Abstract:By introducing the equivalent coefficient weighting into the 1st natural modal frequency response function （FRF） to represent the higher modal contribution， an improved simplified model for calculating the wind-induced response of super tall buildings constructed by tuned mass-damper-inerter （TMDI） is derived. The analytical solution for the coefficient is further refined， utilizing the maximum tolerable error criteria of both the H₂ and H_∞ norms of FRF. Employing a super tall building controlled by TMDI with varied parameters as a case study， the efficacy of TMDI in mitigating wind-induced structural responses is evaluated based on numerical model， generalized single-degree-of-freedom model and the improved model. Furthermore， the impact of different parameters on the accuracy of both the generalized single-degree-of-freedom model and the improved model is analyzed. Across a spectrum of TMDI cases， the computational efficiency of the three models is scrutinized. Results indicate that the improved model in this study offers superior accuracy in estimating the structure’s 1st modal response compared to the generalized single-degree-of-freedom model. Moreover， it demonstrates greater computational efficiency when compared to the numerical model across a diverse range of TMDI configurations.

Abstract:In order to study the seismic performance of the multi-story glulam frame-CLT core tube structure with buckling-restrained brace， a 1∶2 scaled model of a 5-story glulam frame-CLT core tube structure with steel-wood buckling-restrained braces was designed and manufactured， then a shaking table test study and finite element analysis simulation were carried out. In the test， the unidirectional seismic action was used to observe the damage of the structure under the earthquake intensity ranging from 7 degree frequently to 9 degree rarely， and the acceleration response， displacement response， inter-story drift angle， working performance of buckling- restrained braces were obtained at each seismic intensity. The results show that the damage of the model is mainly concentrated in the CLT core， some self-tapping screws are pulled out in the buckling-restrained brace， and there is no obvious damage to the glulam frame. The model exhibites good energy dissipation capacity and the damping ratio rises to 12.5%. The acceleration amplification factor of each story is between 1.0 and 5.0， and the acceleration amplification factor of the first story of the model is notably higher than that of other stories. The maximum inter-story displacement angle of the model does not exceed the specification requirements under the 9 degrees rarely， and the overall collapse does not occur. The overall finite element analysis results are in better agreement with the shaking table test results.

Abstract:The diffusion of chloride ions in concrete is a stochastic process. In this paper， the Fick’s second law with random diffusion coefficient is adopted to describe the process. The analytical solution of the probability density function （PDF） of the chloride ion concentration under given boundary conditions is presented by solving the generalized probability density evolution equation （GDEE） about the chloride ion diffusion process. Taking the chloride ion diffusion coefficient obeying a lognormal distribution as an example， the correctness of the analytical solution is confirmed by numerically solving GDEE. It is shown that： the chloride ion diffusion process is a non-stationary spatio-temporal random field； the probability density and variability of the chloride ion concentration distribution not only depend on the stochasticity of the diffusion coefficient， but also are related to the developmental stage of the diffusion process.

Abstract:In order to explore the effect of creep deformation characteristics of soft rock on the stress distribution and deformation law of surrounding rock， taking the carbonaceous slate as a research object， based on indoor experiments， theoretical methods and numerical analysis， the Nishihara model is improved to describe the whole creep characteristics of the rock. Then， the improved Nishihara model is expanded to a transversely isotropic creep model， and a three-dimensional equation is derived under the assumption of transverse isotropy. Finally， the creep experimental data are fitted and the creep parameters are inverted according to the three-dimensional equation. The results show that the creep curve of horizontal bedding rock samples changes with the load， and the graded creep deformation of horizontal bedding rock samples and vertical bedding rock samples shows obvious anisotropy. The improved Nishihara model can well adapt to the deformation characteristics of the creep curve of the carbonaceous slate. The parameter value change of the transversely isotropic creep model fully reflects the effect of stress level on the creep of the carbonaceous slate， thus verifying the adaptability and rationality of the improved Nishihara model.

Abstract:In this paper， a digital twin （DT） based just-in-time （JIT） construction safety control strategy for slope engineering is proposed. To further enhance the efficiency and precision， DT technology is integrated into the JIT safety control framework for slope construction， and a scene-entity-relationship-incident-control （SERIC） modeling method is introduced for cross-domain sharing and reuse. The proposed JIT construction safety control strategy for slope construction has been implemented in the Huangmeishan Tunnel reconstruction and expansion project， incorporating advanced methods for data acquisition， interpretation， and aggregation. The results demonstrate that the application of the JIT construction safety control strategy significantly improves the automation level， construction efficiency and accuracy of slope construction.

Abstract:For the deconstruction of components in aging structures， a mathematical model for process planning is developed based on resource flow constraints， flexible work breakdown， and selective dismantling strategies. A multi-agent reinforcement learning method is then proposed to optimize the deconstruction planning with recovery efficiency as the core objective. Case study results indicate that， compared to the genetic algorithm， the method achieves a higher carbon emission return-to-consumption ratio， demonstrating the effectiveness of intelligent algorithm-based optimization in promoting the energy efficiency and carbon reduction.

Abstract:Autonomous driving technology is a pivotal focus for the future development of the transportation and automotive sectors， with the speed decision serving as a critical component that directly impacts the driving safety and efficiency. As autonomous vehicles undergo testing and trial operations， new expectations for speed decision have emerged. Driving speed planning process should not only integrate seamlessly into mixed traffic environments， but also meet personalized needs and comfort requirements of drivers. In this paper， the current research landscape and key issues related to global speed decision are reviewed. Methodologies of local speed decision are analyzed， and the simulation evaluation methods employed in this field are summarized. Finally， future development plan for speed decision is mapped out，emphasizing the importance of improving the ability to coordinate heterogeneous traffic participants， balancing legal and social norms， integrating knowledge-driven and data-driven approaches， enhancing the unification of global and local speed， and evaluating algorithms on a larger scale.

Abstract:To effectively test the interaction capability of autonomous vehicles， a dynamic strategy-explainable heterogeneous multi-agent adversarial inverse reinforcement learning （MA-AIRL-2） model is established， focusing on interaction behaviors at unprotected left-turn intersections. Left-turning and straight-going vehicles are modeled as heterogeneous agents， and goal-directed gain rewards are introduced to improve the learning efficiency. Finally， the model is trained using the open-sourced dataset SinD， and evaluated on interaction strategies， vehicle trajectory， macroscopic traffic flow features dimensions， respectively. The results show that： the MA-AIRL-2 model can accurately portray the competitive and cooperative strategies during vehicles’ dynamic interaction and achieve the replication of strategy diversity； MA-AIRL-2 model can generate the closest realistic trajectory and traffic flow features when compared with the baseline models， and the trajectory accuracy is improved by 26% and 21% with respect to the homogeneous strategy and the no-gain reward model， respectively.

Abstract:To accurately grasp the key points of management and education on bus drivers， the influencing factors of risky driving behaviors among bus drivers were identified from the perspectives of demographic characteristics， physical and psychological health， and company safety education. A questionnaire survey on bus drivers’ behavioral and psychological health was conducted， with a total of 10 201 valid samples collected. Four types of machine learning models were compared， and the internal influence mechanism was explained based on the explainable machine learning framework. The results indicate that： CatBoost achieves a mean absolute error （MAE） of 0.190 6 and a root mean square error （RMSE） of 0.267 1， having the best overall performance. Psychological health is the most important factor. With severe degrees of the driving anger， anxiety， and job burnout， bus drivers are more likely to commit risky driving behavior. The impact of insomnia and depression on risky driving behavior shows the non-linear relationship. The moderate and lower degree of insomnia and depression are positively correlated with risky driving behaviors， while the severe insomnia and depression are negatively correlated with risky driving behaviors； although the importance of company safety education is relatively low， bus drivers who regularly participated in various types of company safety education are less prone to risky driving behaviors.

Abstract:By using kinematics and video demonstrations methods， drivers were trained to take on rational decision-making measures under conflict conditions. Data from 30 participants’ emergency avoidance driving simulator experiments under lateral conflict scenarios at intersections were collected. A generalized linear mixed model was used to analyze differences in avoidance reaction， learning effects， and decision characteristics before， after， and one month after training intervention. The results show that the training intervention significantly improves the accuracy of rational decision-making in both horizontal and vertical evasive reactions in short term. However， after one month of the training intervention， the positive impact of driver training on evasive operations significantly decreases.

Abstract:Based on the distinctive constraint of budget and penetration rate inherent in incentive-based travel demand management （IBTDM） scheme design， a bottleneck model was employed to analyze the impact of departure time offsets on travelers’ travel costs and the optimal IBTDM incentive scheme. Then， the incentive scheme considering departure time offset costs was compared with the scheme based on departure time， which does not consider travelers’ original travel patterns， aiming to assess the applicability and effectiveness of different incentive strategies. The results indicate that the relative magnitude of travelers’ unit departure time offset cost θ and unit travel time cost α significantly shapes the structure of optimal incentive schemes considering departure time offsets. When θ<α， the incentive amount decreases first and then increases over time； when θ>α， the incentive amount increases first and then decreases； and when θ=α， the incentive amount remains constant. Furthermore， the comparative effectiveness of the incentive scheme considering departure time offsets versus the incentive scheme based on departure time， in mitigating system congestion under budget constraints， also depends on the relative values of θ and α. Therefore， administrations must carefully evaluate travelers’ unit travel costs to determine the most efficient incentive schemes. Generally， θ<α more accurately reflects the real scenario， in which case the incentive scheme considering departure time offsets is more effective in alleviating system-wide congestion.

Abstract:In this paper， we use a combination of the ridership of rail stations and the proportion of rail trips to characterize the passenger attraction of rail stations. In order to investigate the factors influencing the passenger attraction of rail stations， the “door-to-station + station-to-door” and “station-to-station” rail trip quality measurement method is proposed， from the perspective of the trip chain. Taking the urban and suburban areas of Shanghai as an example， a multi-scale geographically weighted regression （MGWR） model of the factors influencing the passenger attraction of rail stations is built， with the ridership of rail stations and the proportion of rail trips as dependent variables， respectively. The regression results show that the ridership in urban stations is strongly influenced by the quality of “door-to-station + station-to-door” trips. And the ridership in suburban stations is mainly influenced by built environment and the quality of “station-to-station” trips， especially the ratio of travel time between car and rail. The proportion of rail trips in suburban people is primarily influenced by the functional mix of the site and the distance from the city center， and some remote sites are also heavily influenced by the rail trip quality.

Abstract:To improve the passing efficiency of Three Gorges-Gezhou Dams Hub， a co-scheduling problem of approach channel and ship lock group （CSP-AC&SLG） with synchronous moving process was studied. In consideration of the ship freeboard height and steering skill， a mathematical model of CSP-AC&SLG was formulated， aiming to maximize the water resource utilization， minimize the average waiting time of ships and minimize the average total energy consumption of ships. A multi-objective hybrid heuristic algorithm （MOHHA） was proposed to solve the problem. MOHHA makes use of dual chain coding scheme and staged group right-shift decoding strategy to respectively represent and decode each solution. A fuzzy correlation entropy analysis method was triggered to assess multi-objective solutions. An adaptive parameter adjustment strategy and variable neighborhood simulated annealing were then inserted to enhance the resolution ability of the algorithm. Simulation results demonstrate that MOHHA can deal with CSP-AC&SLG， realizing the dual goals of improving the navigation efficiency and saving the energy consumption of ships.

Abstract:The sound field performance for an automobile aero-acoustic wind tunnel comprises three aspects of free-sound field， background noise and low-frequency pressure fluctuation. The latest test results of Tongji Aero-acoustic Wind Tunnel has validated that the background noise OASPL at 160 km?h^-1 wind speed is 61 dB（A）， and the low-frequency pressure fluctuation coefficient is lower than 0.5% within the wind speed range from 50 to 200 km?h^-1. Meanwhile， it was found that the belt suction of rolling road systems adds the background noise below 80 km?h^-1 wind speeds， yet no effects on the low-frequency pressure fluctuation. According to ISO 3745∶2012， the cut-off frequency of the free-sound field is down to 50 Hz， while the radius of the free-sound field at and above 63 Hz cut-off frequencies reaches 9 m inside test section plenum.

Abstract:A numerical simulation method with ANSYS Fluent software and user-defined function （UDF） was employed to model and simulate the coupled two-dimensional heat transfer process involving internal/external fluids and the pipe wall. Taking theoretically calculated outlet temperatures and heat transfer coefficients as benchmarks， through the computational fluid dynamics （CFD） investigation， the heat transfer characteristics of supercritical hydrogen in different inlet states and surrounding air domains were explored under the condition of external air liquefaction. It is shown that within the liquid hydrogen temperature range dominated by the air liquefaction （50 to 70 K），the temperature of the external air has a negligible impact on the overall heat transfer effectiveness. Even when the air velocity reaches 8 m·s^-1， the enhancement in the overall heat transfer coefficient is only approximately 20%. Neither forcing convection to increase air velocity， altering the ambient temperature， nor modifying the inlet mass flow rate can significantly improve the overall heat transfer coefficient within the temperature range. Instead， the coefficient is predominantly governed by the heat transfer coefficient associated with air liquefaction， meaning that the temperature of the cryogenic fluid is the dominant influencing factor.

Abstract:Based on the housing transaction data of Beijing in 2020， the differences among three types of housings using logit model are analyzed， and the STAR model and 2STQR model are constructed to explore the social differentiation of housing prices in the property rights sub-markets and property rights-price level composite sub-markets. It is found that there are differences in structure features， neighborhood features， and location features among the three types of housings； compared with public housing buyers， commercial housing buyers pay more attention to structure and location features， while affordable housing buyers show stronger willingness to pay for neighborhood features； and the differences in willingness to pay change in a complex way with the increase in the price of the housings. The findings of the study reveal the differences in housing preferences among buyers from sub-markets with different property rights， and provide an important reference for the development and location of public housing and affordable housing projects.