Hasil untuk "Motor vehicles. Aeronautics. Astronautics"

J. Anderson

Xiao Wang, Mengyu Wang, Xueqian Bai et al.

To advance research in multi-agent reinforcement learning (MARL) for pursuit–evasion scenarios, this paper introduces a novel algorithm called Expert Knowledge and Opponent Modeling Multi-UAV Deep Deterministic Policy Gradient (EO-MADDPG). EO-MADDPG consists of two key components: the integration of expert knowledge and real-time sampled data and the prediction of evader UAV actions. The expert knowledge includes a multi-UAV formation control algorithm and an encirclement strategy, which incorporates consensus algorithms and Apollonius circle guidance. Additionally, the network-training framework is optimized by integrating information about opponent actions under a fixed policy for improved prediction accuracy. The experiments focus on three vs. one and three vs. two scenarios, where pursuer UAVs utilize EO-MADDPG and evader UAVs follow fixed policies with Gaussian perturbations. Experimental results show that EO-MADDPG achieves success rates of 99.9 ± 0.3% and 97.5 ± 1.4% (mean ± std over five seeds) in three vs. one and three vs. two pursuit–evasion simulations, respectively, outperforming the baseline MADDPG (72.7 ± 6.0% and 64.4 ± 34.4%). Ablation studies and cooperative landmark tasks further demonstrate improved training stability and interpretability.

Kartik Loya, Phanindra Tallapragada

This work presents a hybrid physics-informed and data-driven modeling framework for predictive control of autonomous off-road vehicles operating on deformable terrain. Traditional high-fidelity terramechanics models are often too computationally demanding to be directly used in control design. Modern Koopman operator methods can be used to represent the complex terramechanics and vehicle dynamics in a linear form. We develop a framework whereby a Koopman linear system can be constructed using data from simulations of a vehicle moving on deformable terrain. For vehicle simulations, the deformable-terrain terramechanics are modeled using Bekker-Wong theory, and the vehicle is represented as a simplified five-degree-of-freedom (5-DOF) system. The Koopman operators are identified from large simulation datasets for sandy loam and clay using a recursive subspace identification method, where Grassmannian distance is used to prioritize informative data segments during training. The advantage of this approach is that the Koopman operator learned from simulations can be updated with data from the physical system in a seamless manner, making this a hybrid physics-informed and data-driven approach. Prediction results demonstrate stable short-horizon accuracy and robustness under mild terrain-height variations. When embedded in a constrained MPC, the learned predictor enables stable closed-loop tracking of aggressive maneuvers while satisfying steering and torque limits.

Sandeep Gupta, Roberto Passerone

This article investigates the robustness of vision systems in Connected and Autonomous Vehicles (CAVs), which is critical for developing Level-5 autonomous driving capabilities. Safe and reliable CAV navigation undeniably depends on robust vision systems that enable accurate detection of objects, lane markings, and traffic signage. We analyze the key sensors and vision components essential for CAV navigation to derive a reference architecture for CAV vision system (CAVVS). This reference architecture provides a basis for identifying potential attack surfaces of CAVVS. Subsequently, we elaborate on identified attack vectors targeting each attack surface, rigorously evaluating their implications for confidentiality, integrity, and availability (CIA). Our study provides a comprehensive understanding of attack vector dynamics in vision systems, which is crucial for formulating robust security measures that can uphold the principles of the CIA triad.

CHEN Yao, XIONG Zhenghui, LUO Junwei et al.

Traditional ultrasonic automated non-destructive testing is a great challenge in the inspection of aviation components with complex surfaces. Complex surfaces can interfere with the formation of the focus in the sound beam，and the waveform transformation generated when the sound beam is incident is more complex. All these will lead to a decrease in the ultrasonic testing capability and a significant reduction in the obtained echo signal-to-noise ratio. Under the background of intelligent manufacturing，the development of rapid and low-cost manufacturing of aviation components has been seriously restricted. The paper analyzes the ultrasonic propagation of complex surface media，and summarizes the technical difficulties of automatic ultrasonic detection of complex surface components. The paper also describes the development status of three kinds of automatic ultrasonic imaging detection of complex surfaces，which are based on flexible phased array ultrasonic probe，ultrasonic C-scan imaging detection based on industrial robot and phased array ultrasonic imaging detection for complex surfaces. The advantages and limitations of three kinds of automated ultrasonic imaging detection are analyzed，and the challenges faced by various ultrasonic imaging technologies are reviewed. The key technology to break through the automatic ultrasonic imaging detection of complex aerospace components under the background of intelligent manufacturing is proposed. The paper introduces the future technical requirements for the development of advanced imaging algorithms for automated inspection and the intelligent recognition and classification of defects in the ultrasonic testing of complex aviation components. The key detection technologies based on digital twin detection path planning and the design and manufacture of massive channel phased array ultrasonic sensors，which are urgently needed to be broken through under the background of intelligent manufacturing，have been proposed.

Bohan Sun, Xiaoliang Wang, Zhefeng Yu

This study explores the impact of flap gaps on aircraft cabin vibrations, focusing on the dynamic responses of flap structures under aerodynamic excitations. Flaps are crucial for aircraft control, especially during takeoff and landing. Common issues in flap systems, such as wear on slide tracks, bearing clearances, and bushing slippage, introduce gaps that amplify vibration amplitudes. A time–domain method, employing a spring–beam model, was used to convert aerodynamic loads into structural loads, providing detailed insights into flap dynamics and their influence on the aircraft. Computational fluid dynamics software was utilized to calculate aerodynamic force distributions, and the effects of varying gap sizes on constraint forces and vibration responses were analyzed. The results indicate that increasing gap width reduces the maximum constraint force at the flap, altering the vibration characteristics. The study also reveals that higher vibration responses occur in the fore section of the fuselage due to modal superposition of frequencies near those of the flap support forces. Enhancing the stiffness of gap elements can mitigate abnormal vibrations, thereby improving aircraft performance and passenger comfort. This paper offers a quantitative analysis of the effects of flap gaps, serving as a technical reference for diagnosing faults based on cabin vibrations. These findings are critical for advancing aircraft design and maintenance practices, contributing to improved safety and reliability under extreme flight conditions.

Artem Karpenko, Yuriy Torba

In this study, a numerical model was developed for performing 3D computational fluid dynamics (CFD) simulations of the total temperature separation phenomenon in rotating flows, and the accuracy of this model was validated. This study aims to confirm that the CFD simulation results accurately reflect the physical processes occurring in a real swirling fluid flow. This study examines a rotating fluid flow in a counterflow Ranque-Hilsch vortex tube. The commercial three-dimensional CFD solver Ansys CFX 2024 R2 is used for calculations in a steady-state formulation. To determine the required mesh resolution, a grid convergence study was conducted using the Grid Convergence Index (GCI). The obtained indicators meet the GCI requirements in detailed studies with low error margins (GCI ≤ 5%). The selected mesh for further calculations comprises 13 million elements. This study includes simulations of the vortex tube at various cold stream mass fractions and with different turbulence models. Specifically, the standard k-ε model, the Shear Stress Transport (SST) model, and two Reynolds Stress Models were considered: the Baseline Reynolds Stress Model (BSL RSM) and the Speziale-Sarkar-Gatski Reynolds Stress Model (SSG RSM). The results of the 3D CFD simulations were compared with experimental data, particularly analyzing the difference in total temperatures between the hot and cold outlets of the vortex tube at varying static pressures at the hot outlet. Based on the analysis, all considered turbulence models are capable of detecting the total temperature separation in rotating flows. However, the standard k−ε turbulence model demonstrated the best agreement with the experimental data in terms of the degree of temperature separation. Therefore, it is recommended for use in energy separation calculations in rotating flows.

YANG Huixin, SONG Zeming, DONG Kaikai et al.

Regarding the robust trajectory planning and control of multi-constrained rendezvous for space tumbling targets in an elliptical orbit, a robust tracking model predictive control algorithm based on the "tube" is proposed. Firstly, the engineering constraints and uncertainties in the rendezvous process are analyzed and modeled, and the uncertainty propagation of the nominal system is quantified to obtain its compressed constraints. Then, a robust tracking model predictive rendezvous controller, consisting of a feedforward term and a feedback term, is designed. The feedforward term drives the system state to the target state with the least energy consumption and the best tracking accuracy. The feedback term ensures that the system still satisfies the engineering constraints under uncertainty. The digital simulation results indicate that the controller thus designed can successfully complete the elliptical orbit rendezvous task with a shorter prediction horizon and guaranteed stability. Compared with the classical robust model predictive control (MPC) algorithm, the proposed algorithm significantly reduces the computational complexity and has the advantages of autonomy, robustness and safety.

ZHANG Hui, DING Qiang, LIU Jide

The creep rupture properties of DD419 single crystal superalloys，fabricated at varying pouring temperatures were examined under conditions of 850 ℃/650 MPa，1050 ℃/190 MPa and 1100 ℃/130 MPa. SEM，EDS and TEM were used to analyze the microstructure and component segregation to study their effects on the durability. The results show that as the pouring temperature decreases，the primary dendrite spacing of the alloy widens，the eutectic content and the number of micropore increase，and the γ′ phase size diminishes. Under high temperature/low stress（1100℃/130 MPa），the γ′ phase size exerts a more pronounced influence on durability than do micropore and residual eutectic content. The finely dispersed γ′ phase enhances the alloy’s durability under all three test scenarios，with the alloy poured at 1500 ℃ exhibiting optimal durability. At intermediate temperature/high stress condition（1050℃/190 MPa），the γ′ phase is intersected by numerous dislocations, and dispersed γ′ phase may contribute to dislocation pile-ups. Concurrently，the alloy maintains good elongation at different pouring temperatures；however，as the pouring temperature decreases，section shrinkage decreases under all three test conditions. Pouring temperature has a negligible impact on the the alloy’s fracture morphology. Specifically, the γ′ phase near the fracture surface of the specimen tested under 850 ℃/650 MPa condition remains cubic morphology，with a mixed -mode fracture mechanism. Under other durability parameters，the γ′ phase assumes a rafted configuration，leading to an all-micropore aggregation fracture mechanism.

Hao Cui, Ke Zhang, Minghu Tan et al.

We present a novel approach to generating a cooperative guidance strategy using deep reinforcement learning to address the challenge of cooperative multi-missile strikes under uncontrollable velocity conditions. This method employs the multi-agent proximal policy optimization (MAPPO) algorithm to construct a continuous action space framework for intelligent cooperative guidance. A heuristically reshaped reward function is designed to enhance cooperative guidance among agents, enabling effective target engagement while mitigating the low learning efficiency caused by sparse reward signals in the guidance environment. Additionally, a multi-stage curriculum learning approach is introduced to smooth agent actions, effectively reducing action oscillations arising from independent sampling in reinforcement learning. Simulation results demonstrate that the proposed deep reinforcement learning-based guidance law can successfully achieve cooperative attacks across a range of randomized initial conditions.

Felix Embacher, David Holtz, Jonas Uhrig et al.

Autonomous vehicles often have varying camera sensor setups, which is inevitable due to restricted placement options for different vehicle types. Training a perception model on one particular setup and evaluating it on a new, different sensor setup reveals the so-called cross-sensor domain gap, typically leading to a degradation in accuracy. In this paper, we investigate the impact of the cross-sensor domain gap on state-of-the-art 3D object detectors. To this end, we introduce CamShift, a dataset inspired by nuScenes and created in CARLA to specifically simulate the domain gap between subcompact vehicles and sport utility vehicles (SUVs). Using CamShift, we demonstrate significant cross-sensor performance degradation, identify robustness dependencies on model architecture, and propose a data-driven solution to mitigate the effect. On the one hand, we show that model architectures based on a dense Bird's Eye View (BEV) representation with backward projection, such as BEVFormer, are the most robust against varying sensor configurations. On the other hand, we propose a novel data-driven sensor adaptation pipeline based on neural rendering, which can transform entire datasets to match different camera sensor setups. Applying this approach improves performance across all investigated 3D object detectors, mitigating the cross-sensor domain gap by a large margin and reducing the need for new data collection by enabling efficient data reusability across vehicles with different sensor setups. The CamShift dataset and the sensor adaptation benchmark are available at https://dmholtz.github.io/camshift/.

Jin Zhang, Xiaoran Qin, Ming Zhang Department of Transportation Engineering et al.

With the increasing development of intelligent transportation systems and advancements in aviation technology, the concept of Advanced Air Mobility (AAM) is gaining attention. This study aims to improve operational safety and service quality within Urban Air Mobility (UAM) through a trajectory-based operation (TBO). A multi-layer operational risk assessment model is introduced to capture the effects of aircraft failure scenarios on critical urban entities, including ground personnel, vehicles, and in-flight UAVs (unmanned aerial vehicles). Based on this, a single-aircraft track planning model is designed to balance operational risk and transportation cost under the performance constraints of eVTOL (electric Vertical Take-off and Landing) aircraft. A customized track planning algorithm with safety buffer zones is used to identify the most efficient flight paths. Additionally, a multi-aircraft scheduling optimization model is proposed to minimize delays and reduce mid-air collision risks. Experimental results show that the presented approach improves both efficiency and safety, providing practical solutions for UAM operations.

Yang Zhou, Z. Bai, Lindy Zeng et al.

Stress concentration is a complex problem in material mechanics, especially in Aeronautics and Astronautics applications, the concentrated stress will do great harm to the safe operation of the launch vehicle. Therefore, it is a common practice to simulate and verify the structural design or manufacturing process in the use of materials and dynamics. This paper first introduces the problem of stress concentration in aerospace, demonstrates the necessity and importance of studying this problem, at the same time, puts forward a stress concentration solution based on finite element method, which simplifies the geometric model by using symmetrical characte-ristics, so as to reduce the batch data of analysis. Compared with the conventional analysis using Patran and NASTRAN, its geometric model has a higher degree of discretization, The superiority and reliability of this method are verified by two examples. The results show that this finite element method is more accurate and effective in dealing with stress concentration problems, especially complex geometric models, and has high value for engineering practice.

Tong Liu, Huawei Lou, Haitao Huang

The Stewart structure with six degrees of freedoms (6-DOF), which has the advantages of big bearing capacity, strong stiffness and high accuracy, has been widely used in the indoor automatic assembly of aeronautics and astronautics. In this paper, a novel design method for the Stewart structure with limited installation space is proposed to extend the workspace by parameter optimization using genetic algorithm, and the performance is further verified via co-simulation. Firstly, the target function, which describes the workspace that satisfies both kinematic and dynamic constraints, is built with four parameters: the joints circle diameters and the angles between adjacent joints of both bottom and top platforms. Subsequently, the genetic algorithm is used to optimize the target function with the four parameters whose bounds are limited by vehicle-mounted installation space and structure interference. Finally, the geometry model is built in LMS. Motion and is studied using co-simulation method with the hydraulic system in AMEsim and control system in Simulink. With the cylinders motion accuracy within ±0.5 mm, the final position of top platform is consistent with the theoretical calculations, and the position error could be controlled within ±0.7 mm.

Tianhao Hou, Hongyan Xing, Wei Gu et al.

Wind speed and direction are critical meteorological elements. Multi-rotor unmanned aerial vehicles UAVs are widely used as a premium payload platform in meteorological monitoring. The meteorological UAV is able to improve the spatial and temporal resolution of the elements collected. However, during wind measurement missions, the installed anemometers are susceptible to interference caused by rotor turbulence. This paper puts forward a wind pressure orthogonal decomposition (WPOD) strategy to overcome this limitation in three ways: the location of the sensors, a new wind measurement method, and supporting equipment. A weak turbulence zone (WTZ) is found around the airframe, where the turbulence strength decays rapidly and is more suitable for installing wind measurement sensors. For the sensors to match the spatial structure of this area, a WPOD wind measurement method is proposed. An anemometer based on this principle was mounted on a quadrotor UAV to build a wind measurement system. Compared with a standard anemometer, this system has satisfactory performance. Analysis of the resulting data indicates that the error of the system is ±0.3 m/s and ±2° under hovering conditions and ±0.7 m/s and ±5° under moving conditions. In summary, WPOD points to a new orientation for wind measurement under a small spatial–temporal scale.

Alessandro A. Quarta, Giovanni Mengali

The aim of this paper is to investigate the performance of a robotic spacecraft, whose primary propulsion system is an electric solar wind sail (E-sail), in a mission to a heliostationary point (HP)—that is, a static equilibrium point in a heliocentric and inertial reference frame. A spacecraft placed at a given HP with zero inertial velocity maintains that heliocentric position provided the on-board thrust is able to counterbalance the Sun’s gravitational force. Due to the finite amount of storable propellant mass, a prolonged mission toward an HP may be considered as a typical application of a propellantless propulsion system. In this respect, previous research has been concentrated on the capability of high-performance (photonic) solar sails to reach and maintain such a static equilibrium condition. However, in the case of a solar-sail-based spacecraft, an HP mission requires a sail design with propulsive characteristics that are well beyond the capability of current or near-future technology. This paper shows that a medium-performance E-sail is able to offer a viable alternative to the use of photonic solar sails. To that end, we discuss a typical HP mission from an optimal viewpoint, by looking for the minimum time trajectory necessary for a spacecraft to reach a given HP. In particular, both two- and three-dimensional scenarios are considered, and the time-optimal mission performance is analyzed parametrically as a function of the HP heliocentric position. The paper also illustrates a potential mission application involving the observation of the Sun’s poles from such a static inertial position.

Dawei Wang, Weizi Li, Lei Zhu et al.

Intersections are essential road infrastructures for traffic in modern metropolises. However, they can also be the bottleneck of traffic flows as a result of traffic incidents or the absence of traffic coordination mechanisms such as traffic lights. Recently, various control and coordination mechanisms that are beyond traditional control methods have been proposed to improve the efficiency of intersection traffic by leveraging the ability of autonomous vehicles. Amongst these methods, the control of foreseeable mixed traffic that consists of human-driven vehicles (HVs) and robot vehicles (RVs) has emerged. We propose a decentralized multi-agent reinforcement learning approach for the control and coordination of mixed traffic by RVs at real-world, complex intersections -- an open challenge to date. We design comprehensive experiments to evaluate the effectiveness, robustness, generalizablility, and adaptability of our approach. In particular, our method can prevent congestion formation via merely 5% RVs under a real-world traffic demand of 700 vehicles per hour. In contrast, without RVs, congestion will form when the traffic demand reaches as low as 200 vehicles per hour. Moreover, when the RV penetration rate exceeds 60%, our method starts to outperform traffic signal control in terms of the average waiting time of all vehicles. Our method is not only robust against blackout events, sudden RV percentage drops, and V2V communication error, but also enjoys excellent generalizablility, evidenced by its successful deployment in five unseen intersections. Lastly, our method performs well under various traffic rules, demonstrating its adaptability to diverse scenarios. Videos and code of our work are available at https://sites.google.com/view/mixedtrafficcontrol

Arya Rachman, Jürgen Seiler, André Kaup

Autonomous vehicles are equipped with a multi-modal sensor setup to enable the car to drive safely. The initial calibration of such perception sensors is a highly matured topic and is routinely done in an automated factory environment. However, an intriguing question arises on how to maintain the calibration quality throughout the vehicle's operating duration. Another challenge is to calibrate multiple sensors jointly to ensure no propagation of systemic errors. In this paper, we propose CaLiCa, an end-to-end deep self-calibration network which addresses the automatic calibration problem for pinhole camera and Lidar. We jointly predict the camera intrinsic parameters (focal length and distortion) as well as Lidar-Camera extrinsic parameters (rotation and translation), by regressing feature correlation between the camera image and the Lidar point cloud. The network is arranged in a Siamese-twin structure to constrain the network features learning to a mutually shared feature in both point cloud and camera (Lidar-camera constraint). Evaluation using KITTI datasets shows that we achieve 0.154 ° and 0.059 m accuracy with a reprojection error of 0.028 pixel with a single-pass inference. We also provide an ablative study of how our end-to-end learning architecture offers lower terminal loss (21% decrease in rotation loss) compared to isolated calibration

Fabian Immel, Richard Fehler, Mohammad M. Ghanaat et al.

High Definition (HD) maps are necessary for many applications of automated driving (AD), but their manual creation and maintenance is very costly. Vehicle fleet data from series production vehicles can be used to automatically generate HD maps, but the data is often incomplete and noisy. We propose a system for the generation of HD maps from vehicle fleet data, which is tolerant to missing or misclassified detections and can handle drives with multiple routes, generating a single complete map, model-free and without prior reference lines. Using randomly selected drives as pivot drives, a step-wise lateral sampling of detections is performed. These sampled points are then clustered and aligned using Expectation Maximization (EM), estimating a lateral offset for each drive to compensate localization errors. The clustered points are replaced with the maxima of their probability density function (PDF) and connected to form polylines using a modified rectangular linear assignment algorithm. The data from vehicles on varying routes is then fused into a hierarchical singular map graph. The proposed approach achieves an average accuracy below 0.5 meters compared to a hand annotated ground truth map, as well as correctly resolving lane splits and merges, proving the feasibility of the use of vehicle fleet data for the generation of highway HD maps.