Anhui University of Technology
Recent publications
The canal waterways of China still play an important role in the logistics and transportation industry. Unmanned technology helps to reduce costs and improve the safety of navigation. Real-time environmental awareness is vital to making unmanned surface vehicles (USVs) come true. This paper proposes a new lightweight environmental awareness method based on deep convolutional neural networks ( DCNN ) and a mixed attention mechanism for USVs in canals, which can simultaneously perform ship detection, segmentation, and surface and background segmentation tasks. The features of the ships, surface, and background are extracted by a shared feature extraction backbone network and hybrid attention mechanism, which improves the efficiency of visual environmental awareness. In addition, a dataset named USV-Canal is constructed to enrich the features of canal waterways for environmental awareness, which contains typical canal scenes and 3443 ship objects. To improve the generalization, multiple public datasets are mixed with the USV-Canal dataset to build an integrated dataset to train our model, which boasts diversity in scene types and ship classes. The comparative and field experiments’ results show that 40.9% of mAP , 95.8% of mIoU, and 5 frames per second ( FPS ) inference speed can be achieved, and have good generalization, which can meet the requirements of environmental awareness of low-speed ships in canal waterways Note to Practitioners —The trained and validated model can ultimately be deployed on unmanned surface vehicles, and the required hardware platform is NVIDIA’s Jetson Nano, which is used for real-time perception of surrounding ships and navigable surfaces during navigation. The information can be integrated into the guidance, navigation, and control (GNC) system of USVs, achieving obstacle avoidance and ensuring safe navigation. It is vital to make autonomous navigation come true.
Through millions of years of evolution, bones have developed a complex and elegant hierarchical structure, utilizing tropocollagen and hydroxyapatite to attain an intricate balance between modulus, strength, and toughness. In this study, continuous fiber silk composites (CFSCs) of large size are prepared to mimic the hierarchical structure of natural bones, through the inheritance of the hierarchical structure of fiber silk and the integration with a polyester matrix. Due to the robust interface between the matrix and fiber silk, CFSCs show maintained stable long-term mechanical performance under wet conditions. During in vivo degradation, this material primarily undergoes host cell-mediated surface degradation, rather than bulk hydrolysis. We demonstrate significant capabilities of CFSCs in promoting vascularization and macrophage differentiation toward repair. A bone defect model further indicates the potential of CFSC for bone graft applications. Our belief is that the material family of CFSCs may promise a novel biomaterial strategy for yet to be achieved excellent regen-erative implants.
In this article, we obtain that compact simple Lie groups Sp(n)(n=4k+626)Sp(n)(n=4k+6\ge 26) admit at least two new non-naturally reductive Ad(Sp(k+2)×Sp(k+2)×Sp(k+2)×Sp(k))Ad(Sp(k+2)\times Sp(k+2)\times Sp(k+2)\times Sp(k))-invariant Einstein metrics, and we prove that these Einstein metrics are not geodesic orbit metrics. Furthermore, we construct two left invariant non-geodesic orbit Einstein–Randers metrics on Lie group Sp(n).
Although leniency from leaders is a frequent occurrence in workplaces, there is no research to date regarding the reactions of employees when observing leaders display leniency towards their peers. This paper applies deonance theory to argue that employees interpret leader leniency as a form of comparative grievance, which shapes feelings of envy towards their peers. Subsequently, employees hold the grantor and recipient involved in leadership leniency responsible for this unfavorable situation. Responding out of deontic reactions, it is predicted that employees will react by socially undermining their favored peers and avoiding interaction with lenient leaders. It is also proposed that observers’ rivalry and mindfulness moderate these responses. Two studies (Study 1, N = 314, and Study 2, N = 458) were conducted to empirically test our model. As expected, the results revealed that employees who witnessed leader leniency reacted by socially undermining peers and avoiding interaction with leaders. Our results also found that employees who observe leader leniency react unfavorably harming both their peers and supervisors; and that this reaction peaks when both employees rival with their coworkers and lack mindfulness. The paper concludes by discussing the implications of these findings and suggesting directions for future research.
A compatible distributed optical fiber sensing system based on spatial division multiplexing Raman anti-stokes scattering light and Rayleigh scattering light is proposed and experimentally demonstrated. The sensing fibers for temperature monitoring and vibration event location are separated spatially. To ensure the temperature measurement accuracy, a new temperature calibration algorithm is proposed to overcome the interference caused by the instability of the laser source in the Raman anti-stokes scattering light based temperature measurement scheme. And, for vibration event location by the polarization state analysis of the time domain Rayleigh scattering light, a new threshold method by the subtraction between normalized trace data is adopted to exclude the fiber attenuation factor, and then improve the vibration event location performance. In the experiment, with a multi-mode fiber (MMF) of ~ 12 km and a single mode fiber of ~ 15 km, the temperature error range obtained is within − 1.77 ℃ to 1.56 ℃ and the root mean square error is 1.16 ℃ with a temperature measurement range from 20 ℃ to 90 ℃, and the vibration events are accurately located.
Autonomous collision avoidance technology is the core of unmanned surface vehicles (USVs). Deep reinforcement learning (DRL) is a new approach to avoid collision for USVs. However, most research is based on the assumption of a fixed number of obstacles and ignores the collision prediction to improve safety. To address this problem, a novel "prediction-decision" collision avoidance model based on the deep deterministic policy gradient (DDPG) is proposed. Firstly, a radiation-shaped state space is designed to make the DDPG that can be used in time-varying scenarios with stochastic obstacles. Then, the velocity obstacle (VO) is combined with the state space for training to realize the collision prediction. Subsequently, reward functions are designed using a reward-shaping technique to improve training efficiency and safety. Finally, virtual simulation experiments based on Unity3D and field tests are conducted to verify the algorithms performance. The results show that it can take safe collision avoidance actions in unknown environments and with generalization ability.
New ten tetrakis(benzoxazine) calix[4]resorcinarenes modified by organic amine through Mannich reaction were obtained in this paper. Compounds 2 ~ 11 were characterized by infrared spectroscopies, nuclear magnetic resonance spectroscopies. The structure of compounds 2, 10 and 11 was characterized by single crystal X-ray diffraction. Compound 7 possesses a MIC value of 3.13 μg/mL against S. Aureus, demonstrating a favorable inhibitory effect, and a MIC value of 6.25 μg/mL against E. Coli, likewise showing a good inhibitory effect. The UV and ¹H NMR titration experiments were conducted to explore the host–guest chemistry between compound 10 and acetonitrile molecules, indicating that compound 10 exhibited encapsulation behavior towards acetonitrile molecules through hydrogen bonding interactions. Then Hirshfeld surface analysis showed that H⋅⋅⋅H, C − H⋅⋅⋅O, C − H⋅⋅⋅π, O − H⋅⋅⋅O played an important role in the crystal accumulation of compounds 2, 10 and 11.
Designing asymmetrical structures is an effective strategy to optimize metallic catalysts for electrochemical carbon dioxide reduction reactions. Herein, we demonstrate a transient pulsed discharge method for instantaneously constructing graphene-aerogel supports asymmetric copper nanocluster catalysts. This process induces the convergence of copper atoms decomposed by copper chloride onto graphene originating from the intense current pulse and high temperature. The catalysts exhibit asymmetrical atomic and electronic structures due to lattice distortion and oxygen doping of copper clusters. In carbon dioxide reduction reaction, the selectivity and activity for ethanol production are enhanced by the asymmetric structure and abundance of active sites on catalysts, achieving a Faradaic efficiency of 75.3% for ethanol and 90.5% for multicarbon products at −1.1 V vs. reversible hydrogen electrode. Moreover, the strong interactions between copper nanoclusters and graphene-aerogel support confer notable long-term stability. We elucidate the key reaction intermediates and mechanisms on Cu4O-Cu/C2O1 moieties through in situ testing and density functional theory calculations. This study provides an innovative approach to balancing activity and stability in asymmetric-structure catalysts for energy conversion.
Longitudinal crack is a typical surface defect for slab of steel. Accurate prediction of longitudinal crack is of great significance to improve slab quality. However, in actual production, the quantity distribution of normal and longitudinally cracked slabs is extremely unbalanced, which brings great challenges to the subsequent modeling and prediction. To solve the above problems, this paper proposes a prediction method for the longitudinal crack under the condition of data imbalance. Firstly, multiple sampling methods (SMOTE, BorderlineSMOTE, SMOTE-ENN and SMOTE-Tomk) were used to construct feature data sets respectively to alleviate the problem of data imbalance. Then, based on the data sets after sampling processing, the prediction models of the longitudinal crack were constructed by using Random Forest (RF), Support Vector Machine (SVM), Logistic Regression (LR), XGBoost and LightGBM as classification algorithms. The models’ hyperparameters were determined by Bayesian optimisation and the optimal classification algorithm was selected according to the evaluation metrics. Meantime, SHapley Additive exPlanations (SHAP) was used to analyse the model and verify the influence of input parameters on the model output. The results show that, compared with other models, the combined model of SMOTE sampling and LightGBM classification algorithm can better deal with the problem of imbalanced data for prediction of the longitudinal crack. The Recall of normal and longitudinally cracked slabs are 90.57% and 84.62%, respectively, false alarm rate is 9.44% and AUC is 0.93. At the same time, the training time of this model is about 0.15 s, and prediction time is less than 0.01 s. The time consumed in each stage is significantly shorter than other models. It shows obvious advantages and provides a reliable method for predicting the longitudinal crack.
Humanoid robots, characterized by their anthropomorphic design, have become increasingly common in various service areas. Nevertheless, the majority of current affective designs of humanoid robots primarily concentrate on the physical appearance while overlooking its (audiovisual) integration with voice. In this study, we propose simultaneously designing the appearance and voice of humanoid robots using Kansei Engineering, an effective method for optimizing the affective design of products. We first selected representative humanoid robots with different appearances and voices and constructed kansei space to capture users’ affective needs for these robots. Then, we decomposed appearances and parameterized voices to extract design features and orthogonalized these design features to generate prototypes. After that, we conducted an evaluation experiment to acquire users’ affective evaluations on the combinations of appearance and voice. Based on the data, relationship models between design features and users’ kansei images and holistic preferences were constructed using the back-propagation neural network. Furthermore, optimization design models were formulated and resolved through the genetic algorithm. Also, we conducted a validation experiment, and the results demonstrated that the optimized design schemes look harmonious in appearance, sound warmth in voice, and achieve a high level of audiovisual compatibility. The results suggest that the proposed approach can effectively optimize the audiovisual affective design of humanoid robot appearance and voice. Moreover, it can not only provide methodological support for the affective design of robots and other voice-based smart products but can also help to improve the affective experience quality and facilitate the application of robots in service areas.
To address the challenge of accurately capturing tool wear states in small sample scenarios, this paper proposes a tool wear prediction method that combines XGBoost feature selection with a PSO-BP network. In order to solve the problem of input feature selection and parameter selection in BP neural network, a double-layer programming model of input feature and parameter selection is established, which is solved by XGBoost and PSO. Initially, vibration and cutting force signals from CNC machining are preprocessed using time-domain segmentation, Hampel filtering, and wavelet denoising. Subsequently, time-domain, frequency-domain, and time–frequency domain features are extracted from the preprocessed data using FFT and wavelet packet decomposition, followed by feature screening for tool wear mapping via Pearson correlation and XGBoost feature importance analysis as model input. Finally, PSO is employed to optimize BPNN parameters. Experimental results show that PSO outperforms other algorithms in training the tool wear prediction model, with XGBoost feature selection reducing model construction time by 57.4% and increasing accuracy by 63.57%, demonstrating superior feature selection capabilities over Decision Tree, Random Fores, Adaboost and Extra Trees. These findings suggest that the proposed method can effectively predict tool wear in real-world CNC machining, contributing to improved production efficiency, reduced tool replacement frequency, and lower maintenance costs, thereby providing valuable insights for industrial applications.
In response to the escalating issue of antibiotic pollution in water bodies, with tetracycline (TC) serving as a representative example, this study introduced a novel magnetic nano cobalt @ nano zero valent iron (nCo@nZVI) composite material. To synthesize this material, the rheological phase reaction method was employed to produce sheet-like nZVI, followed by the liquid-phase reduction method to formulate the nCo@nZVI compound. Various advanced characterization techniques, including FESEM, HRTEM, EDS, XPS, XRD, BET, and FTIR, were utilized to systematically evaluate the physical, chemical properties, and structure of the material.Moreover, the study experimentally assessed the TC removal efficiency of nCo@nZVI, exploring the impacts of pH, temperature, and initial heavy metal ion concentration on this efficiency. It is worth noting that, under conditions of a neutral pH of 7, a temperature of 20 °C, and a material dosage of 1 g/L, the initial TC concentration of 20 mg/L in the wastewater was reduced to nearly zero (or completely removed) within 120 min. The adsorption kinetics and isotherm analysis revealed that the TC adsorption process by nCo@nZVI conforms to the pseudo-second-order kinetic model and Langmuir isotherm model, suggesting a predominantly chemical adsorption mechanism. The adsorption capacity derived from the Langmuir model was 25.33 mg/g.Further thermodynamic investigations demonstrated that the TC adsorption by nCo@nZVI is a spontaneous process. Additionally, the material primarily removes TC through an adsorption-degradation mechanism within the Fenton system. This eco-friendly and cost-effective material retains a removal rate of 65.87% after five cycles of regeneration treatment and can be recycled and reused under the influence of an external magnetic field, showcasing significant potential for the remediation of antibiotic-contaminated sites. Graphical abstract
With the widespread adoption of selective catalytic reduction (SCR) denitrification technology and the enforcement of ultra-low emissions regulations, the annual production of spent SCR catalyst, classified as hazardous waste, has been increasing. This study presents an innovative method of integrating spent SCR catalysts into carbon-containing pellets processed in a rotary hearth furnace. The research examines the impact of the addition of spent catalyst on the comprehensive performance and reduction characteristics of carbon-bearing pellets. The results indicate that the drop number for both green and dry pellets peaks at 5.0 wt-% spent catalyst addition. When the addition of spent catalyst reaches 7.5 wt-%, both the number of drops for dry pellets and the compressive strength reach their peaks. Additionally, the compressive strength of direct reduction iron (DRI) gradually increases from 1794 N/pre to 3476 N/pre. The metallisation and zinc removal rates of DRI are maintained at high levels, exceeding 90% and 99%, respectively. The interaction between TiO 2 and FeO in the spent catalyst forms titanium–iron compounds, reducing low-melting eutectic formation. The catalyst's active sites also promote CO production, which enhances iron oxide reduction and leads to the early formation of iron grains, contributing to increased DRI compressive strength. This process not only provides an effective way to manage spent SCR catalysts but also improves pellet performance.
In the CaO–SiO 2 –Al 2 O 3 –MgO–TiO 2 –Na 2 O system slag, the desulphurisation kinetics of pig iron desulphurisation were studied. The results show that the sulphur content reduced with the increase of the temperature, the alkalinity, the lowest content was 0.01% and 0.04%, respectively. With the increase of the content of Na 2 CO 3 and MgO, the sulphur content in hot metal reduced from 0.33% to 0.04% and 0.05%, respectively. In addition, the content of Al 2 O 3 and TiO 2 has a great influence on the desulphurisation rate. The sulphur content in hot metal firstly decreases and then increases with increasing Al 2 O 3 and TiO 2 addition, The activation energy was estimated to be 535.71 kJ/mol at the temperature from 1350 °C to 1450 °C. The desulphurisation slag was analysed by SEM-EDS. A thick layer of calcium sulphide was formed at the interface of slag–metal, which has a negative effect on the desulphurisation reaction. When the blast furnace slag was added Na 2 CO 3 of the plant experiment, desulphurisation rate and the sulphur element in the slag increased. Compared to previous desulphurisation effect without Na 2 CO 3 , the desulphurisation efficiency increases from 50% to 83.3%. The desulphurisation efficiency is improved by optimising the thermodynamic conditions of KR stirring desulphurisation, which is beneficial to reduce desulphurisation flux consumption and the secondary desulphurisation ratio.
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.
393 members
Donghong Wang
  • School of Materials Science and Engineering
Shibiao Ren
  • School of Chemistry and Chemical Engineering
Fa-bin Cao
  • Anhui Provincial Key Lab of Metallurgy Engineering and Resources Employments
Junchao Xu
  • School of Energy and Environment
Information
Address
Ma’anshan, China