Abstract: In recent years, deep neural networks (DNN) have attracted increasing attention because of their excellent performance in computer vision and natural language processing. The success of deep learning is due to the fact that the models have more layers and more parameters, which gives them stronger nonlinear fitting ability. Furthermore, the continuous updating of hardware equipment makes it possible to quickly train deep learning models. The development of deep learning is driven by the greater amounts of available annotated or unannotated data. Specifically, large-scale data provide models with greater learning space and stronger generalization ability. Although the performance of deep neural networks is significant, they are difficult to deploy in embedded or mobile devices with limited hardware due to their large number of parameters and high storage and computing costs. Recent studies have found that deep models based on a convolutional neural network are characterized by parameter redundancy as well as parameters that are irrelevant to the final model results, which provides theoretical support for the compression of deep network models. Therefore, determining ways to reduce model size while retaining model precision has become a hot research issue. Model compression refers to the reduction of a trained model through some operation to obtain a lightweight network with equivalent performance. After model compression, there are fewer network parameters and usually a reduction in the computation required, which greatly reduces the computational and storage costs and enables the deployment of the model in restricted hardware conditions. In this paper, the achievements and progress made in recent years by domestic and foreign scholars with respect to model compressionwere classified and summarized and their advantages and disadvantages were evaluated, including network pruning, parameter sharing, quantization, network decomposition, and network distillation. Then, existing problems and the future development of model compression were discussed.
Abstract: Given the limitations of manufacturing precision, manufacturing cost, and other factors, contact clearance is always inevitable at the contact interface of components. The contact clearance leads to the reduction of heat flux during the heat transfer process. The effect of thermal contact resistance is significant, particularly in the fields of aerospace, microelectrical technology, and cryogenic superconductor that are closely related to the operating temperature. The thermal contact resistance is affected by many factors, such as size, shape, space of asperity, mechanical properties of the material, external pressure, and temperature. Moreover, these factors usually interact with each other and need to be coupled. Thus, how to describe the thermal contact resistance accurately and build the appropriate prediction model are the key problems that should be resolved during engineering calculations. On the basis of the research results of domestic and international scholars, the current research state of thermal contact resistance during theoretical calculations and engineering applications were presented. The theoretical calculation, experimental measurement, and digital simulation methods used to analyze macroscopic thermal contact resistance were summarized, and the advantages and disadvantages of these methods were indicated. The effects of different factors on macroscopic thermal contact resistance were briefly discussed. On the basis of the combined cooling experiments of low-temperature superconducting magnet coils of the China Fusion Engineering Test Reactor, the effects of heat flow direction, temperature, and pressure on the thermal contact resistance of superconducting magnet components, such as stainless-steel jacket, dielectric insulation material, and quench protection material, were analyzed. Moreover, the reason why the effects of temperature and external pressure result in the change of thermal contact resistance was investigated from the perspective of thermal mechanics. Finally, given the accuracy and convenience requirements for the calculation of thermal contact resistance during engineering practice, the future research direction was indicated.
Abstract: Under the action of negative temperature, the static strength of a rock increases; however, the rock will tend to be brittle, failure strain will decrease, and the rock will also bear the action of internal ice heave force, which leads to complex dynamic behavior of rocks under high strain rate loading. In addition, the geotechnical structures in cold regions are prone to sudden engineering disasters under dynamic disturbance. In this study, a dynamic impact experiment of low-temperature frozen red sandstone was carried out to investigate the temperature effect on the dynamic mechanical properties of red sandstone under high strain rate. Based on the damage theory and energy theory, the effects of different negative temperatures on the strength, damage variables, and energy dissipation of red sandstone were analyzed, and the reasons for the dynamic mechanical strength deterioration of red sandstone at lower negative temperatures were explored using fracture morphology analysis. Research shows that the low negative temperature (after -30℃) can cause a "frostbite" red sandstone, resulting in a sharp decrease in the dynamic mechanical strength of rocks under high strain rate, and transient engineering disasters can easily occur under dynamic disturbance. According to the fracture morphology analysis, the low negative temperature will cause a large number of cracks to be generated at the interface between the components in the red sandstone. The plastic deformation ability of the crack tip is poor, and the crack can easily lose stability and expand under high strain rates, resulting in the low-stress brittle failure. However, due to the complex mineral composition of the cementitious materials, they are more susceptible to negative temperature. Therefore, under the double action of dynamic load and negative temperature, the damage usually occurs first at the cementitious materials and then results in the fracture of the whole red sandstone.
Abstract: During the mining of deeply metal ore bodies, the accumulation and release of the strain energy of the surrounding rock is one of the causes of catastrophes. However, there are a large number of random distribution joints and fractures in a rock mass, which makes the evolution of strain energy more complicated and the catastrophe more difficult to predict. Therefore, five phyllites with different bedding dip angles were selected for uniaxial loading and unloading tests to investigate the effects of bedding dips on energy evolution and rock burst tendency during deformation and failure of phyllites. The strain energy evolutions of each rock sample are similar, showing energy accumulation before the peak stress and energy release and dissipation after the peak stress. However, with the increase of the bedding dip angle, the energy storage limit, residual elastic energy, and maximum dissipation energy show U-shape, and the minimum value is obtained at 60° by fitting. With the increase of the bedding dip angle, the ratio of the elastic energy of rock samples changes in an inverted U-shape before the peak, and the maximum value is obtained at 60°, indicating that the minimum work is done for bedding dip angle at 60° before peak. Moreover, the maximum elastic energy efficiency changes slightly with the increase of the bedding dip, which shows that the influence of bedding dip angle on the energy storage efficiency is small before the peak. After the peak, the decrease range of the elastic energy ratio is 60°→45°→ 30°→ 90°→0°, indicating that the post-peak fracture of the rock sample with 90° is the least developed and shows the greatest lithologic brittleness. A new criterion modified impact energy index (W) was established by combining the advantages of elastic deformation energy index (Wet) and impact energy index (Wcf). The W value of rock samples is calculated as 60°→45°→30°→90°→0° from small to large.
Abstract: With the development of liquid production and molecular synthesis technology, the application of soft particle solutions has become increasingly widespread. Soft particle solutions are also used in oil exploitation technology. The soft particles can be elastically deformed through the pores, and the whole process produces a resistance effect on flow. After breaking through the tunnel, the original shape is restored and continuously moved to the deep part of the oil layer. The soft particles do not only block the porous medium but also increase flow resistance. Moreover, they can generate deformation and break through the pores under a certain pressure to reach the depth of the reservoir. The microscopic forces mainly include Van der Waals force, electrostatic force, spatial configuration force, and surface tension. The effect of the spatial configuration force caused by the deformation of the soft particles affected by the tube wall action is considered to address the problem that micron-sized soft particle solutions in microtube deviate from the Poiseuille law. On the basis of Navier-Stokes theory, the flow velocity distribution and flow expression of the polymer solution in the tube were derived. A particle deformation factor was introduced to characterize the effect of the spatial configuration force. A mathematical model of microtube flow was established by considering the spatial configuration force. From the micro-scale flow characteristics experiment, the microtube flow in micron-sized soft particle solution was obtained. As evidenced by the results, when the tube diameter is smaller than the particle diameter, the flow velocity considering the spatial configuration force is closer to the experimental data than the Poiseuille flow under the same pressure gradient. Through the analysis of influencing factors, the spatial configuration force cannot be neglected in the microtube flow. Compared with the Poiseuille flow, the spatial configuration force increases and affects the microtube flow when the microtube size decreases. When the particles are non-spherical and the minimum projected area is the same, the greater the degree of deviation from the spherical particles and the greater the effect of the spatial configuration force.
Abstract: Metal ions have a very important influence on the flotation process of minerals, and some of them can activate minerals and thus yield an improvement in the flotation effect. Some production practices have shown that Pb2+ has an activation effect in the process of cassiterite flotation, which can improve the rate of cassiterite recovery. Styrene phosphonic acid is the most commonly used collector in the production of cassiterite flotation. In this study, the activation effect of Pb2+ in cassiterite flotation when styrene phosphonic acid is used as a collector is revealed by single mineral flotation tests. The activation mechanism of Pb2+ in the process of styrene phosphonic acid collecting cassiterite is assessed by contact angle measurement, zeta potential determination, IR spectroscopy, and solution chemistry analysis. The results of single mineral flotation tests indicate that Pb2+ can increase the floatability of cassiterite when the pH is 2.0~8.0, and at a pH of 4.0, the recovery rate of cassiterite reaches maximum, 93.78%, which is 5.33% higher than the recovery rate without Pb2+. The results of zeta potential determination, IR spectroscopy, and solution chemistry analysis show that the styrene phosphonic acid can be adsorbed on the cassiterite surface in form of chemical adsorption, causing the zeta potential of the surface to shift toward the negative direction, and the Pb2+ can promote the adsorption of styrene phosphonic acid on the cassiterite surface, making the zeta potential of the surface lower. Moreover, the Sn4+ on the cassiterite surface can be replaced with Pb2+ and the hydrolyzed species PbOH+ in solutions can interact with Sn-OH on the surface to form the surface complex Sn-O-Pb+, which may lead to an increase in the number of active sites on the cassiterite surface, promoting the adsorption of styrene phosphonic acid on the cassiterite surface and resulting in the activation of cassiterite.
Abstract: In steelmaking process, nonmetallic inclusions are often considered to be detrimental to the mechanical properties and product quality of steel as they influence the microstructure of the steel matrix to a large extent, and thus, much industrial efforts are being made to promote inclusion removal by upward flotation. From this point of view, inclusions with large size are favorable; however, quality problems or mechanical defects are more likely to happen if some of them remain in the steel. In addition, fine nonmetallic inclusions can be utilized as nucleation sites of acicular ferrite during phase transformation to improve the steel strength by promoting the formation of a fine-grained structure; this procedure is known as oxide metallurgy. In both cases, the key issue is to control the size of inclusion particles. The main factor affecting inclusion size is the collision, agglomeration, and coalescence behavior of inclusions in the molten steel. Interfacial characteristics between inclusions and steel melts are known to have a significant influence on this coalescence behavior. To analyze this influence mechanism in depth, physical and numerical simulation methods were applied to investigate the effects of inclusion type, interfacial tension, and viscosity on droplet coalescence. Based on the similarity principle, water and organic reagents were chosen to simulate molten steel and liquid nonmetallic inclusions, respectively, in the physical modeling part. The results indicate that the coalescence tendency of inclusion droplets is closely related to the physical properties of the droplets. The interfacial tension between the droplet phase and the continuous phase promotes the mutual aggregation of droplets, while the viscosity of droplets plays an inhibitory role during the aggregation process. Therefore, it is feasible to achieve aggregation or dispersion of inclusions in liquid steel by changing interfacial or viscosity parameters, thereby realizing flexible control of the inclusions particle size.
Abstract: The direct preparation of materials from high-temperature slag is an effective way for the integrated utilization of slags and their thermal energy. In this paper, with ferronickel slags and blast furnace slags as the main raw materials, glass ceramics were prepared by the Petrurgic method, a one-step heat-treatment method by direct crystallization of the slag melt during its cooling process. The ratio of ferronickel slags and blast furnace slags, Mg2+ content, and the effect of nucleating agent TiO2 on the microstructure and properties of the products were analyzed by X-ray diffraction, scanning electron microscopy, and mechanical property test. The results show that glass ceramics with excellent properties can be prepared by crystallization at 900℃ and annealing at 650℃ for slag melts during the cooling process. When the content of Mg2+ increased, and the precipitated crystal was a single-pyroxene-group mineral, the glass ceramics exhibited the highest mechanical properties. The content of pyroxene group mineral increased with the increasing ferronickel slag or MgO content. When the content of the two slags reached 90% (50% ferronickel slags and 40% blast furnace slags) with the addition of 2% MgO, the prepared glass ceramics presented a compact structure containing single-pyroxene-group minerals, including diopside, ordinary pyroxene, and clinopyroxene, and the best mechanical properties with flexural strength of 210 MPa and compressive strength of 1162 MPa. However, the further increase in ferronickel slag or MgO content led to the precipitation of forsterite, which significantly deteriorated the mechanical properties of glass ceramics. The increasing content of TiO2 caused no change in the type of crystals in the glass ceramics. Appropriate doping (2% in the experiments) increased the content of diopside, but excessive doping inhibited the crystal growth and reduced its performance.
Abstract: Al-Zn-Mg-Cu alloys are widely used due to their excellent properties. For the 7056 aluminum alloy developed on the basis of 7055 aluminum alloy, exploring its aging characteristics and the effects of rare earth elements on its microstructure and mechanical properties has a great significance to promote the use of the alloy. In this paper, the as-cast alloy is subjected to homogenization treatment, extrusion, solution treatment, and aging treatment. The effect of adding 0.2% Sc to the 7056 aluminum alloy on the microstructure and properties of the alloy was investigated by analyzing the chemical composition of the alloy, observing the microstructure of the alloy in different states, observing the precipitated phase by transmission electron microscopy (TEM) and testing the hardness and tensile properties of the alloy after heat treatment. The experimental results show that the addition of Sc significantly refines the microstructure of the grains, and the as-cast grains decrease from 100-500 μm to about 50 μm. The addition of Sc element greatly improves the plasticity of the alloy. After the aging treatment, the elongation after fracture of the alloy increased from 10.82% to 13.60%, but the yield strength reduced from 668 MPa to 657 MPa. By comprehensively calculating the grain size and precipitation phase strengthening, the reasons for the decrease of the yield strength of the peak-aged 7056 aluminum alloy were analyzed in detail. Theoretical calculations show that when 0.2% of Sc is added to the alloy, after peak aging treatment, the strength of the alloy will decrease by 12.005 MPa, which is close to the test value of 11 MPa. Through the research, the best single-stage aging system condition for the 7056 aluminum alloy was found to be 120℃+16 h, and the corresponding peak hardness and strength were 195.2 HV and 714 MPa, respectively. At this time, the main strengthening phase of the alloy was a disk-shaped and short rod-shaped MgZn2 phase, which was about 4 to 6 nm in size, and the alloy also had a spherical Al3Zr phase with a size of about 20 nm.
Abstract: Lithium-ion batteries have been widely used in various industries because of their high energy density, long life cycle, and green ring. In recent years, with the rapid development of consumer electronics, mobile wearable devices, and especially electric vehicles, the energy density requirements of the lithium-ion battery have progressively increased, promoting the development of lithium-ion batteries of higher specific capacity and longer life cycle. The commonly used graphite negative electrodes have a low theoretical capacity of 372 mA·h·g-1, which does not meet the current requirements. Silicon is a very promising lithium-ion battery anode material because of its high theoretical specific capacity of 4200 mA·h·g-1, low price, and eco-friendliness. However, silicon experiences high volume expansion (~300%) during charging and discharging, leading to severe loss of electrical contact with conductive agents and current collectors along with capacity degradation. Thus, using pitch as a soft carbon raw material and nano-Si and commercial graphite as active materials, a silicon/graphite/carbon composite was successfully synthesized using the high-temperature pyrolysis method, and micron-scale carbon fiber was formed in situ. The silicon/graphite/carbon composite material has many advantages: the void between the graphite sheet provides an effective space for the volume expansion of silicon, the coating of the asphalt pyrolysis carbon material can inhibit the volume effect in the nano-Si and increase its electronic conductivity to a certain extent, and the micro-sized carbon fiber enhances the long-range conductivity and structural stability of the material, thus greatly improving the cycle performance of the negative electrode material. The electrochemical test show that the silicon/graphite/carbon composite anode material delivers a reversible capacity of 650 mA·h·g-1 at 200 mA·g-1 and a capacity retention rate of 92.8% after 500 cycles at a current density of 500 mA·g-1. The capacity decay rate per cycle was only 0.014%, indicating excellent cyclic performance.
Abstract: To obtain a low-cost anode with low oxygen evolution potential and high catalytic activity for zinc electrowinning, Pb-0.2%Ag alloy was coated on an aluminum matrix surface by extrusion cladding technology, and a film layer with high catalytic performance was formed on the surface of the Pb-0.2%Ag alloy and Al-rod-Pb-0.2%Ag anode by anodization in a fluorine-containing sulfuric acid solution. The thickness and hardness of the film were studied using a microscopic image analyzer and digital microhardness tester, and the ultimate tensile strengths of the two anodes were compared using an electronic tensile tester. The phase, morphology, and electrochemical performance of the Al-rod-Pb-0.2%Ag and Pb-0.2%Ag anode surface film were investigated using X-ray diffractometry, scanning electron microscopy, cyclic voltammetry, anodic polarization, and electrochemical impedance spectroscopy. The results show that the Al-rod-Pb-0.2%Ag anode surface forms a dense and thick oxide film layer more easily than the Pb-0.2%Ag anode and the hardness of the film layer is increased by 41.64%; moreover, the main phase is β-PbO2, and the oxide film layer exhibits good electrocatalytic activity. The ultimate tensile strength of the new anode was 1.3 times that of the traditional anode, which greatly improves the mechanical properties of the anode material. Analytical data of anodic polarization curves reveal that the Al-rod-Pb-0.2%Ag/PbO2 anode shows low oxygen evolution potential (1.35 V vs MSE, 500 A·m-2) and high exchange current density (7.079×10-5 A·m-2) in zinc electrowinning system. Analytical data of cyclic voltammetry and EIS curves indicate that the Al-rod-Pb-0.2%Ag/PbO2 anode has higher electrocatalytic activity, larger surface roughness, and smaller charge transfer resistance. In the zinc electrowinning experiment, the average cell voltage of the fence-like Al-rod-Pb-0.2%Ag/PbO2 anode was 75 mV less than that of the traditional Pb-0.2% Ag anode, and the production of anode slime was greatly reduced.
Abstract: As science and high-tech have developed, stealth technology has gained increasing prominence in the military field. Application of stealth technology can improve the survival, defense, and attack capabilities of military equipment, thus it has become a focus in the field of modern military science. As the core part of radar stealth technology, absorbing materials are widely required by various industries. For military equipment such as ships operating in the marine environment, absorptive coating can not only make the military equipment effectively invisible, but can also enhance the corrosion protection capability of the equipment itself. Once the surface of an absorptive coating is corroded, not only will its corrosion resistance become compromised, but its absorbing performance may also be affected, leading to threats and hidden dangers to the safety of the weapons and equipment. At present, most researchers are paying more attention to the effect of absorbent particles on absorbing properties during studies of absorptive coatings. However, after addition of absorbent particles, the effect of the absorptive coating on a material's absorbing properties is unknown when corrosion resistance is constantly changing. Therefore, research in this area is of great significance in selection of surface absorbing coatings for marine weapons and equipment. In this study, FeSiAl electromagnetic shielding coating, based on Q235 cold-rolled steel, was used as the experimental material. By changing curing conditions, the optimal curing environment for electromagnetic shielding coating was explored. At the same time, the neutral salt spray test, electromagnetic shielding performance test, and electrochemical impedance test were applied to study the variations in absorption and corrosion resistance of the coating after curing in natural conditions during the salt spray period. Results show that curing under an electromagnetic field can impair the corrosion resistance of the coating. Increasing the content of the absorbing agent was not conducive to improving the absorbing properties of the coating, and impaired the corrosion shielding properties of the coating. After the long-term salt spray test, absorbing properties of the coating decreased with decreasing corrosion shielding properties.
Abstract: To cope with the increasingly stringent emission regulations, major automobile manufacturers have been focusing on the development of new energy vehicles. Fuel-cell vehicles with advantages of zero emission, high efficiency, diversification of fuel sources, and renewable energy have been the focus of international automotive giants and Chinese automotive enterprises. Establishing a reasonable energy management strategy, effectively controlling the vehicle working mode, and reasonably using battery energy for hybrid fuel-cell vehicles are core technologies in domestic and foreign automobile enterprises and research institutes. To improve the equilibrium between fuel-cell hydrogen consumption and battery consumption and realize the optimal energy distribution between fuel-cell systems and batteries for plug-in fuel-cell electric vehicles (PFCEVs), considering vehicles as the environment and vehicle control as an agent, an energy management strategy for the PFCEV based on reinforcement learning algorithm was proposed in this paper. This strategy considered the immediate return and future cumulative discounted returns of a fuel-cell vehicle's real-time energy allocation. The vehicle simulation model was built by Matlab/Simulink to carry out the simulation test for the proposed strategy. Compared with the rule-based strategy, the battery can store a certain amount of electricity, and the integrated energy consumption of the vehicle was notably reduced under different mileages. The energy consumption in 100 km was reduced by 8.84%, 29.5%, and 38.6% under 100, 200, and 300 km mileages, respectively. The hardware-in-loop-test was performed on the D2P development platform, and the final energy consumption of the vehicle was reduced by 20.8% under urban dynamometer driving schedule driving cycle. The hardware-in loop-test results are consistent with the simulation findings, indicating the effectiveness and feasibility of the proposed energy management strategy.
Abstract: Compared with manned aircraft, unmanned aerial vehicles (UAVs) are affordable and convenient for high-risk missions. Therefore, UAVs are increasingly playing an important role in military and civilian fields. Today, UAVs have been exploited to perform special missions carrying some important equipment. However, influenced by the constraints of single UAV's performance and load, it has become a research hotspot that multi-UAVs perform search cooperatively. The process is to minimize the uncertainty of the unknown area and to find the target as much as possible. In terms of cooperation among UAVs, the more effective method based on search map is used. And search process optimization on the basis of distributed model predictive control (DMPC) or traditional swarm intelligence algorithms are adopted, but these methods have some limitations. Due to the behavior of swarm intelligent individual have the characteristics of the decentralization, distribution, and overall self-organization, which match with the requirements of the localization, distribution and robustness of the UAV cooperate search. Nevertheless, the traditional swarm intelligence algorithms have low search efficiency and are easy to fall into local optimum. To solve the problem of repeated search, static targets and low efficiency in cooperative search for multi-UAVs, a method based on improved pigeon-inspired optimization (PIO) and Markov chain was proposed. Firstly, a honeycomb environmental model similar to the sensor detect region was established to reduce repeated search for the area. Secondly, Markov chain with the Gaussian distribution was used to represent dynamic movement of targets. Thirdly, Cauchy mutation and Gaussian mutation were introduced into the map and compass operator and the landmark operator of PIO, respectively. Meanwhile, simulated annealing (SA) mechanism was exploited to reserve the worse individual, which effectively reduced the problem that PIO was easy to fall into local optimum. Finally, the algorithm was compared with other swarm intelligence algorithms through simulation experiments. The results show that the new method is effective and available.
Abstract: The operational environment of offshore wind turbine towers is complex, and the harsh service environment makes them more vulnerable to damage under conditions of complex stress such as sea water scouring. The scouring pit has a great influence on the vibration of wind turbine towers. It is of great importance to study the dynamic response to earthquakes of wind turbine towers under scouring depths. The research object of this study was a wind turbine tower at a wind farm in Jiangsu Province, which had a seven-degree seismic fortification in the area. Based on finite element simulation, on-site monitoring and a shaking table test of the offshore wind tower, and considering pile-soil interaction in a refined model, variation in the natural vibration period of the structure under different scouring depths and the influence of different scouring depths on the dynamic response of the structure under seismic excitation were studied. On-site monitoring results show that the #6 wind turbine structure is seriously eroded by sea water, and the vibration amplitude is clear compared with the #15 wind turbine built in the same period. These aspects indicate that the influence of scouring depth on the structure could not be ignored. Analysis of numerical simulation show that scouring depth has a great influence on the high-order mode of the structure, which lengthen the natural vibration period of the structure by a maximum of 33%. On account of scouring, constraints of the soil layer on the highly flexible structure were weakened, and the structure produced considerable vibration, which could lead to damage of structures such as wind towers. Results also indicate that when encountering a seven-degree rare earthquake, power generation should immediately be stopped. The variation curves of the shaking table test and the numerical simulation results were more uniform, and the trend coincided well, which fully verified the accuracy of the numerical simulation.
Monthly, started in 1955 Supervising institution：Ministry of Education Sponsoring Institution：University of Science and Technology Beijing Editorial office：Editorial Department of Chinese Journal of Engineering Publisher：Science Press Chairperson：Ren-shu Yang Editor-in-Chief：Ai-xiang Wu ISSN 2095-9389CN 2095-9389
