Abstract: The global steel yield heavily relies on the continuous-casting process in the modern steel industry. Thus, the implementation of protected teeming in continuous-casting tundish for high-quality clean steel production is essential. Ladle shroud is a refractory device that contains the teeming stream between ladle and tundish. The invention and application of the ladle shroud play a significant role in the development of continuous-casting technology and considerably influence the performance of protected teeming, including protecting molten steel from reoxidation and contamination from air/slag/refractory/ladle filler sand during both steady-state and transient casting periods (i.e., first-heat filling, ladle change, and tundish emptying). In this review, the sources and forms of atmospheric contamination of molten steel in the tundish were addressed and the possible solutions to the problems encountered during the use of the ladle shroud were proposed by considering the invention, industrial trials, and progressive optimization in the early stage of the development of continuous casting. The ladle shroud has been proven to effectively protect the teeming stream, which, however, is closely associated with the structural design and operating practice. Thus, the effect of the structural design of the ladle shroud, including two types of industrialized ladle shroud and several new designs, on protected teeming was analyzed and the advantage of the trumpet-shaped ladle shroud over the conventional ladle shroud in terms of production efficiency and molten steel cleanness was emphasized. The materials of the ladle shroud have also been progressively enhanced to prolong its service life and achieve a stable service performance. The influence of operational parameters, including the immersion depth of the ladle shroud in tundish and misalignment, was also discussed. On the basis of current research and development work in the steelmaking continuous-casting field, the future development direction of the ladle shroud was identified to be structure–function integration, characterized by long service life, lightweight, multifunction, and eco-friendly manufacturing.
Abstract: With the rapid development of the automobile industry, the development and application of lightweight automobile steel are increasingly extensive. The second- and third-generation automobile steels with a tensile strength of over 1000 MPa are usually of duplex structure. Through solid solution strengthening, precipitation, deformation, fine grain strengthening, and other strengthening methods, a large number of defects are formed in the matrix, which makes the steel more sensitive to hydrogen in the service process and prone to hydrogen embrittlement under very small hydrogen dissolution conditions. The high-Mn content steels Fe?Mn?C and Fe?Mn?Al?C steels have high stacking fault energy, which not only influences their strength and toughness but also significantly affects their service performance. Based on the composition of twinning-induced plasticity (TWIP) steel of the Fe?Mn?C system, adding a small amount of Al element to form Fe?Mn?(Al)?C steel can not only reduce the steel density and improve the steel strength and toughness but also change the steel microstructure to a certain extent; the effect on the microstructure reduces the steel susceptibility to hydrogen embrittlement. However, when the Al content is high, low-density steel with a more complex structure is formed, and the precipitates are more, which leads to a more significant sensitivity to hydrogen embrittlement. In this paper, the permeation, dissolution, and diffusion behavior of H in Fe?Mn?(Al)?C high-strength-and-toughness-steel; the interaction between H and the matrix structure, the precipitated phase, and lattice defects; the model of H in steel; the hydrogen embrittlement mechanism; and the methods of hydrogen embrittlement evaluation were summarized based on the structure, second phase, and crystal defects of Fe?Mn?(Al)?C high-strength-and-toughness steel. The related research work and the latest developments of the hydrogen embrittlement of Fe?Mn?(Al)?C high-strength-and-toughness steel were reviewed. The development direction of the hydrogen embrittlement microstructure mechanism of high-strength-and-toughness steel was revealed by combining first-principle calculations, molecular dynamics simulation, and physical experiments such as hydrogen atom microprinting technology and three-dimensional atomic probe analysis.
Abstract: Phosphorus slag (PS) is an industrial waste discharged during the refining of yellow phosphorus by a high-temperature electric furnace. The slag is generally deposited on the surface in situ, and this uses up considerable areas of land and causes environmental pollution problems. The glass phase content in electric furnace PS is over 90%, indicating that the PS has potential cementitious property. As a cement admixture, PS has a retarding effect on cement, thus, its application to cement has become less adopted. Based on the hydration characteristics of PS and the application conditions of mine filling, the feasibility of alkali-activated cement–PS used as cementitious material (CPCM) to cement sulfur tailings was studied. Quicklime, NaOH, and Na2SiO3 were used as the activators to investigate the CPCM curing performance in a laboratory, and the compatibility of CPCM with sulfur tailings was evaluated. The results show that when the PS-to-cement mole ratio is 1∶1, quicklime is 3% of PS, the final setting time of CPCM is 300 min, and the CPCM strength for 28 days is 40.6 MPa, indicating that the material can replace P.O 42.5 cement in the subsequent filling process. When the Na2SiO3 is 4% of PS, the final setting time of CPCM is 39.3% less than that of cement, and its strength for 7 days is 31.1% higher than that of cement. Compared with cement, CPCM shortens the setting time of backfill by 8 hours, and its strength does not deteriorate after curing for 28 days. Therefore, CPCM is suitable for cementing sulfur tailings. X-ray diffraction (XRD), scanning electron microscopy (SEM), and energy-dispersive X-ray spectroscopy (EDS) analyses show that the CPCM hydration products are mainly Ca(OH)2 and C?S?H. Moreover, the PS hydration resulted in a decrease in the Ca(OH)2 content, and a compact C?S?H gel with a low Ca/Si mole ratio is formed. This not only improves the later strength of CPCM but also prevents the strength deterioration of the backfill.
Abstract: Heap leaching is a widely used solution mining technology that enables various kinds of low-grade ores to be processed economically. The solution flow characteristics are very important factors in the leaching process, and they influence both the overall recovery and kinetics of the system. The properties of fluid flow in porous media are associated with the pore structure, which is influenced by the grain size and shape. To study the influence of ore particle size gradation on the pore structure of the heap leaching system, a micro-CT scanning test was conducted in ore columns with two grain size gradation types, and images of the internal structure of the leaching columns were obtained. A 3D digital pore model of the two leaching columns was then established, and the spatial distribution characteristics of 2D and 3D porosity were analyzed. The pore network models of the two columns were then extracted from the reconstructed 3D binary pore structures using the maximal ball fitting method, and the effect of the ore particle size distribution on the pore throat radius, throat length, pore throat volume, shape factor, and coordination number was analyzed. The results show that the porosity of the column comprising well-graded ore particles is lower than the column with uniformly graded grains. In addition, the 2D and 3D porosities of the well-graded ores show a relatively high degree of heterogeneity compared to those of the more uniformly graded ores. The ore particle size gradation has a significant influence on pore size and pore connectivity, but it has a minimal influence on the pore throat shape factor. The number of large pores increases with a decrease in the amount of fine ore, and the pore throat radius, throat length, and pore throat volume also correspondingly increase. When the uniformity of ore particle gradation is enhanced, the proportion of isolated pores decreases and the proportion of the number of high coordination pores increases.
Abstract: Deep-cone thickening of unclassified tailings is one of the key technologies in the field of cemented paste backfill. Flocculation and settling behavior of unclassified tailings constitute key research topics of the deep-cone thickening technology. Based on the measurement of the floc chord length during the unclassified tailings flocculation process, this study investigated the flocculation and settling processes independently, which is different from the traditional research. First, the average chord length of the floc was used as the index to study the flocculation behavior of the unclassified tailings under different conditions. Then, the initial settling rate of the suspension–supernate interface was used as the index to analyze the settling behavior of the unclassified tailings slurry under varying flocculation conditions. Under different flocculation conditions, the unclassified tailings particles were flocculated rapidly and the average chord length of the floc increased rapidly to the peak value. Then, it decreased gradually with the shear time until it reached a stable state. It was found that the average chord length of the flocs of the unclassified tailings and the initial settling rate of the suspension–supernate interface of the flocculated, unclassified tailings slurry vary with flocculation conditions. For the scope of this study, the optimal flocculation conditions were determined as follows: the flocculant used was Magnafloc 5250, the solid mass fraction was 10%, the flocculant dosage was 10 g·t?1, the flocculant mass fraction was 0.025%, and the shear rate was 94.8 s?1. Under such flocculation conditions, the peak value of the average chord length of the floc was 620.63 μm, the average chord length after the flocculation was 399.57 μm, and the initial settling rate of the suspension–supernate interface of the flocculated tailings slurry was 4.61 mm·s?1. The initial settling rate model of the suspension–supernate interface, applicable only to the tailings used in this study, was established preliminarily based on the average chord length of the flocs. The initial settling rate of the suspension–supernate interface increased with the increase in the average chord length of the flocs, providing a reference for the control of flocculation and settling parameters and the optimization of the equipment structure to improve the flocculation settling efficiency of unclassified tailings slurry in actual production.
Abstract: In view of the influence of the load contact conditions on Brazilian test results, the acoustic emission (AE) monitoring system was used to conduct a Brazilian test of hard and brittle low-porosity sandstone under linear/non-linear load contact conditions. The standard Brazilian discs with a diameter of 50 mm and a thickness of 25 mm were instrumented with a three-dimensional sensor array containing eight Nano30 sensors. All the discs were equipped with identical three-dimensional sensor arrays. At the same load rate, the Brazilian discs were quasi-statically loaded under both linear/non-linear loads. The Richter 8 acquisition system continuously recorded waveform signals from eight channels from load application to brittle failure. Under the linear/non-linear load conditions, 1131 and 931 AE events were successfully located by a P-wave automatic picking and collapsing grid search algorithm. Under the linear/non-linear load condition, the crack initiation points were both away from the disc center. For non-central crack initiation, the tensile strength test may underestimate the true value. A pole density analysis of the planes under nonlinear load conditions shows that the local distortion of the fracture is greater than that under linear load. The evolution of the 3D damage to the disc shows that the load area of the disc significantly affects the cumulative time of damage, amount of energy liberation and stability of the crack propagation. The moment tensor decomposition was performed on the effective AE events, and the isotropic (ISO) component, the pure double-coupled (DC) and the compensated linear vector dipole (CLVD) component frequency percentage were obtained. The classification method was applied to quantitatively analyze the focal mechanism. The results show that the Brazilian test is not sensitive to the load contact conditions, and the focal mechanism of both cases can be interpreted as the initiation, propagation, and penetration of the tensile and shear microcracks approximately along the load direction.
Abstract: Mercury, a heavy metal, can seriously harm human bodies and the environment due to its characteristics of high toxicity, biological enrichment, and long-range migration. The non-ferrous metal smelting industry is one of the main sources of atmospheric mercury pollution in China. Therefore, controlling atmospheric mercury emissions from non-ferrous smelting plants is very important. The wet cleaning process has been widely applied in the purification of smelting flue gas because of its advantages such as a high removal efficiency, stable operation, and low cost. During the wet purification process, thiourea is usually added because it can reduce the oxidation potential of mercury and react with mercury to form stable coordination ions, resulting in the high-efficiency removal of mercury from high-sulfur smelting flue gas. However, mercury recovery from scrubbing solutions containing mercury and thiourea obtained from the wet cleaning process is difficult. In this study, a novel technology to recover mercury from the thiourea mercury solution via electrodeposition was proposed and investigated. The linear potential scanning method was applied to obtain the reduction potential of mercury. It was determined that the optimal potential of the mercury electrodeposition process should be controlled between ?0.55 V and ?0.45 V because the presence of ferric ions, copper ions, and sulfite ions did not seriously affect the electrodeposition of mercury. Controlled potential electrolysis was employed to efficiently recover mercury from thiourea mercury solution, and the effects of key parameters, including electrolyte type and concentration, electrolyte temperature, stirring rate, and electrolytic time, on the mercury recovery efficiency were explored. The optimal process conditions are as follows: a cathode material of copper sheet, electrolyte of 0.24 mol·L?1 Na2SO4, electrolyte temperature of 30–40 ℃, stirring speed of 100–300 r·min?1, $ {\rm{SO}}^{2-}_{3} $ concentration of 8 mmol·L?1, and electrolytic time of 5 h. Under the optimal process conditions, the mercury recovery efficiency mercury is over 98%. The electrolytic products on the cathode are elemental mercury, and the corresponding purity is over 99%.
Abstract: Titanium alloys are widely used in aviation industry because of their high specific strength, corrosion resistance, and heat resistance. They are widely used in aircraft engine compressor to improve the thrust-to-weight ratio of an aircraft engine. However, they are easily burning because of their low thermal conductivity and high combustion heat. Under some conditions, titanium blades rubbing with their casees to generate a large amount of heat, and finally burns. To meet the requirements of advanced aero engines and prevent the burning of titanium alloys, we must understand the mechanism of titanium alloys combustion. In this study, TC4 titanium alloys coated with Cr coatings with different thicknesses (0, 15, 30, and 60 μm) were subjected to oxygen-enriched atmosphere under different oxygen pressures. The effect of chrome plating thickness on the combustion properties of TC4 titanium alloys was reported, and microstructure analyses were carried out through SEM, EDS and XRD. Results show that chrome plating thickness has no obvious effect on the critical oxygen pressure of TC4 when the Cr layer thickness is less than 30 μm. The pressure threshold of TC4 increases from 0.07 MPa to 0.15 MPa, when the Cr layer thickness increases to 60 μm, which is about two times higher than the pressure threshold of the substrate. Burning velocity decreases as the Cr layer thickness increases, indicating that a thick Cr layer can effectively inhibit the flame propagation speed. In the underlying action mechanism during combustion, surface Cr enters the molten pool via diffusion and melting and precipitates with Al and V in the alloy to form a Cr-, Al-, and V-rich dispersion cloth phase. The combination of Al and O is reduced, thereby hindering of O diffusion and reducing the burning rate.
Abstract: Supercapacitors are usually used in new energy storage devices, communication technology, military, and aerospace fields due to their long lifecycle and high power density. Presently, it is imperative to find the electrode materials with low cost and excellent capacity. MXenes have received increasing attention due to their unique physical and chemical properties. They not only have superior electrical conductivity but also contain abundant surface groups (?OH, ?F or ?O); therefore, they are regarded as versatile 2D materials. MXenes can generate higher volumetric capacitance than that of graphene. However, MXene nanosheets are inclined to stack together, limiting the electrochemical properties of supercapacitors. In this work, an MXene (Ti3C2Tx) was obtained by etching an MAX (Ti3AlC2) phase using HF. To expand the interlayer spacing of Ti3C2Tx, the liquid-phase intercalation method was adopted. After the interlayer spacing was expanded, V2O5 nanosheet (NSV) and V2O5 nanobelt (NBV) were loaded on the MXene surface by a facile hydrothermal process. Their structure and morphology were characterized using different techniques, such as X-ray diffraction, Brunauer–Emmett–Teller surface area measurements, and field-emission scanning electron microscopy. The results show that the interlayer spacing of MXene is increased after liquid-phase intercalation, and NSV and NBV are uniformly loaded on the MXene surface. Moreover, the specific surface areas of the NSV/MXene and NSV/MXene nanocomposites are higher than that of the MXene; therefore, the nanocomposites can provide more active sites for electrochemical reactions. The electrochemical performances of the nanocomposites were investigated in 1.0 mol·L?1 Na2SO4 and 1.0 mol·L?1 LiNO3 aqueous solutions. The specific capacitances of V2O5, MXene, NSV/MXene, and NBV/MXene are 8.1, 15.7, 96.8, and 88.5 F·g?1 in 1.0 mol·L?1 Na2SO4, respectively. When they are tested in 1.0 mol·L?1 LiNO3, their specific capacitances are 64.6, 46.7, 180.0, and 114.0 F·g?1, respectively. Therefore, the NSV/MXene nanocomposite is a potential electrode material for supercapacitors.
Abstract: With the growing of CO2 corrosion problem in multiphase oil and gas in-field pipelines, carbon steel can no longer meet the continuously growing demand for energy consumption. At the same time, the water content in the gathering pipelines and the complex phase distribution of the oil and water phases make the service environment of the pipeline steel increasingly demanding. Recently, the low Cr-containing steel, which shows an excellent performance-price ratio with a better CO2 corrosion resistance, is expected to replace the carbon steel used for pipelines. However, the application of 3Cr is limited under the conditions of oil-water flows, especially those with corrosion inhibitor. For example, the absolute value of the uniform corrosion rate is still relatively high in environments of high-carbon dioxide, and using corrosion inhibitor in the application of Cr-containing low-alloy steels is still necessary. Some researchers found that the corrosion inhibitor of imidazoline quaternary ammonium salt can better control the corrosion caused by carbon dioxide in the application of 3Cr steel. Since the corrosion resistance of Cr-containing low-alloy steel depends on the formation of corrosion products, it is highly susceptible to corrosion inhibitors, and research on its compatibility with corrosion inhibitors is still lacking. In this study, the corrosion resistance of 3Cr steel and the effect of corrosion inhibitor on the resistance were evaluated in an oil-water two-phase environment by using a high-temperature and high-pressure autoclave combined with SEM (scanning electron microscope), XRD (X-ray diffraction), confocal Raman spectroscopy, and electrochemical impedance spectroscopy. The results show that the corrosion scales formed on the 3Cr steel consist of two layers, and the inner layer is a Cr-rich layer in this environments, exhibiting good resistance to CO2 corrosion under the conditions of oil-water flows. However, after adding 100 mg·L?1 corrosion inhibitor of seventeen alkenyl amide ethyl imidazoline quaternary ammonium salt, 3Cr steel has not been effectively protected from corrosion. The analysis of the corrosion product and electrochemical tests revealed that competition exited between alkane molecules, corrosion inhibitor molecules and Cr-rich layers and the alkanes interfered with the ordered arrangement of the corrosion inhibitor and thus affected the corrosion resistance of 3Cr steel.
Abstract: With the development of hypersonic technology, the demand for thermal protection material is continuously increasing. Carbon/carbon composites are widely used as thermal protection materials in the nose and in the leading edge of hypersonic vehicles owing to their high latent heat and good resistance to high temperatures. The flow field around the aircraft affects the heat transfer and ablation of carbon/carbon composites, changing the thickness and shape of the thermal protection layer. The ablation of carbon/carbon composites alters the flow field distribution, thus conversely affecting the ablation of carbon/carbon composites. To predict the heat transfer and ablation of carbon/carbon composites, a multi-field coupling model was established to predict the transient temperature distribution, ablation rate, and ablation profile of carbon/carbon composites in hypersonic aerothermal environments. The thermochemical non-equilibrium effects of the flow field, heat transfer of the material, and ablation of the material surface were considered in the modeling. The wall temperature and heat flux in the stagnation area change significantly. The initial heat flux is higher and the stagnation heat flux at 1 s is 17.22 MW?m?2. As time passes, the wall temperature increases, the temperature gradient in the stagnation area decreases, the heat flux decreases, and the stagnation heat flux at 30 s is 10.22 MW?m?2. As the temperature of the stagnation area is high, the material at the surface reacts actively and the ablation is more serious, whereas only a small amount of ablation occurs on the side of the model. The shape of the material model changes after the ablation, the leading-edge radius increases, and the ablation depth at the material stagnation point is 17.47 mm at 30 s. The results show that, in the hypersonic aerodynamic thermal environment, the carbon/carbon composites have a certain ablation recession, which leads to change in the external flow field and thermal load. The multi-field flow-heat-ablation coupling model can be used to predict the response of thermal protection materials, which can provide some reference for the design of thermal protection systems.
Abstract: Planetary gearboxes have one or several planet gears rotating around the sun gear while revolving along their axle. This unique gear structure results in the simultaneous meshing of the planet gear with both sun and ring gears. Because of the high transmission ratio and large bearing capacity of its compact structure, planetary gearboxes have been extensively used in a variety of industrial applications. Therefore, planetary gearbox fault diagnosis is essential to ensure safe and efficient industrial manufacturing. Acoustic signal analysis provides an effective and noninvasive method for detecting potential faults in the planetary gearbox. However, the theoretical foundation of planetary gearbox fault signatures in acoustic signals is ambiguous. In this work, the planetary gearbox acoustic signal model of the resonance frequency region under the nonstationary state is structured by amplitude and frequency modulation, and the gear fault characteristics of the acoustic signals are explicitly derived. Given that resonance frequency is independent of rotational speed, the resonance frequency can be distinguished from speed-related frequency components. This lays the foundation for extracting the gear fault characteristics of the resonance frequency region. Moreover, the planetary gearbox often runs under time-varying speed conditions, and the fault frequency components are time-varying. To overcome the limitations of the traditional time–frequency analysis method in limited time–frequency resolution or cross-term interferences, the appropriate time–frequency analysis method is essential. In this work, the high-order synchrosqueezing transform is exploited to identify the time-varying fault characteristics of the planetary gearbox acoustic signal. Owing to the step of squeezing the energy distributed along instantaneous frequency in frequency direction, time–frequency representation by synchrosqueezing transform achieves a high time–frequency resolution. The high-order interpretation of instantaneous frequency further improves the capability to capture the time–frequency details. The acoustic signal model and corresponding fault characteristics of the planetary gearbox in the resonance frequency region are verified by both numerical simulations and laboratory experiments. The gear defect within the planetary gearbox is successfully diagnosed via the high-order synchrosqueezing transform.
Abstract: Transmission error in gear system and backlash are important factors that affect the accuracy of precision transmission systems. The main sources of the complete cycle transmission errors and periodic backlash are eccentric errors due to imperfections in machining and assembling. Therefore, analyzing the transmission error in gear system and backlash under the effect of eccentricity error is necessary. Scholars from around the world have conducted extensive research on the transmission error of the gear system. They observed that the back-side contact of the gear tooth generally happens in high-speed light-load conditions, or when alternate load torque occurs or anti-backlash gear has been used. However, the existing literature does not evidently show the calculation method of the transmission error of the back-side tooth mesh. Therefore, analyzing the transmission error of the back-side tooth mesh is of great significance. The present work develops a calculation model of drive- and back-side (two-sided) transmission error of an involute gear pair with eccentricities to construct the equivalence of time-varying backlash calculation formula to demonstrate the continuous measurement of gear backlash based on two-sided transmission error that can also be realized in theory. The experimental transmission error and the corresponding hysteresis were acquired under various load torques and different initial conditions, such as initial position of the gear unit. The continuous backlash curve of the gear pair was obtained from the two-sided transmission curves, and the backlash of the whole cycle could be predicted. The result of the experiment shows that the continuous measurement curve agrees with the mechanical hysteresis method, and the prediction of backlash perfectly reflects the variation range and trend of backlash. Moreover, both the continuous measurement method of backlash and prediction of backlash demonstrate that the model is practicable and have higher efficiency, and the overall data retain instructional significance, reference value for nonlinear research, anti-backlash control, and gear accuracy of gear transmission.
Abstract: The adder circuit is the core component of the high-performance system-on-chip (SoC). It is also important in image and voice encryption. The full adder circuit is a basic unit with a very high reuse rate among all the units. Therefore, the design of an adder with high energy efficiency is of great significance for the optimization of digital circuit systems. In recent years, numerous researchers have studied the design of advanced adder circuits, which are characterized by high speed and low power consumption. To reduce the hardware overhead, an increasing number of adder circuits utilize the transmission tube logic to reduce the number of transistors. However, this method also brings about several negative effects, such as threshold loss and performance degradation. In this paper, by studying the swing recovery logic and full adder circuit, we proposed a full adder design scheme based on swing restored pass-transistor logic (SRPL). First, the threshold loss mechanism of the circuit was analyzed, and the characteristics of the high-efficiency transmission of high-level and low-level transistors were considered; then the design method of the swing recovery transmission tube logic was developed. We used a symmetric structure to design an XOR/XNOR circuit without delay deviation output. The two-shot MOS tube was used to compensate the threshold loss to realize the full swing output of the XOR/XNOR circuit. Finally, we fused the designed XOR/XNOR circuit to the full adder structure and used the 4T XOR sum circuit and the improved transmission gate carry circuit to implement the high-performance full adder for swing recovery. In the TSMC 65 nm process, the logic function of our method was verified by HSPICE simulation. Compared with the conventional approach, the delay is reduced by 10.8%, and the power-delay product (PDP) is reduced by more than 13.5%. The design method of low delay and full swing output of the SRPL circuit can be further applied to the design of other logic circuits, further promoting the practical process of the SRPL circuit.
Abstract: Traffic sign detection and recognition, which are important to ensure traffic safety, have been a research hotspot. In recent years, with the rapid development of automated driving technology, significant progress has been made in developing more accurate and efficient deep learning algorithms for traffic sign detection and recognition. However, these studies mainly focus on foreign traffic signs and do not consider the low-illumination conditions in practical application, which is a common scene. Therefore, many challenges still exist in the application of traffic sign detection and recognition in traffic scenes. To solve the problems of easy omission and inaccurate positioning for traffic sign detection and recognition under complex illumination conditions, the enhanced YOLOv3 (You only look once) detection algorithm, a traffic sign detection and recognition method combining real-time adaptive image enhancement and the YOLOv3 frame was proposed. First, a large and complex illumination traffic sign dataset for Chinese traffic was constructed; it included globally low illumination, locally low illumination, and sufficient illumination images. Then an adaptive enhancement algorithm was proposed for low-illumination images, which can enhance the difference between traffic signs and background by adjusting the brightness and contrast of the images. Finally, high-quality and discrimination images as input were transmitted to the YOLOv3 network framework, and traffic sign detection and recognition were performed. To reduce the influences of the prior anchor box setting accuracy and the imbalance between the background and foreground on the detection accuracy, the clustering algorithm for the prior anchor box and loss function for the network were optimized. The results of the comparison experiment with the LISA dataset and complex illumination traffic sign dataset for Chinese traffic show that the proposed enhanced YOLOv3 detection algorithm has higher regression accuracy and category confidence than the published YOLOv3 algorithm for traffic signs; the recall and precision are higher by 0.96% and 0.48%, respectively, which indicates the application potential of the proposed algorithm in actual traffic scenarios.
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
