Xiaojun Xing

Xiaojun Xing

Auto-generated profile, based on publicly available information

Xiaojun Xing

Northwestern Polytechnical University

Research Focus & Metrics

Academic Career Length

20Years (2007-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
16 April 2024 in IEEE Transactions on Vehicular Technology
Verified
Journal Article
Closed access
XXXiaojun Xing
Xiaojun Xing
ZZZhiwei Zhou
Zhiwei Zhou
YLYan Li
Yan Li
Bing XiaoBing Xiao
Bing Xiao
YXYilin Xun
Yilin Xun
Multi-unmanned aerial vehicle (multi-UAV) cooperative trajectory planning is an extremely challenging issue in UAV research field due to its NP-hard characteristic, collision avoiding constraints, close formation requirement, consensus convergence and high-dimensional action space etc. Especially, the difficulty of multi-UAV trajectory planning will boost comparatively when there are complex obstacles and narrow passages in unknown environments. Accordingly, a novel multi-UAV adaptive cooperative formation trajectory planning approach is proposed in this paper based on an improved deep reinforcement learning algorithm in unknown obstacle environments, which innovatively introduces long short-term memory (LSTM) recurrent neural network (RNN) into the environment perception end of multiagent twin delayed deep deterministic policy gradient (MATD3) network, and develops an improved potential field-based dense reward function to strengthen the policy learning efficiency and accelerates the convergence respectively. Moreover, a hierarchical deep reinforcement learning training mechanism, including adaptive formation layer, trajectory planning layer and action execution layer is implemented to explore an optimal trajectory planning policy. Additionally, an adaptive formation maintaining and transformation strategy is presented for UAV swarm to comply with the environment with narrow passages. Simulation results show that the proposed approach is better in policy learning efficiency, optimality of trajectory planning policy and adaptability to narrow passages than that using multi-agent deep deterministic policy gradient (MADDPG) and MATD3.

22 October 2025 in IEEE Transactions on Vehicular Technology
Verified
Journal Article
Closed access
XXXiaojun Xing
Xiaojun Xing
YMYuanqiang Ma
Yuanqiang Ma
YLYichen Lei
Yichen Lei
YLYan Li
Yan Li
Bing XiaoBing Xiao
Bing Xiao
Traditional trajectory planning algorithms for multi-UAVs face challenges such as difficulty in establishing cooperative mechanisms and poor adaptability to dynamic obstacle environments. To address these limitations, an enhanced reinforcement learning algorithm, based on the multi-agent deep deterministic policy gradient algorithm (MADDPG) and attention mechanism, is proposed for multi-UAV rendezvous trajectory planning in unknown complex environments. Firstly, the algorithm innovatively introduces an attention mechanism in deep learning into the centralized critic network of the MADDPG, enabling the model to dynamically adjust attention in complex environments and enhance learning efficiency; secondly, a dense reward function model based on guiding points is developed, combining attractive and repulsive forces, effectively addressing the issue of sparse rewards, accelerating the algorithm's convergence rate, and bettering policy learning efficiency; thirdly, an Ornstein-Uhlenbeck (OU) noise network is incorporated to well balance exploration and exploitation during the training process; finally, in the static obstacle environment, dynamic obstacle environment and extended composite scenarios, this algorithm was compared with MADDPG, MATD3, and IDDPG. The results show that the improved algorithm can effectively avoid collisions, successfully rendezvous at the target point, and achieve the minimum decision steps, the shortest trajectory length and the highest rendezvous success rate. Especially in scenarios with multiple dynamic obstacles, the improved algorithm can adjust the UAV flight path in real-time and successfully avoid all dynamic obstacles.

15 January 2025 in IEEE Transactions on Industrial Informatics
Verified
Journal Article
Closed access
XXXiaojun Xing
Xiaojun Xing
YLYiming Luo
Yiming Luo
BHBing Han
Bing Han
LQLinfeng Qin
Linfeng Qin
Bing XiaoBing Xiao
Bing Xiao
High-accuracy fault diagnosis is a crucial way to improve the reliability of electro-hydrostatic actuator (EHA) in transport airplane. Due to the limitation of aircraft structure, it is extremely difficult to obtain EHA fault data and to accurately assess the type of fault occurrence, so a process methodology is proposed for test signal excitation in an EHA simulation environment to obtain multisource fault data. Based on two EHA fault dataset, a multisequence fusion network (MSFN) is proposed for end-to-end fault diagnosis. MSFN has the advantages of light weight, high efficiency, and the ease of expansion, utilizes multiscale wide kernel convolutional neural network (CNN), deep dilated CNN, and long short-term memory network for parallel feature extraction, and the improved fusion channel and spatial attention mechanism for cross-fusion of different sequence features. Experimental results show that MSFN has high prediction accuracy and robustness under different intensity noise, achieving 98.56% average accuracy when SNR = 20, can effectively realize the rapid fault diagnosis of EHA under multiple complex working conditions.

1 April 2024 in IEEE Transactions on Aerospace and Electronic Systems
Verified
Journal Article
Closed access
XXXiaojun Xing
Xiaojun Xing
RWRui Wang
Rui Wang
QGQianchao Gong
Qianchao Gong
Bing XiaoBing Xiao
Bing Xiao
The accurate recovery of first-stage booster of rocket or spacecraft is one of the most significant research issues in the aerospace field, of which parafoil system is an effective low-cost reliable roadmap for the first-stage booster of rocket recovery. Nevertheless, as a lightweight and underactuated aircraft, the parafoil and first-stage booster combination (PFC) is extremely vulnerable to low-altitude terrain obstacles (like mountains) and wind turbulences (like longitudinal thermal turbulences), simultaneously undergoes recovery energy consumption limitation. In order to guarantee flight safety and upwind precise landing of the PFC at low altitude environments, a terminal trajectory planning method, avoiding low-altitude static terrain obstacles and longitudinal thermal turbulences dynamic obstacles, is proposed in this paper. The method generates initial PFC trajectory by Model Predictive Static Programming (MPSP) algorithm, then obstacles-avoiding terminal trajectory of PFC is planned using integrating slack variable and sliding mode control theory, of which firstly transforms the obstacle avoidance inequality constraints into equality constraints by introducing slack variables, secondly formulate appropriate sliding mode surface to manipulate the change of slack variables to always satisfy the inequality constraints of safe distance between the PFC and the obstacles safety boundary, thirdly plans the trajectory of safe flight and upwind landing based on MPSP algorithm. Furthermore, the trajectory optimization is done by differential dynamic programming (DDP) for least energy consumption. The numerical simulation results show that the planed trajectory always maintains safe distance between the PFC and the obstacles, meets the requirements of upwind landing,makes the control quantity more continuous and smoother. The hardware-in-the-loop experiments show that the planned terminal trajectory is continuous and smooth with excellent traceability and comparatively small tracking deviation.

27 August 2024 in IEEE Transactions on Aerospace and Electronic Systems
Verified
Journal Article
Closed access
XXXiaojun Xing
Xiaojun Xing
LQLinfeng Qin
Linfeng Qin
YLYiming Luo
Yiming Luo
LFLei Feng
Lei Feng
Bing XiaoBing Xiao
Bing Xiao
The parafoil system is an effective low-cost reliable methodology to fulfill the precise recovery and reuse of rocket booster. As a lightweight underactuated aircraft, the parafoil with booster combination (PBC) motion characteristics are strongly nonlinear and extremely vulnerable to unknown disturbances, such as model uncertainty and wind disturbance. So, a prescribed performance-based backstepping(PPB) attitude tracking controller of PBC with fixed-time disturbance observer(FDO) is proposed. Firstly, a 9-degree-of-freedom (DOF) multi-body flexible nonlinear model of the PBC is built based on kinematic, dynamic and computational fluid dynamics (CFD) analyses. Then, to ensure the attitude error converges to an arbitrarily small residual set, with steady-state error and settling time less than the prespecified values, the prescribed performance control is employed to improve traditional backstepping method which deals with the system nonlinearity effectively. Furthermore, to restrain the complex unknown complex disturbances and enhance the robustness of PPB controller, a FDO is introduced to accurately estimate the unknown external environmental and internal model uncertainties disturbances of the PBC. Finally, numerical simulation experiments and hardware-in-loop experiments of PBC attitude tracking control are conducted thoroughly, whose results demonstrate that the PPB controller with FDO proposed in this paper can satisfy the prescribed performance and robustness requirements and achieve precise recovery within the error requirements under unknow complex disturbances.

11 September 2024 in IEEE Transactions on Instrumentation and Measurement
Verified
Journal Article
Closed access
XXXiaojun Xing
Xiaojun Xing
RWRui Wang
Rui Wang
BHBing Han
Bing Han
CWCihang Wu
Cihang Wu
Bing XiaoBing Xiao
Bing Xiao
The growing significance of aerial refueling requires that receiver aircraft could perform autonomous aerial refueling tasks in flight. In this regard, precise docking is a key but challenging issue. To address this problem, a drogue trajectory prediction method based on transfer learning and attention mechanism is proposed in this article. The long short-term memory (LSTM) neural network is introduced as the base to learn temporal correlations between time series trajectory data of a drogue. To further boost the network performance, transfer learning strategy and the attention mechanism are involved in the model construction. Prior knowledge about physical models in similar domains can be passed to the network through transfer learning, and larger weights can be adaptively assigned to more important features. The effectiveness of the proposed method is verified through the comparisons with ARIMA, RNN, LSTM, and attention-based LSTM models, while effects of transfer learning and attention mechanism are visualized. When implementing this approach to perform predictive docking, a high success rate is achieved in the ground experiment, which shows great potentials for industrial applications.