Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Yuzhu Bai has a total of 27 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    A High-Precision Adaptive Sequential Convex Programming Method for Non-Coplanar Transfer Trajectory Optimization in Constellation Aggregation

    by Zihui Ma, Rong Chen, Yuzhu Bai

    With the development of constellations and satellite chains, trajectory planning for constellation satellites has gradually attracted research attention. To address the infeasibility issues of trapezoidal sequential convex programming (T-SCP) in this strong non-convexity problem, this paper proposes high-precision adaptive SCP based on the Hermite Simpson method for Low-Earth-Orbit (LEO) constellation aggregation and aimed at enhanced local reconnaissance with practical constraints, including collision avoidance. The method introduces an adaptive trust region and slack variables, which avoid convergence failures due to strong nonlinearity or large state variations over long distances and accelerate the convergence rate. Furthermore, a third-order accurate Hermite Simpson method is adopted along with a adaptive iterative collision warning mechanism, improving accuracy while reducing computational cost. Simulation results demonstrate that the proposed method improves computational efficiency by 6% and 21% compared with the T-SCP and pseudospectral method, respectively. When the number of discrete nodes is 50, it achieves position errors of 284.7 m and velocity errors of 0.33 m/s, and these errors are far lower than those of T-SCP. Monte Carlo simulations across 10,000 scenarios validate the robustness of the proposed method with a 91% overall success rate and 100% for LEO.

  • Open Access

    Spacecraft Safe Proximity Policy Based on Graph Neural Network Safe Reinforcement Learning

    by Heng Zhou, Jingxian Wang, Monan Dong, Yong Zhao, Yuzhu Bai, Rong Chen

    Spacecraft safe proximity, as a critical component of on-orbit servicing missions, primarily encounters the following two challenges: the partial observability of the environment surrounding the service spacecraft and the necessity to evade uncertain obstacles. A safe reinforcement learning algorithm based on a graph neural network is proposed to address the constrained Markov decision problem in partially observable scenarios for spacecraft safe proximity missions. A graph neural network mechanism is introduced to solve the problem of dynamic variations in the quantity and location of obstacles in the observation area of the service spacecraft. The graph attention network is used to facilitate the extraction of feature information from the graph structure, which is then utilized as input for the subsequent reinforcement learning algorithm. The Soft Actor–Critic–Lagrangian algorithm is adopted to deal with the problems of tuning reward function parameters and balancing safety and optimality. By introducing Lagrange multipliers, the constrained optimization problem is transformed into an unconstrained optimization problem. In order to verify the effectiveness of the algorithm proposed in this paper, a spacecraft safe proximity environment model with dynamic obstacles is constructed, and the GAT-SACL algorithm proposed in this paper is validated by the Monte Carlo shooting method. The results show that the GAT-SACL algorithm possess excellent exploratory characteristics and delivers significant advantages in balancing optimality and safety.


  • Open Access

    Research on Spacecraft Observation and Counter-Observation Game Based on Clustering Multi-objective Optimization genetic Algorithm

    by Zhongxu Zheng, Rong Chen, Chenyu Shan, Yuzhu Bai, Yong Zhao

    With the increasing complexity of the space environment and continuous advancements in satellite maneuverability, orbital game theory has become a crucial research direction in aerospace engineering. To address the need for prior target information acquisition in missions such as space debris removal and on-orbit servicing, a scenario of observation and counter-observation similar to the pursuit-evasion game is proposed. Leveraging the Multi-objective optimization algorithm, we systematically investigate the observation and counter-observation game (OCOG) problem under continuous-thrust conditions for two satellites. First, a sequential decision-making model for continuous-thrust maneuvers is established, transforming the OCOG problem into a bilateral optimization problem incorporating multiple constraints, including maneuverability, illumination conditions, relative distance, line-of-sight (LOS) angular velocity, and relative velocity. Second, a switching-type payoff function characterizing effective imaging duration is constructed, and a hyperbolic tangent function is introduced to achieve continuous smoothing of discrete switching characteristics, effectively resolving differentiation challenges at constraint boundaries. Furthermore, a clustering multi-objective optimization genetic algorithm (CMOGA) is proposed, which employs non-dominated sorting and crowding distance computation to effectively maintain the Pareto front distribution of the solution set. By introducing a clustering mechanism, the algorithm ensures solution diversity while significantly reducing computational complexity. Comprehensive simulations and comparative analyses demonstrate the proposed method’s significant advantages in collision avoidance, robustness, and optimality. This research provides an effective maneuvering strategy for future space orbital games.

  • From Chemical Reaction to Orbital Assembly: Fast Cost Evaluation and Global Energy Minimization

    by Jingxian Wang, Chenyu Shan, Heng Zhou, Yuzhu Bai, Yong Zhao, Rong Chen, Quanjing Cheng

    The high computational complexity of target assignment and trajectory planning in on-orbit assembly tasks limits its development. This paper develops a cross-scale framework for target assignment and global energy optimization inspired by chemical reactions. Aiming at the target assignment problem, according to the theory of molecular graph energy, an innovative rapid cost evaluation method based on the configuration transfer matrix energy is proposed. The method directly quantifies the difference in energy consumption of the assignment strategies by calculating the traces of sparse matrices, avoiding the redundant trajectory optimization costing iterations in traditional methods. For the trajectory planning problem, this paper draws on the idea of manifold dimension reduction for the minimum energy path of a chemical reaction to map the cluster configuration to the manifold space. The high-dimensional trajectory optimization problem is transformed into an energy-optimal path solution on the manifold. The effectiveness of the target assignment cost evaluation method is proved by theoretical derivation and simulation calculation. Finally, the advantages of the manifold algorithm in terms of low energy consumption, low initial guessing dependence, and fast computation are verified by comparative analysis.

  • Open Access

    Formation Maneuver Control Based on the Augmented Laplacian Method

    by Xinzhe Zhou, Xuyang Wang, Xiaoming Duan, Yuzhu Bai, Jianping He

    This paper proposes a novel formation maneuver control method for both 2-D and 3-D space, which enables the formation to translate, scale, and rotate with arbitrary orientation. The core innovation is the novel design of weights in the proposed augmented Laplacian matrix. Instead of using scalars, we represent weights as matrices, which are designed based on a specified rotation axis and allow the formation to perform rotation in 3-D space. To further improve the flexibility and scalability of the formation, the rotational axis adjustment approach and dynamic agent reconfiguration method are developed, allowing formations to rotate around arbitrary axes in 3-D space and new agents to join the formation. Theoretical analysis is provided to show that the proposed approach preserves the original configuration of the formation. The proposed method maintains the advantages of the complex Laplacian-based method, including reduced neighbor requirements and no reliance on generic or convex nominal configurations, while achieving arbitrary orientation rotations via a more simplified implementation. Simulations in both 2-D and 3-D space validate the effectiveness of the proposed method.

  • Path Planning and Control for Multiagent Traversing Numerous Obstacles

    by Rong Chen, Jingxian Wang, Heng Zhou, Yuzhu Bai, Yong Zhao, Xiaoqian Chen

    To ensure that multiagent’s collision-free traversal in a complex 3-D space with numerous obstacles, this article proposes an online centralized path planning and distributed tracking control method. First, considering the position and overall size of the numerous obstacles in 3-D space, a threat model is established by using Gaussian mixture model, which assigns a different threat level to each location in the 3-D space, with closer obstacles posing a greater threat. Subsequently, five rules are defined based on the threat model, namely, target steering, minimal threat, threat threshold, direct connection, and mixed threat. A combination of these rules is used to design a shortest, collision-free, and fast path planning algorithm for the multiagent. To ensure holistic optimality, multiagent path planning is designed as a centralized computation. A distributed tracking control algorithm is then designed based on model predictive control. Finally, the proposed methods are verified by both numerical simulations and semiphysical simulations based on the air-float experimental system, proving the stable generation of collision-free traversal paths in numerous obstacles, and the accurate tracking control. The simulations show that the proposed methods have the characteristics of minimal path length, high-efficiency computation, robust security, general applicability, and small control error.


  • Open Access

    Rapid Sequence Generation for Active Debris Removal Mission Based on Attention Mechanism and Pointer Network

    by Zhijun Chen, Yuzhu Bai, Yong Zhao, Xiaoqian Chen

    This paper proposes a method based on the attention mechanism combined with a pointer network for rapidly generating sequences of Active debris removal (ADR) missions. Initially, the paper introduces the problem and modeling of ADR missions, establishing the multi-target planning problem as a time-dependent traveling salesman problem (TDTSP) considering the time dimension. Subsequently, during the spacecraft transfer process, it considers using continuous-thrust transfer, utilizing Pontryagin’s minimum principle to transform the fuel-optimal continuous-thrust transfer problem into a two-point boundary value problem (TPBVP), then rapidly generating a continuous-thrust dataset to train a surrogate model for evaluating continuous-thrust cost. Furthermore, PN-attention is proposed, which is derived from a pointer network (PN) model that incorporates attention computation, aimed at learning the sequence planning of ADR problems. In the validation and discussion section, PN-attention’s performance in solving ADR problems is compared with various heuristic algorithms, verifying its effectiveness, optimality, generalizability, and solution speed.