Astrodynamics最新文献

Private lunar missions are faced with the challenge of robust autonomous navigation while operating under stringent constraints on mass, power, and computational resources. This work proposes a motion-field inversion framework that uses optical flow and rangefinder-based depth estimation as a lightweight CPU-based solution for egomotion estimation during lunar descent. We extend classical optical flow formulations by integrating them with depth modeling strategies tailored to the geometry for lunar/planetary approach, descent, and landing—specifically, planar and spherical terrain approximations parameterized by a laser rangefinder. Motion field inversion is performed through a least-squares framework, using sparse optical flow features extracted via the pyramidal Lucas–Kanade algorithm. We verify our approach using synthetically generated lunar images over the challenging terrain of the lunar south pole, using CPU budgets compatible with small lunar landers. The results demonstrate accurate velocity estimation from approach to landing, with sub-8% error for complex terrain and on the order of 1% for more typical terrain, as well as performances suitable for real-time onboard applications. This framework shows promise for enabling robust, lightweight onboard navigation for small lunar missions.

The emergence of sub-CubeSat spacecraft, such as PocketQubes and ChipSats, enables new possibilities for cost-effective and scalable space missions. Their miniaturisation allows rapid deployment of satellite swarms, facilitating distributed scientific measurements and technological operations which are optimal for swarm missions. Previous research has explored spacecraft swarms and beamforming separately, often assuming fixed satellite configurations and ignoring orbital dynamics, control, and system constraints. No studies have assessed beamforming performance in swarms with relaxed positioning and attitude requirements. This paper presents the first integrated analysis combining a sub-CubeSat satellite platform, swarm mission design, swarm-keeping strategy, and distributed beamforming. The paper also evaluates beamforming performance using a sub-CubeSat-sized platform, highlighting the challenges and advantages of using such miniaturised platforms. First, the low Earth orbit dynamics of a swarm is modelled, including orbital perturbations, and a swarm-keeping strategy using a micropropulsion system is adopted to control the swarm geometry to meet the beamforming requirements. By integrating swarm dynamics and control with beamforming principles, the beamforming performance is evaluated. The paper shows that beamforming is achievable even with the limited capabilities of PocketQubes, by adjusting swarm parameters such as element count, intersatellite spacing, frequency, type of antenna, and antenna pointing, among others.

This work considers the problem of autonomous contingency station-keeping in lunar halo orbits under engine failures and communication delays with the Earth. The objective is to develop a control design methodology that autonomously maps observations to control actions, enabling the spacecraft to return to its nominal orbit while adapting to errors in both the magnitude and direction of thrust. We investigate various measurement sets that include direction, range, and radial velocity with respect to the Moon. The proposed control function is parameterized by a neural network and optimized via reinforcement meta-learning methods, which are well suited to handling uncertainty and contingencies. This approach is applicable to missions in both unstable and stable halo orbits associated with the L1 and L2. The results include performance analysis of the designed control functions, characteristic velocities required for station-keeping, and station-keeping precision. Findings are presented for both unstable and nearly stable halo orbits that are used in real missions.

Space-based gravitational-wave detection missions typically deploy three spacecraft in a widely spaced triangular formation in deep-space heliocentric or high Earth orbits. Maintaining high-precision coherence across this distributed, large-scale, and multi-degree-of-freedom system is critical to long-term, stable, and precise detector operations. High-accuracy orbit determination is foundational to mission success. Although a variety of tracking and measurement techniques exist, achievable orbit-determination accuracy is constrained by tracking coverage, systematic measurement errors, formation geometry, orbit-control capability, and the geometry of ground-based tracking networks. This paper presents a systematic overview of orbit-determination requirements for different mission architectures, analyses the performance and technical characteristics of ground-based and space-based tracking methods applicable to spacecraft in heliocentric and geocentric orbits, and discusses current challenges and future directions in high-precision orbit determination technologies to enable reliable, precise operation of space-based gravitational-wave detectors.

The Moon will play a central role in upcoming human space exploration efforts, with numerous missions planned in the near future. Designing trajectories for spacecraft in the cislunar region requires advanced methods that account for the dynamics of complex multi-body systems. This work presents a sequential convex programming (SCP) framework for optimizing both impulsive and continuous-thrust trajectories within the restricted three- and four-body models, with a subsequent transition to high-fidelity ephemeris models. We employ a pulsating–rotating frame, use a homotopic approach, and leverage the Sundman transformation to improve convergence. The proposed method is demonstrated through simulations in the Sun-perturbed Earth–Moon system and the Moon-perturbed Sun–Earth system. The high accuracy and robustness make SCP an excellent choice even for solving highly nonlinear problems.

Tethered-net capture of space tumbling debris involves complex deployment and collision processes. The ejection parameters and target characteristics are critical mission parameters in the contact process. To better assess the effects of these factors, a novel time-series predictive framework is proposed. First, an accurate dynamic model of tethered-net capture is established using the finite element method (FEM). Next, a tumbling target capture device is designed to experimentally validate the established dynamic model. Finally, a predictive model is constructed to predict the entire net contact process. It is proposed using a bidirectional long short-term memory (BiLSTM) neural network and optimized by the Bayesian optimization method. The results show that the predictive model exhibits strong generalization and high predictive accuracy, achieving 96% accuracy for the entire capture process and 98% accuracy during net deployment. Additionally, the sensitivity analysis indicates that the ejection speed and target tumbling rate are the key parameters during the contact process. This work provides valuable guidance for the design and control of future active debris removal (ADR) missions.

Crater-based navigation (CBN) uses the ubiquitous impact craters of the Moon observed on images as natural landmarks to determine the six degrees of freedom (6Dof) pose of a spacecraft. To date, CBN has primarily been studied in the context of powered descent and landing. These missions are typically short in duration, with high-frequency imagery captured from a nadir viewpoint over well-lit terrain. In contrast, lunar mapping missions involve sparse, oblique imagery acquired under varying illumination conditions over potentially year-long campaigns, posing significantly greater challenges for pose estimation. We bridge this gap with STELLA—the first end-to-end CBN pipeline for long-duration lunar mapping. STELLA combines a mask R-CNN-based crater detector, a descriptor-less crater identification module, a robust perspective-n-crater pose solver, and a batch orbit determination back-end. To rigorously test STELLA, we introduce CRESENT-365—the first public dataset that emulates a year-long lunar mapping mission. Each of its 15,283 images is rendered from high-resolution digital elevation models with SPICE-derived Sun angles and Moon motion, delivering realistic global coverage, illumination cycles, and viewing geometries. Experiments on CRESENT+ and CRESENT-365 show that STELLA maintains metre-level position accuracy and sub-degree attitude accuracy on average across wide ranges of viewing angles, illumination conditions, and lunar latitudes. These results constitute the first comprehensive assessment of CBN in a true lunar mapping setting and inform operational conditions that should be considered for future missions.

The alternative text for this image may have been generated using AI.

The geodesy of irregularly shaped small bodies presents fundamental challenges for gravitational field modeling, particularly as deep space exploration missions increasingly target asteroids and comets. Traditional approaches suffer from critical limitations: spherical harmonics diverge within the Brillouin sphere where spacecraft typically operate, polyhedral models assume unrealistic homogeneous density distributions, and existing machine learning methods like GeodesyNets and physics-informed neural networks (PINNGM) require extensive computational resources and training time. This work introduces MasconCubes, a novel self-supervised learning approach that formulates gravity inversion as a direct optimization problem over a regular 3D grid of point masses (mascons). Unlike implicit neural representations, MasconCubes explicitly model mass distributions while leveraging known asteroid shape information to constrain the solution space. Comprehensive evaluation on diverse asteroid models including Bennu, Eros, Itokawa, and synthetic planetesimals demonstrates that MasconCubes achieve superior performance across multiple metrics. Most notably, MasconCubes demonstrate computational efficiency advantages with training times approximately 40 times faster than GeodesyNets while maintaining physical interpretability through explicit mass distributions. These results establish MasconCubes as a promising approach for mission-critical gravitational modeling applications requiring high accuracy, computational efficiency, and physical insight into internal mass distributions of irregular celestial bodies.

Hopping exploration on the asteroid surface is a promising technique for comprehensive asteroid research. Controllable high-precision hopping is still challengeable due to asteroids’ irregular microgravity and rugged terrain. This paper proposes a feasible hopping exploration strategy with a flexible lander. The dynamics for efficient hopping simulation are introduced, which consist of the asteroid’s dynamics model, the lander’s flexible dynamics, the collision dynamics, and the coupling dynamics. One control strategy for high-precision hopping is proposed, which applies simple feed-back control and the nearest neighbor method to get the desired hopping moment. A series of signal hopping simulations on a smooth microgravity plane are discussed. The single hopping result combined the control strategy is applied for directional hopping exploration of smooth and rugged microgravity planes, showing the feasibility of the proposed exploration strategy. What is more, one scenario of hopping exploration on asteroid Bennu is presented, demonstrating the potential of the flexible lander in controllable high-precision asteroid hopping exploration.

This study explores the decision making capabilities of large language model (LLM) artificial intelligence (AI) agents to automate learning in lunar landing missions. In particular, the work investigates the use of AI agents to minimise human intervention in training a lunar lander by providing high-level strategic guidance to a reinforcement learning (RL) agent within the complex simulation environment of Kerbal Space Program (KSP). To that end, LLM AI agents are utilised to interpret a lander manual, extract key information to construct the reward function of the RL algorithm, and dynamically refine it based on training performance. A comparative case study evaluates the effectiveness of GPT-3.5-Turbo, GPT-4, and Meta-Llama-3-70B, in optimising the RL training process for lunar landings. Extending this approach, the study further explores AI-assisted hyperparameter optimisation (HPO) to streamline RL training. Instead of relying on traditional, computationally expensive methods like grid search and Bayesian optimisation, a zero-shot tuning approach is introduced, where an AI agent configures RL hyperparameters from a single instruction prompt without iterative refinements. Using GPT-4o, this method is applied to four RL algorithms (DQN, PPO, A2C, and SAC Discrete), demonstrating significant improvements in training efficiency, convergence speed, and reduced human effort, particularly for SAC Discrete. These findings highlight the potential of LLMs to automate both reward function design and hyperparameter tuning, thus advancing AI capabilities in space exploration and autonomous navigation tasks.

The alternative text for this image may have been generated using AI.