Search Results (8,150)

Accurate environmental perception is fundamental to safe autonomous driving; however, most existing multimodal systems rely on fixed or heuristic sensor fusion strategies that cannot adapt to scene-dependent variations in sensor reliability. This paper proposes Cross-Modal Adaptive Attention (CMAA), a unified end-to-end Bird’s-Eye-View (BEV) perception framework that dynamically fuses camera, LiDAR, and RADAR information through learnable, context-aware modality gating. Unlike static fusion approaches, CMAA adaptively reweights sensor contributions based on global scene descriptors, enabling the robust integration of semantic, geometric, and motion cues without manual tuning. The proposed architecture jointly performs 3D object detection, multi-object tracking, and motion forecasting within a shared BEV representation, preserving spatial alignment across tasks and supporting efficient real-time deployment. Experiments conducted on the official nuScenes validation split demonstrate that CMAA achieves 0.528 mAP and 0.691 NDS, outperforming fixed-weight fusion baselines while maintaining a compact model size and efficient inference. Additional tracking evaluation using the official nuScenes tracking devkit reports improved tracking performance, while motion forecasting experiments show reduced trajectory displacement errors (minADE and minFDE). Ablation studies further confirm the complementary contributions of adaptive modality gating and bidirectional cross-modal refinement, and a stratified dynamic analysis reveals consistent reductions in velocity estimation error across object classes, motion regimes, and environmental conditions. These results demonstrate that adaptive multimodal fusion improves robustness, motion reasoning, and perception reliability in complex traffic environments while remaining computationally efficient for deployment in safety-critical autonomous driving systems. Full article

This paper investigates the influence of rolling resistance on fuel consumption in Class 8 heavy-duty vehicles, with a focus on a modeling approach through variations in the rolling resistance coefficient ($C_{rr}$) across different driving scenarios. Leveraging TruckSim’s multibody modeling approach for vehicle dynamics and MATLAB/Simulink co-simulation capability, the study provides insights into how tire rolling resistance affects energy efficiency under varying conditions while enabling controlled, repeatable comparisons across various scenarios. Results show that across the evaluated scenarios, increases in $C_{rr}$ impact the vehicle’s speed, fuel consumption, engine torque, and crankshaft spin. Specifically, increasing $C_{rr}$ from 0.004 to 0.013 was found to lead up to 68% higher fuel consumption in high demand scenarios. These findings aim to guide efforts to optimize tire design and vehicle performance that help achieve improved fuel efficiency. Full article

Research on centennial-scale precipitation variability within the Australian summer monsoon (AUSM) remains limited, particularly regarding its driving mechanisms and the sustainability-relevant implications for long-term water security and climate adaptation. Here, we use the TraCE-21ka transient simulation, which credibly reproduces the centennial periodicities documented in Holocene proxy records, to attribute the physical drivers of AUSM centennial variability. Attribution is conducted by contrasting the all-forcing (AF) simulation with four single-forcing experiments that isolate the effects of orbital parameters, ice sheets, meltwater flux, and greenhouse gases. Among these experiments, the meltwater-forcing run best reproduces the centennial periodicities found in the AF simulation, indicating that meltwater input is the leading contributor to Holocene AUSM centennial variability. We further identify a dynamical pathway in which Atlantic Meridional Overturning Circulation (AMOC) variability acts as the key mediator linking meltwater perturbations to Australian hydroclimate. The enhanced AMOC amplitude during the meltwater interval (0.14 at 9–8 ka BP), compared with much weaker fluctuations during the non-meltwater interval (0.01 at 4–3 ka BP), is accompanied by a ~200-year periodicity in AUSM precipitation. This periodicity arises through an interhemispheric teleconnection: a strengthened AMOC cools Southern Hemisphere sea surface temperatures, reduces moisture availability for northern Australia, and promotes large-scale subsidence that suppresses monsoon rainfall. By contrast, during 4–3 ka BP, when meltwater forcing was negligible, weaker AMOC variability coincides with warmer Southern Hemisphere sea surface temperature (SST), favoring cyclonic circulation over northwestern Australia, enhanced moisture convergence, and stronger ascent, ultimately intensifying AUSM precipitation. Beyond advancing process understanding, these results provide a sustainability-oriented framework for interpreting low-frequency hydroclimate variability relevant to Australia’s water resources and climate adaptation. Specifically, the identified meltwater–AMOC–SST–AUSM pathway offers a physical basis for developing and evaluating long-horizon indicators of monsoon-driven rainfall variability, informing monitoring strategies and scenario planning for drought–flood risk management, water allocation, and climate-resilient infrastructure. By linking centennial-scale monsoon variability to an identifiable remote driver, this study contributes to quantifying and contextualizing natural hydroclimate variability that can confound near-term trends, thereby supporting more robust sustainability assessments, adaptation policy design, and integrated water-resource management under ongoing climate change. Full article

Ensuring the safety of Autonomous Driving Systems (ADS) at urban intersections remains challenging due to complex interactions between vehicles and traffic management infrastructure. This study validates an ADS equipped with connected perception using Infrastructure-to-Vehicle (I2V) communication within a combined virtual and hybrid testing approach. The validation follows the overall structure and methodology of the SUNRISE Safety Assurance Framework (SAF), which is applied in detail where required by the scope of the study. Five representative urban intersection scenarios, covering both nominal driving conditions and safety-critical edge cases, are evaluated using virtual simulations in MATLAB/Simulink (2014b) and hybrid experiments integrating OMNeT++ (5.7.1)/Veins (5.2)/SUMO (1.12.0) with real-world components. Key Performance Indicators (KPIs) related to safety, decision-making, longitudinal control, passenger comfort, and V2X communication performance are analyzed. The results show strong consistency between virtual and hybrid testing, with ego vehicle speed deviations below 2 km/h and trigger distance differences under 3 m. V2X communication achieves a near-perfect Cooperative Awareness Message (CAM) delivery ratio, with an average latency of approximately 142 ms. While this latency remains within the tolerance of the deployed ADS, the overall end-to-end delay highlights opportunities for further optimization. The study demonstrates how the SUNRISE SAF can effectively structure ADS validation, identifies critical scenarios such as right-of-way violations by non-priority obstacles, and provides insights into improving connectivity handling and low-speed braking behavior for Cooperative, Connected, and Automated Mobility (CCAM) systems in urban environments. Full article

This paper proposes a reinforcement learning (RL)-based control method for robust path tracking of a 4-wheel independent steering (4WIS) mobile robot in outdoor rough terrain environments. Traditional wheeled robots typically suffer from limitations including mobility constraints in narrow spaces, path deviations caused by ground slip, and reduced traction on rough terrain. To address these challenges, we designed a 4WIS mobile robot and implemented an architecture that independently controls the steering and driving of each wheel. The RL state space is defined by look-ahead path information, robot pose, velocity, and tracking errors, while the action space consists of target angular velocity and steering angle. To ensure robust performance, we applied random path and terrain generation and implemented domain randomization for sensors and actuators based on empirical GPS and motor data. The proposed controller was validated against the Pure Pursuit algorithm through dynamic simulations and real-world experiments. In simulations mimicking outdoor terrain, the controller reduced lateral and heading RMSE by 6.32% and 16.00%, respectively. In actual outdoor environments, it reduced these errors by 21.54% and 4.78%, respectively. These results demonstrate that the proposed controller provides superior robust tracking performance in unstructured outdoor environments. Full article

To investigate the regulatory effects of emotional states on the evolution of fatigue in metro drivers, this study conducts an experimental investigation based on an urban rail transit driving simulation platform. A total of 21 participants complete a 90 min simulated driving task, during which electroencephalogram (EEG) and electrocardiogram (ECG) signals are synchronously collected from drivers for fatigue assessment and emotion recognition, respectively. An emotion recognition model based on a multi-scale convolutional neural network (MSCNN) combined with an attention mechanism is constructed. The proposed model uses ECG signals to classify three emotional states—neutral, positive, and negative—where the neutral state is defined as an emotionally undefined baseline that is neither positive nor negative. The model achieves a classification accuracy of 86.96% on the DREAMER dataset. By temporally aligning the emotion recognition results with EEG frequency-domain fatigue indicators, the results show that fatigue exhibits the highest growth and largest fluctuation in amplitude under negative emotions, demonstrating a pronounced fatigue-accelerating effect. Under positive emotions, fatigue decreases considerably and has smaller fluctuations, indicating a certain buffering and restorative effect. In contrast, the neutral emotional state exhibits intermediate and transitional fatigue characteristics. This study innovatively integrates ECG-based emotion recognition with EEG-based fatigue assessment to reveal the mechanisms based on which emotions influence fatigue in metro driving tasks from a physiological perspective. This work provides a basis for emotion-aware fatigue monitoring and safety intervention strategies. Full article

Large language models (LLMs) exhibit strong semantic reasoning capabilities for autonomous driving decision-making; however, their substantial inference latency poses a critical challenge for real-time closed-loop vehicle control. This study proposes an engineering-oriented framework to enable latency-constrained LLM-based decision-making by integrating bird’s-eye-view (BEV) structured perception with low-bit quantized inference. The BEV perception module compresses multi-view visual inputs into structured semantic representations, thereby reducing input redundancy and enhancing inference efficiency. In addition, 4-bit post-training quantization (PTQ), combined with an optimized inference engine, is employed to alleviate computational and memory bandwidth constraints during autoregressive decoding. Experiments conducted on the CARLA simulation platform under car-following, overtaking, and mixed driving scenarios—validated through 500 independent trials—demonstrate that the proposed framework substantially reduces end-to-end inference latency while maintaining stable decision-making performance. The results indicate that the system satisfies the 10 Hz real-time control requirement and significantly improves control quality, as evidenced by reduced collision rates and lower Average Jerk compared with both traditional imitation learning (Behavioral Cloning, BC) and the Transformer-based TransFuser baseline. Furthermore, sensitivity analyses confirm the robustness of the framework under environmental degradation and perception noise, underscoring the practical feasibility of deploying LLMs for safe and reliable closed-loop autonomous driving. Full article

The advent of 5G/6G technologies and the pervasive deployment of IoT devices are driving the emergence of demanding applications that necessitate ultra-low latency, high bandwidth, and significant computational power. Traditional cloud computing models fall short in meeting these stringent requirements. To address this, Software-Defined Edge Networks (SDENs) have emerged as a promising architecture, yet efficiently managing their heterogeneous and geographically distributed resources poses substantial challenges for optimal application provisioning. In response, this paper proposes a novel framework for intelligent task offloading, which reframes the intricate multi-component application task offloading problem as a Virtual Network Embedding (VNE) challenge within a SDEN environment. We introduce a comprehensive model where complex applications are represented as Virtual Network Requests (VNRs). In this model, each VNR consists of virtual nodes that demand specific computing and storage resources, as well as virtual links that demand specific bandwidth and must adhere to maximum tolerable delay constraints. To dynamically solve this NP-hard VNE problem in the face of stochastic VNR arrivals and dynamic network conditions, we leverage Deep Reinforcement Learning (DRL). Specifically, a Soft Actor-Critic (SAC) agent is employed at the SDN controller. This agent learns a sequential decision-making policy for mapping virtual nodes to physical edge servers and virtual links to network paths. To guide the agent towards efficient resource utilization, we define the reward for each successful embedding as the long-term revenue-to-cost ratio. By learning to maximize this reward, the agent is naturally driven to find economically viable allocation strategies. Comprehensive simulation experiments demonstrate that our SAC-based VNE approach significantly outperforms other baselines across key metrics, affirming its efficacy in dynamic SDEN environments. Full article

Aiming at the challenge of simultaneously controlling ride comfort and wheel grounding performance for mining dump trucks, this paper proposes a multi-dimensional synergistic optimization control (MDSOC) strategy based on model predictive control (MPC) for active hydro-pneumatic suspension. First, an accurate hydro-pneumatic suspension and hinged mining truck full-vehicle-dynamics model is established, and the model accuracy is validated through actual vehicle testing. Subsequently, an MDSOC-MPC for active hydro-pneumatic suspension is constructed to minimize the mean square root of the three-axis acceleration of the body, pitch angle, roll angle, and wheel dynamic tire load. Comparative analysis is performed with traditional single-MPC longitudinal, lateral, and vertical control, and the simulation results showed: under emergency braking conditions, the root mean square (RMS) value of the pitch angle is reduced by 18.2%; under single and double-shift conditions, the RMS values of the roll angle are reduced by 40.4% and 30%, respectively; under D-class random road, the RMS values of the longitudinal, lateral, and vertical body acceleration are significantly reduced by 22%, 21.5%, and 21.2%, respectively, while the RMS values of pitch angle and roll angle are reduced by 22.5%, and 20.2%, respectively, systematically improving riding comfort, vehicle wheel contact, and driving safety. This study provides a theoretical basis and feasible engineering methods for the active control of hydro-pneumatic suspension systems in heavy engineering vehicles. Full article

Theoretical studies of transonic accretion onto black holes reveal a wide range of possible solutions, broadly classified into smooth flows and flows featuring shocks. Accretion solutions that involve the formation of shocks are particularly intriguing, as they are expected to naturally produce observable variability features. However, despite their theoretical significance, time-dependent studies exploring the stability and evolution of such shocked solutions remain relatively scarce. To address this gap, we perform simulations of transonic accretion flows around a black hole in an ideal magnetohydrodynamic framework. Our simulations are initialized using boundary conditions derived from semi-analytical hydrodynamical models, allowing us to explore the stability of these flows under varying magnetic field strengths. Our results indicate that mildly magnetized flows in a uniform vertical magnetic field alter the accretion dynamics through magnetic pressure, with the resulting force imbalance driving oscillations in the shock front. Variations in the emitted luminosity arising from shock oscillations appear as quasi-periodic oscillations (QPOs), a characteristic feature commonly observed in accreting black holes. We find that the QPO frequency is determined by the radial position of the shock front: oscillations occurring closer to the black hole produce frequencies of tens of hertz, whereas shocks located farther out yield sub-hertz frequencies. Full article

The formation of longitudinal queues is critical for biological and artificial swarm systems to achieve efficient long-distance navigation. However, the “black-box” nature of conventional deep reinforcement learning models often obscures the microscopic rules driving the emergence of such ordered behaviors. To address this challenge, this paper proposes an interpretable computational model of collective behavior based on Projective Simulation and Episodic Compositional Memory, which enables individuals to learn decision-making strategies within a transparent state–action network. Simulation results demonstrate that the swarm can self-organize into stable and highly elongated longitudinal queues. Crucially, through visualization of microscopic strategies, we reveal a deterministic target-priority mechanism: when local neighbor alignment conflicts with global target orientation, individuals learn to strictly prioritize the target direction, serving as the key driving force for queue formation. Further parametric analysis indicates that the action space granularity exerts a nonlinear impact on stability, identifying moderate control precision as the optimal choice. This study not only provides a transparent computational explanation for the self-organization mechanism of queues in collective motion but also offers a theoretical foundation for designing interpretable swarm navigation systems. Full article

Compared with rule-based and optimization energy management strategies, online optimal energy management control strategies for a dual-mode power-split hybrid electric vehicles (PSHEVs) are able to achieve better fuel economy and real-time performance. Global online optimization of a finite time domain energy management strategy based on the hybrid model predictive control (HMPC) algorithm is proposed in this study. To reduce the computing time, a linearized predictive model is built; because dual-mode PSHEVs can be considered hybrid systems that include continuous and discrete states, the hybrid states can be expressed uniformly. Therefore, a mixed logical dynamic (MLD) predictive model is built based on hybrid system theory, and an HMPC energy management strategy is proposed based on the MLD predictive model. To solve the optimal control problem online to obtain the optimal control sequence, the optimal control problem is converted into a mixed-integer linear programming (MILP) problem. The HMPC-based energy management strategy is compared with dynamic programming (DP)-based and rule-based energy management strategies over two different driving cycles. Simulation results indicate that the HMPC-based EMS achieves 80.60% and 83.79% of the fuel economy performance obtained by the DP-based EMS. In comparison, the rule-based EMS only achieves 66.46% and 70.51% of the DP-based control performance. Therefore, the HMPC-based energy management strategy is favorable for real-time control while effectively improving fuel economy. Full article

Global, industrialization-driven environmental bottlenecks push manufacturing enterprises toward green transitions; yet, the information asymmetry between central and local governments, and between enterprises and banks, hinders this process. Adopting a systemic–synergistic perspective integrating decentralized governance and green credit, in this study, we investigate the green transition decisions of manufacturing enterprises. We construct a quadrilateral evolutionary game model involving the central government, local governments, enterprises, and banks, employing MATLAB R2022b to simulate the effects of the key parameters. Subject to the model’s structural assumptions and parameter boundaries, three core findings emerge: first, we find that punitive environmental policies outperform incentive-based instruments in driving enterprise emission reduction; second, we find that the adaptive adjustments made by decentralized governance can effectively facilitate green practices among enterprises; third, within this framework, we find that green credit exerts a non-monotonic impact on enterprises’ green transition behaviors; meanwhile banks’ assessments of enterprises’ environmental risks can indirectly promote enterprise abatement by motivating local governments through signal transmission. This study underscores the systemic synergy of decentralized governance and green credit, offering insights for multistakeholder coordination and policy optimization to advance organizational sustainability transitions for the green economy. Full article

Short auditory cues in enclosed built environments (such as elevator calls, access control, navigation, and heating, ventilation, and air conditioning (HVAC) notifications) influence not only usability but also stress and perceptions of well-being in daily indoor life. However, acoustic research remains largely focused on physical properties, and the psychophysiological impact of such short auditory cues remains under-quantified. To address this gap, a neuroscience-based evaluation approach, the Acoustic User Experience and Emotion (AUEX) model, is proposed. This model integrates functional near-infrared spectroscopy (fNIRS), electrodermal activity (EDA), and the User Experience Questionnaire (UEQ). With 33 in-cabin prompt sounds as a controlled typology of short auditory cues in an enclosed setting, we set up a simulated interaction experiment with 20 participants in a driving simulator vehicle cabin to investigate the relationship between acoustic properties and cognitive load, arousal, and user experience. The results show that timbre is the key factor, which was correlated positively with overall UX (r = 0.414) and negatively with prefrontal ΔHbO (CH3: r = −0.368; l-DLPFC: r = −0.449), indicating a decrease in cognitive load and a relaxed affective state. Conversely, high-frequency signals improved pragmatic quality but increased physiological arousal, which negatively affected hedonic assessment. To facilitate the translation of evaluation results into practice, we also completed a design phase that converted the AUEX results into scenario-based parameter targets and prototype designs for functional, warning, and brand/affective cues, illustrating how evidence-based relationships can be translated into design-ready outputs for enclosed built environments. These results confirm the AUEX approach as a transferable method for designing short auditory cues for well-being and provide parameter-level implications for therapeutic and human-centered sound design in smart buildings, intelligent vehicles, and other enclosed built environments. Overall, the AUEX approach provides a transferable evaluation-to-design workflow for short auditory cues in enclosed interactive contexts; however, direct generalization from a single controlled vehicle cabin setting to real-world building environments should be validated through future field studies. Accordingly, the present findings are positioned as evidence from a controlled enclosed case rather than universal conclusions for all enclosed spaces. Full article

Autonomous vehicles (AVs) must make predictable and socially compliant behavioral decisions to ensure safe and efficient interactions with other road users. To address this challenge, this paper proposes a game-theoretic behavioral decision-making model integrated with spatial motion planning to capture the interactive intentions between the ego vehicle (EV) and target vehicle (TV) in pairwise scenarios. First, the study defines an intention representation method that characterizes intentions using spatial area boundaries, feasible speed ranges, and a set of goal points (speed goal points, position-orientation goal points). Second, a spatial motion planning approach is adopted to evaluate the intention, which optimizes the driving scheme using a multi-objective cost function (incorporating pursuit precision, comfort, energy efficiency, and travel efficiency). Finally, the game-theoretic decision-making model is constructed. The Social Value Orientation (SVO) is introduced to quantify drivers’ social preferences, and the payoff function, which integrates safety rewards (based on inter-vehicle distance) and performance rewards (based on motion planning indices), is established. Simulation results verify that the proposed model can effectively address the interactive intention decision-making problem between the AV and other road users and handle different scenarios. Full article