU of 0.8741 and 0.8583 in Scenarios I and II, respectively; as well as superior to the supervised learning method with MIoU of 0.8152 and 0.7552 in Scenarios I and II and other state-of-the-art methods such as ST++ and ST. Accordingly, the proposed method SPCNet has superior performance, beneficially contributing to multi-class object segmentation of 3D tunnel point clouds with great potential for applications in practice.
With the emergence of high-performance materials and the demand for light-weight long-span structures, high-strength concrete-filled-steel-tubular(HS-CFST)structures become increasingly popular. However, relevant research on HS-CFST arch remains scarce, and there is no systematic understanding on its mechanical characteristics. In this study, the in-plane ultimate bearing capacity of parabolic HS-CFST arch with fixed ends under vertical five-point-symmetrical-concentrated-loads along the span was studied experimentally. In total, 15 parabolic arch specimens consisting of different strength of steel tubes and core concretes were designed and tested. The mechanical responses of the arch specimens, including ultimate bearing capacity, failure modes, ductility coefficients, and confinement effects were analyzed. It is found that the ultimate bearing capacity, confinement effect and ductility of a HS-CFST arch are closely related to the strength of steel and core concrete. In addition, finite element models were established to predict the bearing capacity of the HS-CFST arch specimens, where the good agreements showcase the practical usefulness of the proposed HS-CFST arch.
Stability is a prerequisite of a successful real-time hybrid simulation (RTHS), which depends on the time-integration algorithm, the delay-compensation method, the loading system dynamics (mostly characterized as a time delay), and the nonlinearity of the structures being tested. The existing stability analysis methods generally provide a qualitative judgement based on one or some of the factors above, without considering all these factors to estimate quantitative stability margin. However, it is critical to design a robust RTHS system with enough stability margin (i.e., robust RTHS system) to ensure the success of the actual test with inescapable uncertainties. To this end, a relative stability analysis method for robustness design of RTHS systems is proposed. In this method, the critical gain of the stiffness of an experimental substructure is used to quantify the stability limit, and the critical delay is used to quantitatively represent the sensitivity of an RTHS system to the phase discrepancy at the interface. The RTHS system with a nonlinear experimental substructure is modeled by a discrete transfer function in an incremental form, wherein the time-integration algorithm, the delay-compensation method, the loading system dynamics, and the nonlinearities of the experimental substructure are considered. The critical delay and gain are obtained from the discrete transfer function. Two relative stability criteria are proposed based on the critical delay and the critical gain of an RTHS system to facilitate the evaluation of the system's stability performance and the design of a robust RTHS system. To verify the critical delay and gain estimated using the proposed method, numerical simulation of RTHS systems was carried out first. Then RTHS experiments of single-degree-of-freedom models with experimental substructures consisting of a linear spring and a nonlinear stiffness hardening specimen were conducted. The consistent results of the numerical simulation and the physical RTHS tests with those derived from the proposed relative stability analysis method demonstrate its effectiveness and accuracy in determining the stability properties of RTHS systems, which can be further utilized in designing more robust RTHS testing systems.
Previous studies on land use have largely focused on characteristics of quota allocation and spatial distribution based on time series analysis. However, indwelling patterns, combined space–time characteristics and terrain gradient effects are often hidden from the regional territorial spatial planning process. By using spatial econometric techniques, including hot spot analysis, landscape theory, a space–time cube model and a terrain–niche index, this paper aimed to assist regional planning by revealing these hidden characteristics in Guangdong Province, China. The results were as follows: urban construction land had formed an agglomerated region at the center of the study area, surrounded by many smaller satellite cities, which resembled a mainland–island metapopulation pattern. Three hot spot types—oscillating (32.45 % of the region), sporadic (1.85 %) and persistent (6.83 %)—formed the main space–time pattern throughout the study period. The main patterns included persistent hot spots in the Pearl River Delta, approximating a strip pattern of sporadic and oscillating hot spots from the east to west, with persistent cold spots in the north. Dry land cropping showed an increase in area at terrain levels 3–6 while in other terrains the area of dry land cropping decreased over time. These findings indicate that more attention needs to be paid to maintaining or adjusting the shape and equilibrium between the mainland and the island cities in the future. Steeper terrain is normally considered for cropland when conflicts exist for land use, which might decrease cropland quality to some extent. Hence, current territorial and spatial planning should focus on changing spatial disequilibrium and the reduction in cropland quality, which might result from the land policy of “superior occupation and inferior compensation”, to slow regional differences and to achieve sustainable development.
Institution pages aggregate content on ResearchGate related to an institution. The members listed on this page have self-identified as being affiliated with this institution. Publications listed on this page were identified by our algorithms as relating to this institution. This page was not created or approved by the institution. If you represent an institution and have questions about these pages or wish to report inaccurate content, you can contact us here.
2,024 members
Yu Huarong
  • School of Civil Engineering
Zhicheng Dong
  • School of Life Sciences
Information
Address
NO.230 Wai Huan Xi Road, Guangzhou Higher Education Mega Center, 510006, Guangzhou, Guangdong, China
Website
http://www.gzhu.edu.cn/
Phone
+86020-39366232