Abstract:

In view of the problem that most of the existing researches on automatic driving ignore the friction performance of pavement， five kinds of asphalt mixture rutting plate specimens were prepared in this paper. Based on Persson’s surface fractal friction theory and the 3D tire-pavement finite element model， dynamic friction coefficient and adhesion coefficient of asphalt pavement were calculated to characterize the friction performance. Matlab/Simulink software was used to establish the dynamics control module of autonomous vehicles， and the braking pressure of the wheel cylinder was solved by reverse calculation based on the desired braking speed and pavement friction performance of the vehicle to realize the braking process of the vehicle. Finally， CarSim and Matlab/Simulink were used for co-simulation， setting downhill braking and curve braking conditions， and analyzing the influence of factors such as longitudinal slope， curve radius， and pavement superelevation on the braking performance of autonomous vehicles.

Abstract:

This paper proposed an accelerated test method for intelligent driving system based on the surrogate genetic optimization model. First， the Latin hypercube sampling interval in the parameter sampling module was improved by using the weights and optimal solution region features of the scenario element hierarchical analysis method， achieving a synergistic improvement in sampling efficiency and optimization effect. Next， by utilizing parameter sampling results and repeatability screening mechanism， the population diversity of the genetic optimization module was increased， overcoming the local convergence problem of traditional genetic algorithms. Then， the surrogate filtering module based on cyclic update mechanism was used to predict the test results of the scenario， which balanced the contradiction between the efficiency and accuracy of the accelerated algorithm and the application of the surrogate model. Finally， a simulation platform was built to accelerate test process and verification of the intelligent driving system to be tested in the front vehicle speed change scenario of high-dimensional time series decomposition. The results indicate that the method proposed in this paper can effectively search for a large number of key scenarios and improve testing efficiency.

Abstract:

This paper， extracting the scenario feature indexes and risk metrics index from vehicle kinematic status parameters and road infrastructure condition parameters， uses the extreme gradient boosting （XGboost） model and the long short-term memory （LSTM） model for safety risk prediction. Then， it proposes an adaptive cruise control （ACC） optimization method of automatic driving based on safety risk prediction. It selects collision probability， average speed， and standard deviation of speed to evaluate the performance of ACC optimization， and verifies the rationality and effectiveness of the ACC optimization method proposed using Prescan-Simulink co-simulation. The results show that the safety risk-based ACC optimization method is superior to the general ACC. Compared with the XGboost， the LSTM as safety risk prediction model， has a better performance for ACC optimization. The addition of road infrastructure condition parameters for safety risk prediction improves the ACC performance and reduces the collision probability of automatic driving significantly.

Abstract:

This paper provides a comprehensive overview and summary of the application of reinforcement learning in the field of autonomous driving. First， an introduction to the principles and development of reinforcement learning is presented. Following that， the autonomous driving technology system and the fundamentals required for the application of reinforcement learning in this field are systematically introduced. Subsequently， application cases of reinforcement learning in autonomous driving are described according to different directions of use. Finally， the current challenges of applying reinforcement learning in the field of autonomous driving are deeply analyzed， and several prospects are proposed.

Abstract:

A traffic target detection algorithm， fusion attention adaptive pyramid network （FAAP-Net）， is proposed to address the issues of slow speed and low accuracy in traditional vehicle detection techniques， significantly reducing the occurrence of traffic accidents. To mitigate computational complexity， a lightweight complementary pooling structure （CPS） is designed， employing two sets of different pooling combinations in width and height， which maintains a high precision while significantly reducing the floating point operations per second （GFLOPs） and the parameter count of the network. Addressing the information loss during intelligent traffic system feature map generation， the adaptive fusion feature pyramid network （AF-FPN） incorporates the adaptive attention module （AAM） and the feature enhancement module （FEM） to integrate shape features for vehicle detection. Lastly， to address the weak representation of vehicle detail features， a channel-wise grouped attention （SA） mechanism is introduced， enhancing the focus of the backbone network on various vehicle detection details and effectively extracting significant features. The experimental results on the BDD100K dataset demonstrate that the FAAP-Net algorithm achieves a notable improvement， increasing the average precision from 30.3 % to 43.7 %.

Abstract:

In response to the limitations of current multi-object tracking algorithms in handling roadside traffic scenarios， a multi-object tracking algorithm based on roadside cameras was proposed in this paper. First， the one-shot tracking algorithm chosen and a neural network based on FairMOT was built to simultaneously generate both object detection results and appearance feature results， thereby enhancing the real-time performance of the algorithm. Then， a novel data association method was adopted to lessen the effect of occlusion on the tracker. After that， a new motion similarity measurement called buffered intersection over union was introduced to compensate for the errors caused by linear motion prediction models. Subsequently， a velocity-based discriminative algorithm for removing lost trajectories and a history-based position matching algorithm to retrieve the identities of occluded trajectories over lengthy periods of time were developed. Experiments were conducted on the UA-DETRAC public multi-object tracking dataset to verify the effectiveness of the algorithm. Additionally， to demonstrate the applicability of our algorithm in real-world roadside environments， real roadside scene data were collected on open field in the National Intelligent Connected Vehicle （Shanghai） Pilot Demonstration Zone. Finally， comparative experiments between the algorithm proposed and SORT， DeepSORT， ByteTrack and FairMOT algorithms were conducted using real-world roadside scene data. The experimental findings indicate that the proposed algorithm performs better than other algorithms in terms of identification F-score， ID switch， fragmentation， mostly tracked， mostly lost， and multiple object tracking accuracy.

Abstract:

Based on the computational fluid dynamics method （CFD）， the wind field cases with different building environment parameters were systematically simulated and the variation of mean wind speed profile exponents was discussed. The results show that the average building height has a greater impact on the mean wind speed profile exponent than the building density. An empirical formula for the mean wind speed profile exponent with the average building height and building density is proposed as independent variables.

Abstract:

Using finite element modelling， this paper investigates the buckling and post-buckling behavior of K6 single-layer spherical reticulated shells. Based on the proposed criterion to distinguish global buckling modes， the mechanism and load-carrying capacity reduction of global two-modal interaction is revealed from the perspective of deformation. The results show that the stability of K6 single-layer spherical reticulated shells is vulnerable to global interactive buckling， which may lead to significant load-carrying capacity reduction and unstable post-buckling behavior because of the deformation transition from first global eigenmode dominance at the buckling stage to global interactive mode dominance and second global eigenmode dominance at the initial and final post-buckling stages respectively. Imperfection amplitudes can affect the dominance of global interactive buckling for the ultimate behavior.

Abstract:

Capillary water absorption tests on non-carbonated and carbonated concrete specimens with different water-cement ratios were conducted to investigate the effects of carbonation on capillary water absorption performance of concrete. Mercury intrusion porosimetry （MIP）， backscattered electron microscopy （BSE） and thermogravimetric analysis （TGA） were used to identify the pore structure of concrete before and after carbonation. The results show that the water absorption performance of carbonated concrete greatly decreases， and absorption capacity declines by 20.0% to 26.5%， and the water absorption coefficient falls by 30.8% to 37.8%. Compared with non-carbonated concrete， the most probable pore diameter of the carbonated concrete decreases by 13.9 nm to 15.1 nm， with the highest reduction rate reaching 41.8%. The reduction is more significant for concrete with higher water-cement ratios. The carbonation process is found to result in the transformation of calcium hydroxide into calcium carbonate and a decrease in pore diameter. The calculation results indicate that the model proposed by Lucas-Washburn et al. is still effective for carbonated concrete water absorption， and the selection of the most probable pore diameter as the equivalent pore diameter has the best model performance.

Abstract:

By adopting coupled heat and moisture tests in cement-based materials， this paper investigated the effects of temperature and humidity gradients on moisture transport and temperature and humidity distributions. The results reveal that temperature and humidity gradients accelerate moisture transport， and the effect of temperature gradient on moisture transport is more significant than humidity gradient. Moreover， the temperature gradient has a greater effect on temperature distributions than the humidity gradient， and the temperature and humidity gradients have a significant effect on relative humidity distributions. It is， therefore， concluded that during coupled heat and moisture transport in cement-based materials， the moisture transport of cement paste is faster than that of mortar， and the effect of heat transfer should be considered on moisture transport， while the effect of moisture transport could be neglected on heat transfer. The calculated results show that the coupled heat and moisture transport model has a great applicability to cement-based materials

Abstract:

In this paper， a laboratory test device for multi-physical field sensing during soil freezing and thawing was established. Traditional point sensors and distributed optical fibers were used in the test to monitor the temperature field， moisture field， and deformation of the subgrade during freezing and thawing process. Distributed sensing characteristics of the temperature and moisture fields of the soil under single-ended freezing and double-ended thawing condition were investigated， and the monitoring performance of the active heating distributed optical fiber technology was verified. The results show that， under the condition of -15 ℃ single-ended freezing， the soil freezing process is divided into three stages according to the cooling rate， i.e.， the rapid cooling stage， the gradual cooling stage， and the stable equilibrium stage. The freezing front gradually moves down during freezing， and the maximum freezing depth is about 35% of the height of soil column. The pore water is affected by the frost heave migration force to the freezing front and causes the redistribution of soil moisture field before and after freezing and thawing， thus causes frost heave deformation of the soil， which accounts for 7.8 % to 10.9 % of the maximum freezing depth of soil column. The distributed optical fiber can accurately obtain the characteristics of the soil temperature and moisture fields during the freezing and thawing process， and identify the range of the freezing area in the soil. The goodness of fit R² of temperature and moisture monitoring reaches 0.98 and 0.94， respectively.

Abstract:

This study applied the technology acceptance model and information system success model to investigate drivers’ willingness to use the eco-driving guidance assistance system and their preference for information types， using a 7-likert questionnaire. The study shows drivers are inclined to accept the audio warning type of eco-driving guidance. On urban roads， improving the perceived usefulness and perceived ease of use of the guidance assistance system has a beneficial effect on drivers’ acceptance of the system. In contrast， on urban expressways， perceived usefulness and perceived ease of use affect driver willingness to use the system significantly only during peak hours. However， the specific factors affecting drivers’ willingness to use the system in off-peak hours still need to be explored in future experiments.

Abstract:

Low-cycle fatigue tests were conducted on S321 and S321H stainless steels at 0.5%， 0.7%， and 0.9% strain amplitude at room temperature and 650 °C. Fatigue loading history tests were performed on S321 and S321H at room temperature for several life points， and the amount of deformation-induced martensite was measured by quantitative XRD analysis. The results show that deformation-induced martensite generation leads to continuous cyclic hardening in austenitic steels， and the alloys continue to secondary harden after the primary hardening at room temperature， whereas there is no deformation-induced martensite generation at 650 °C， and the alloys enter into cyclic stabilization phase after the primary hardening. The lower the carbon content during cycling， the more deformation-induced martensite there is， and the higher the degree of cyclic hardening of the alloy. At two temperatures， the fatigue life of the two materials in the low strain amplitude are not much different， the greater the strain amplitude， the greater the difference between the fatigue life of the two materials. Therefore， it can be considered to use S321 stainless steel instead of S321H stainless steel. Therefore， it can be considered to use S321 stainless steel instead of S321H stainless steel for service conditions of low strain amplitude.

Abstract:

In order to investigate the intergranular fracture characteristics of HR3C austenitic heat-resistant steel at various creed rates， several techniques are employed to test the properties and observe the microstructure evolution. The quantitative electron backscatter diffraction （EBSD） analysis results show that creep strain has little effect on the grain size， and grains do not exhibit a preferred orientation or no obvious texture after crept at different creep rates. The lowest proportion of low Σ CSL grain boundaries at 150 MPa implies that the primary twin structures is preserved at a low creep rate， while some twin structures are evolved into general grain boundaries at a high creep rate. The EBSD analysis methods well describe the creep micro-strain distribution， which shows that the creep strain is inhomogeneous between grains and strain occurs mostly close to grain boundaries and induces grain boundary migration. The generation and propagation of intergranular microcracks and creep voids are the most significant damage characteristics. These damage phenomena are closely related to the creep rates and the character of the grain boundaries.

Abstract:

Cathode recirculation is particularly suitable for solving the problems of water shortage and high potential of vehicle fuel cell systems under low load or idle conditions by adjusting the oxygen partial pressure and relative humidity of the stack intake. In this paper， a control-oriented nonlinear model of fuel cell system with cathode recirculation function is established. The exhaust gas circulation pump is used to introduce the cathode exhaust gas back to the outlet of the air compressor to realize intake air humidification and oxygen partial pressure regulation. The model can capture a series of dynamic responses brought about by cathode recirculation. The results show that the cathode recirculation can greatly improve the intake air humidity at a low current density. With the increase of the exhaust gas circulation ratio， the humidification dynamic response is faster. The dynamic response of the oxygen partial pressure regulation is slow， but the high potential at small electric densities can still be limited to less than 0.8 V by a reasonable cathode recirculation strategy. A 50 % idle power reduction is achieved by suppressing the voltage at small densities.

Abstract:

Particulate matter （PM） was sampled from three positions of a certain Ⅵ gasoline direct injection （GDI） engine before three-way catalyst （TWC）， after TWC， and after gasoline particle filter （GPF）. Particle morphology was investigated and influences of working conditions， TWC， GPF on particle number （PN）， particle size distribution （PSD）， morphology of particles emitted from engine were analyzed. The results show that the PSD of PN emitted from the engine is unimodal， the PN of particles with a diameter less than 23 nm is higher under lower load and engine speed operating condition. With the increase of engine speed and load， the peak particle diameter increases. The PM of the engine is formed by the accumulation of basic carbon particles with a “core-shell” structure and presents chain， branch， cluster， or other structures. With the increase of load， the particle size increases slightly， and both the overlap of basic carbon particles and the fractal dimension increase. While with the increase of engine speed， all these three parameters of size， overlap， and fractal dimension of PM decrease. With the transport processing of exhaust gas， PN is reduced. TWC does not affect PSD， and the purification efficiency of PN is 41.6%~94.2%. It has a good purification effect on small particles with a diameter less than 23 nm， and the purification efficiency at a lower engine speed and load is higher. The purification efficiency of GPF is about 80%. For particles with a diameter of 23~100 nm， the purification efficiency of GPF is high. However， it has little effect on the purification of particles with a diameter less than 10 nm. TWC and GPF do not change the structure of particles， after TWC and GPF， the overlap of basic carbon particles and the fractal dimension decrease.

Abstract:

This paper aims at investigating the asymmetric fluctuation coupling effect of cross-border capital flows and RMB exchange rate through two dimensions of industrial capital flows and financial capital flows. The result shows that both capital flows （industrial and financial capital flows） and the exchange rate have a significant continuous and clustering fluctuation. Besides， there are differential asymmetric fluctuation coupling effects between the two types of capital flows and the exchange rate. This paper proposes policy recommendations such as giving priority to introducing policy arrangements for coping with industrial capital outflows risks， improving the framework for administration featuring “macro-prudential management plus micro-supervision” and reducing the overshooting risk of foreign exchange market， and building a buffer zone for capital flows by improving the RMB offshore transactions.

Abstract:

Taking the “education” process funded by college as the research object， this paper tries to improve the traditional failure mode and effects analysis （FMEA）. It proposes a fuzzy FMEA model to analyze the failure modes and key risks of the “education” process by combining triangular fuzzy numbers， weights of risk indices， object weights of experts， the Delphi method， and interviews. It is found that the model proposed objectively reflects the importance of risk indices and difference among experts’ discourse power， which means that the model proposed has a higher accuracy and reliability. Suggestions for high-risk failure modes are proposed.