Measurement noise in flight trajectory data degrades the accuracy of aircraft control acceleration identified by estimation algorithms such as the Kalman filter. This increased identification error constrains the reliability of subsequent tasks that rely on control acceleration information, including target recognition and trajectory prediction. This paper proposes a Chaotic Adaptive Bald Eagle Search (CABES)–based identification method. Flight trajectories are segmented using climb and heading change rates to isolate deterministic dynamic parameters, mitigating dynamic uncertainty effects. A global parametric dynamic model represents aircraft-controlled responses through velocity-frame acceleration components. CABES enhances optimization via chaotic mapping for global exploration and adaptive parameter balancing search efficiency. The applicability of the proposed method is verified via CRLB-based identifiability analysis, model-mismatch analysis, Monte Carlo simulations, and real flight trajectory experiments. CABES achieves higher acceleration identification accuracy than EKF, PSO, and BES, with superior error performance under real trajectory disturbances.
To achieve precise pointing of the optical axis of the MUSICO payload aboard the space station along a preset ground push-broom trajectory and enhance its capability to measure target gas concentrations over a large spatial range, this study designs a two-dimensional turntable adaptable to the external platform of the space station cabin. The turntable adopts a “pitch–azimuth” axis system as its core mechanical structure and innovatively integrates three key electronic units—overall main control, secondary power conversion, and servo control into the turntable interior. In this work, the mechanical structure design of the turntable is completed, and computational verification confirms that its pointing accuracy meets the design requirements. Furthermore, multiparameter optimization is conducted on the main load-bearing components, yielding significant results: The overall weight of the turntable assembly is reduced by 11%, and the first-order natural frequency is increased by 11.5%, which effectively improves the structural stability and dynamic stiffness of the turntable. Mechanical tests verify that the turntable can withstand the mechanical environmental conditions during the launch phase. All performance indicators of the turntable designed in this study meet the requirements of the scientific detection mission of the MUSICO payload, providing reliable mechanical support for the subsequent efficient implementation of target gas concentration measurement tasks by the payload.
Multicore processors (MCPs) exhibit significant potential as airborne processors in Integrated Modular Avionics (IMA) systems. For IMA systems employing MCPs, a dual-layer scheduling architecture is adopted to ensure task real-time performance. However, existing scheduling designs based on ARINC 653 primarily focus on the spatial allocation of tasks across different partitions, thereby balancing high-safety-critical tasks only in the spatial domain. They have not addressed the uneven distribution of safety-critical tasks across the temporal dimension. This paper proposes an optimized scheduling approach that considers the safety criticality of tasks in the time domain. Leveraging the ARINC 653 spatial-temporal partitioning mechanism and by quantifying the safety criticality of avionics functions, we develop a task scheduling and high-safety-critical task balancing model for MCPs. This paper adopts an asymmetric multiprocessing (AMP) architecture and employs dual discretization in both spatial and temporal dimensions to achieve statistical balancing of high-safety-critical tasks across multicore IMA system. We conduct experiments on a multinode server that simulates the airborne environment and employ DQN as an offline optimizer to generate static scheduling tables. Experimental results show that, while strictly satisfying the real-time constraints of all task instances, this paper significantly improves the balance of safety criticality across time blocks in the major time frame, achieving improvements of 41.85%, 48.38%, and 51.84% for the three cores, respectively. This work offers a new perspective for enhancing temporal risk management and safety assurance in multicore integrated core processors for safety-critical avionics applications.
Large deployable mesh reflectors (DMRs) play a vital role in spaceborne communication and observation systems because of their low mass and compact stowage capability. Nevertheless, maintaining stringent surface accuracy while mitigating pronounced thermal influences in the orbital environment remains a persistent engineering challenge. This study develops a spatially discretized thermal modeling framework that incorporates direct solar radiation, Earth albedo, a prescribed equivalent Earth-infrared heat input, section-dependent solar-exposure factors, structural self-shadowing, and reflector emission to space. The proposed method is implemented to modeling of an 817-node center-feed reflector and an 849-node offset-feed reflector. Each reflector is discretized into one to 10 spatial sections and analyzed over four idealized geostationary-orbit eclipse-season cycles. Temperature distributions are evaluated at representative spatial nodes, and the influence of spatial discretization is quantified using normalized absolute temperature differences referenced to the 10-section model. For the center-feed reflector, the section-dependent temperature field is further converted into thermal strains and thermally modified stress-free cable lengths, and the resulting shape distortion of its reflecting surface is evaluated using the direct root-mean-square surface error.
The hose–drogue system is highly sensitive to aerodynamic disturbances and aircraft maneuvers during aerial refueling, posing risks to operational safety. This paper develops a multibody dynamic model based on the absolute nodal coordinate formulation (ANCF), incorporating standard atmospheric models, tanker wake effects, and a compliant penalty-based drogue–probe constraint. The transient responses to tanker roll/pitch maneuvers and receiver vertical disturbances are systematically analyzed. Our findings clarify the nonlinear dynamics of the hose–drogue system considering tanker/receiver's attitude changes and provide a foundation for structural risk assessment and control design.
Asteroids have the characteristics of weak gravity, unknown surface topography and physical properties, and low cohesion of asteroid rocks, which make long-term attachment to an asteroid challenging. Considering these characteristics, a claw-spine attachment device was designed, and the effect of anchoring force generation and structural optimization was analyzed. Firstly, a static model between the claw-spine mechanism and the rough asteroid surface was established. Through the static model, the effect of the structural parameters of the claw-spine mechanism on its anchoring performance in the stable state of anchorage was obtained. The key structural parameters affecting the generation of anchoring force were identified to be used as design variables for subsequent analysis. Secondly, the ground test platform of the claw-spine mechanism was built. And the rationality of the claw-spine mechanism and the correctness of the dynamic simulation model were verified by the multicondition ground anchoring test. Third, through multicondition dynamic simulation of the anchoring process of the claw-spine mechanism, the rope tension is identified as the most influential design variable affecting anchoring force. The expression of the anchoring force function was obtained by fitting the simulation results. Constraints are proposed based on the requirement for the spine tip to catch a microprotrusion and the indentation hardness of the asteroid surface. The particle swarm optimization algorithm was used to obtain the optimal solution of a set of structural parameters of the claw-spine mechanism, and the maximum anchoring force of the claw-spine mechanism is 38.2514 N after optimization.
To investigate the mechanical damage of solid propellants under low-temperature conditions, an equivalent modeling approach was employed to establish a mesoscopic particle–embedded model of the hydroxyl-terminated polybutadiene (HTPB) propellant. Uniaxial tensile numerical simulations were conducted on the propellant under varying strain rates and temperatures to examine the debonding damage and binder matrix brittle fracture phenomena at low temperatures. Additionally, a comparative analysis was performed on the matrix damage of propellants containing initial voids under low-temperature conditions. The results indicate that the debonding damage stress of the propellant increases with decreasing temperature and increasing strain rate. Under low-temperature and high-strain-rate conditions, two distinct damage modes were observed: solid particle debonding and binder matrix fracture, accompanied by a pronounced double-peak phenomenon in the stress–strain curve. For propellants containing voids, the stress required to initiate particle debonding and binder matrix fracture was found to decrease, whereas the extent of damage became more severe. Through numerical simulations of solid propellants under low-temperature conditions, this paper elucidates the processes of particle debonding and binder matrix fracture, as well as the stress–strain relationship during tensile loading, from a mesoscopic perspective. These findings contribute to advancing research in micromechanical numerical simulations of solid propellants.
A mathematical model and software for online modeling of the parameters and characteristics of a cruise missile with a pulsejet engine to evaluate the potential use of this type of engine in high-speed unmanned aerial vehicles (UAVs) were developed. The model describes the acceleration after launch and the steady horizontal flight of a cruise missile and allows for the calculation of its flight parameters. Using the developed software, preliminary mathematical flight modeling was performed. The results showed a high degree of agreement between the calculated characteristics and parameters of the cruise missile, including speed, range, and flight endurance, and the technical specifications of the actual Fi-103 (V-1) missile. The simulations demonstrated that the most efficient use of a pulsejet engine occurs at a thrust-to-weight ratio of 0.14–0.15 and a payload weight of 100%–300% of the engine thrust. A combination of parameters allowing a flight range of up to 700 km, corresponding to the operational range of some modern turbojet-powered cruise missiles, was identified. It was also noted that the use of a pulsejet engine can be effective only if mass production is achieved and the intake valve durability problem is resolved.
Traditional ascent trajectory optimization methods for hypersonic vehicles suffer from a high computational burden. This paper proposes an intelligent optimization algorithm based on deep reinforcement learning, where the offline-trained neural network enables fast trajectory generation and online deployment while maintaining all terminal, control, and path constraints. First, the trajectory optimization model—including ascent dynamics, program angle form, and multiple constraints—is established. Then, the trajectory optimization problem is formulated as a Markov decision process, rendering it solvable by reinforcement learning. To accelerate training, the twin delayed deep deterministic policy gradient (TD3) algorithm is improved via behavioral cloning and learning from demonstrations (fD). Specifically, an expert policy is pretrained via behavioral cloning from the optimization dataset, which can warm-start the TD3's Actor. As a result, the early-stage exploration for successful samples is effectively enhanced. Additionally, a supervised auxiliary loss is incorporated into the Actor's update to strengthen the reward signal and improve TD3 convergence. Finally, simulations demonstrate that the proposed behavioral cloning-augmented TD3 from demonstrations (BC-TD3fD) algorithm converges rapidly, whereas the original TD3 fails. Compared with standard baselines, BC-TD3fD reduces the average offline optimization time from 19.268 to 0.149 s, achieving a substantial acceleration over sequential quadratic programming, and decreases the online per-step computation time to 3.204 ms, representing a 94.2% reduction relative to model predictive control based on quadratic programming, thereby ensuring deterministic real-time performance.
This paper describes the rendezvous and docking of spacecraft under uncertain conditions. Vision-based navigation is employed as a cost-effective and practical method to enhance the accuracy of navigation and control in order to estimate the relative position and attitude of the spacecraft. Because of the limited sampling frequency of vision-based sensors, their standalone use fails to provide sufficient accuracy. Therefore, a discrete-time Kalman filter is employed to fuse vision-based navigation data with noisy system dynamics. For spacecraft position and attitude control, an SDRE controller is implemented, and the on–off behavior of cold gas thrusters is modeled using PWPF modulators. To reduce fuel consumption and enhance agility, a multiobjective genetic algorithm is used to optimize the SDRE gains as well as the PWPF parameters. Simulation results demonstrate that this approach not only improves docking accuracy but also reduces impact velocity and increases robustness to uncertainties, indicating its potential for autonomous docking missions.
Unmanned autonomous air combat, characterized by advantages such as zero casualties, low cost, and high maneuverability, represents the primary direction for future air combat development. The key to achieving victory in beyond-visual-range autonomous air combat lies in the drone's ability to accurately analyze threats, advantages, and gains based on the situational awareness of both sides, and subsequently perform rapid occupying point calculation and guidance control. Considering that existing studies do not take radar detection blind spots into account when calculating optimal occupancy points, this article proposes a dual-aircraft cooperative occupying method based on radar detection blind spots and Bézier trajectory optimization. This method is constructed in three steps. First, considering actual radar characteristics, a missile launch zone model that eliminates radar blind spots is established. Next, based on the launch zone model, dual-aircraft cooperative air combat superiority, and threat-gain evaluation functions, the optimal cooperative occupying points are computed using a particle swarm optimization (PSO) algorithm. Finally, accounting for the physical constraints of the aircraft (minimum turning radius, tangential acceleration, and speed limits), a spatiotemporal Bézier cooperative trajectory planning method is developed to ensure synchronized arrival of both aircraft at the designated occupying points. Simulation analyses of two-versus-two scenarios demonstrate that when enemy aircraft approach at varying distances and headings, the proposed cooperative occupying method can calculate reasonable occupying points, whereas the spatiotemporal cooperative Bézier trajectory planning ensures simultaneous arrival at the target points, effectively encircling the enemy aircraft.
Efficient scheduling of airport ground support vehicles is critical for the smooth operation of major hub airports. This paper addresses the complex scheduling scenario involving multiple vehicle types, multiple depots, and task coupling constraints by proposing a novel vehicle routing problem formulation termed MDMT-VRP-TC (Multi-Depot, Multi-Type Vehicle Routing Problem with Task Coupling). To solve this problem, we design a hybrid intelligent optimization algorithm (RL-ALNS) that integrates deep reinforcement learning (proximal policy optimization [PPO]) with adaptive large neighborhood search (ALNS). The algorithm uses reinforcement learning to dynamically guide the selection of destroy and repair operators, configure destruction severity, and set acceptance criteria, significantly improving search efficiency and solution quality. Experimental results based on real-world data from Chengdu Tianfu International Airport demonstrate that the proposed method outperforms traditional exact algorithms and heuristic approaches in terms of solution quality, computational efficiency, and robustness. It also effectively handles dynamic disruptions such as flight delays, providing theoretical support and practical guidance for lean management of airport ground services.
This study proposes an onboard localization method for planetary aerial vehicles, specifically targeting Mars exploration, based solely on a monocular visible-light camera. This method creates a terrain complexity distribution map using satellite imagery and estimates position by matching it with onboard images. Unlike existing template-matching approaches, which are sensitive to lighting or atmospheric conditions, this method is designed for natural terrains such as those on Mars, which have minimal texture. By focusing on terrain roughness and complexity, the algorithm demonstrates robustness against differences in lighting and imaging conditions. Simulations conducted using a virtual fractal terrain with self-similarity demonstrated that this method is robust against changes in brightness distribution caused by differences in terrain texture details and light-source direction. Furthermore, simulations using images of actual Martian canyons and plains demonstrated that accurate position estimation is possible.
The use of uncrewed aerial vehicles (UAVs) for autonomous infrastructure inspections offers significant advantages in terms of safety, efficiency, and cost. However, the effectiveness of such operations is heavily influenced by the availability of Global Navigation Satellite System (GNSS) coverage, particularly in challenging environments such as under bridges or urban canyons. This work evaluates the feasibility of automated UAV pillar inspections through GNSS coverage analysis in two representative case studies: the Freixo Bridge in Porto, Portugal, and the Ribadavia viaduct on the A-52 motorway in Galicia, northwestern Spain. The proposed methodology uses satellite almanac information to estimate GNSS availability beneath these structures and to identify suitable flight windows for autonomous inspection missions. The analysis accounts not only for satellite visibility, but also for the effects of local orography, terrain characteristics, and bridge geometry, which can significantly affect signal propagation through occlusions, reflections, and multipath. In addition, the study includes a comparison between GPS-only and multiconstellation configurations in order to assess the benefits of increased satellite availability in partially occluded environments. To improve the realism of the coverage estimation, experimental GPS and multiconstellation measurements collected beneath the Ribadavia viaduct were used to calibrate and refine the prediction algorithm under real signal-blocking conditions.This approach provides a robust framework for future UAV operations, facilitating adaptive route planning and informed decision-making to ensure reliable and efficient inspections in GNSS-constrained environments.
This paper proposes a molecular-dynamics-inspired decentralized control framework for cooperative path planning and formation coordination of unmanned aerial vehicle (UAV) swarms in dynamic environments. Each UAV is modeled as an interacting particle governed by continuous virtual forces, including attraction, repulsion, and angular alignment terms, yielding an energy-consistent and physically interpretable interaction model. The proposed framework unifies formation keeping, obstacle avoidance, and topology reconfiguration within a single dynamic formulation. By regulating interagent bonding strengths, the swarm can autonomously split, merge, and reassemble without centralized coordination or global communication. Software-in-the-loop simulations under multitarget and adversarial scenarios demonstrate rapid convergence, stable formation maintenance, low communication overhead, and robust collision avoidance. The results indicate that the proposed approach provides a scalable and reliable solution for real-time multi-UAV swarm operations.
During the terminal docking phase of probe-and-drogue aerial refueling, drogue motion is strongly affected by receiver-induced bow-wave interference and atmospheric turbulence, which can degrade docking accuracy and operational safety. To improve disturbance rejection under this close-proximity condition, this paper proposes an annular-wing-based actively stabilized drogue with independent pitch and yaw actuation. A time-domain simulation framework is established by integrating a multibody dynamic model of the hose-drogue system, a CFD-based receiver bow-wave disturbance model, and a Dryden atmospheric-turbulence model. In the bow-wave representation, both local flow velocity and air density are obtained through spatial interpolation of the precomputed flow field, enabling a more complete description of the local aerodynamic environment around the drogue. Based on the aerodynamic characteristics of the annular-wing control surface, a control-oriented linearized model is constructed, and an LQR controller is designed to suppress drogue position and attitude deviations during terminal docking. Numerical simulations are conducted under turbulence-only, bow-wave-only, and combined disturbance conditions. The results show that the proposed configuration exhibits favorable aerodynamic decoupling characteristics near the nominal docking state and can significantly reduce drogue lateral and vertical excursions. Under the representative disturbance cases considered, the peak displacements are suppressed from several tenths of a meter to only a few centimeters, with reduction ratios of approximately 94%-97%, while the required actuator deflections and angular rates remain within practically realizable ranges. These results demonstrate that the proposed annular-wing stabilization concept can effectively enhance drogue stability and docking robustness, providing a compact and practical solution for terminal aerial refueling operations.
Electric pump systems are increasingly being considered as an alternative to turbopumps in liquid rocket engines. Interest is driven by the potential for reduced dry mass and increased system efficiency. This systematic review analyzes 89 peer-reviewed publications from 2008 to 2025. The applicability limits of EPS architecture are defined based on normalized metrics that ensure comparability of results. Sources were selected through a systematic search of leading scientific databases in accordance with the PRISMA procedure. The “battery-to-thrust ratio” indicator was introduced for interarticle comparison. A reduction in the “dry” weight of the engine by 0.23–0.46 kg/kN in the thrust range of 10–100 kN is shown, mainly due to the elimination of the hot turbomachine and a reduction in the number of high-temperature and mechanically loaded components. The current level of EPS already meets the requirements of light and medium launch vehicles. Further development is linked to an increase in the specific energy of batteries to 600 W·h/kg, the improvement of cryogenic heat dissipation solutions, the qualification of large-sized products manufactured using additive manufacturing techniques, and the bench validation of multiphysical digital twins. The implementation of these areas will accelerate the transition to serial EPS engines and expand the possibilities for creating reusable and more environmentally friendly rocket systems.
Multirotor aerodynamic interference exerts a remarkable influence on the aerodynamic characteristics of distributed rotorcraft. It triggers complex flow structures and performance changes, whereas the inherent relationship between flow field characteristics and performance evolution has not yet been clarified. This study performs unsteady numerical simulations of dual-rotor systems with varying axial spacing h and radial spacing l, focusing on the aerodynamic performance and interference mechanisms of coaxial and staggered configurations. Results show that the aerodynamic behavior of dual-rotor systems is governed primarily by geometric parameters and exhibits strong robustness to rotor-speed variations. The parameters h and l are identified as the dominant factors determining interference intensity and performance distribution. In the coaxial configuration, a clear performance asymmetry exists between the upper and lower rotors. Increasing axial spacing enhances the upper rotor efficiency through stronger radial inflow, whereas the lower rotor suffers from increased wake downwash, leading to only modest changes in overall hovering efficiency. In the staggered configuration, the lower rotor plays the primary role in determining system performance and is highly sensitive to radial spacing. As l increases, blade-vortex and vortex-vortex interactions are greatly reduced, transitioning the flow field from a strong-interference regime toward a near-wake-dominated state and gradually improving performance. At a rotational speed of 3000 rpm and h = 0.3, comparative analysis indicates that within the optimal radial-spacing range l = 1.8-2.4, the staggered configuration achieves an up to 24.9% higher overall hover efficiency (FM) compared with the coaxial configuration, and the lower rotor efficiency increases by up to 58.2%. These improvements significantly alleviate the lower-rotor efficiency limitation inherent in coaxial systems, demonstrating the effectiveness of the staggered layout in balancing structural compactness and aerodynamic performance.
With the rapid development of space technology, rigid-flexible coupling systems are widely used in aerospace structures. Attitude-vibration coupling control requires a deep understanding of the rigid-flexible coupling dynamics of these systems. In the presented work, a center rigid body coupled with a rectangular thin plate was studied. The dynamic equation derived based on Hamilton's principle of the rigid-flexible couple structure was introduced, and the direct piezoelectric effect was introduced to detect the sensing signal of the dynamic behavior of the flexible thin plate to reveal the rigid-flexible coupling effect. In case studies, the dynamic response of the rigid-flexible couple thin plate was investigated, the modal sensing signal generated by the piezoelectric patch, and the sensing signal induced by rigid body motion were explored with various piezoelectric sensor positions, sensor size, rigid-flexible coupling characteristics, and mechanical excitation properties. Analysis showed that piezoelectric sensor signals varied with sensor position, primarily governed by modal shape functions. But in rigid-flexible coupled systems, external excitation also significantly impacts piezoelectric sensor effectiveness. With increasing the rigid body inertia, the output signal could decrease greatly, whereas greater mechanical excitation amplifies it. Notably, moment-induced rotation generates substantially stronger signals than force-induced translation, highlighting the system's dynamic complexity.
