Zhang Hui

Dr. Zhang Hui

Inner Mongolia University
orcid

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

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

Representative Publications

View All Publications
13 June 2023 in Ocean Engineering
Verified
Journal Article
Closed access
SGSiyang Guo
Siyang Guo
ZHZhang Hui
Zhang Hui
YGYaming Guo
Yaming Guo

7 February 2025 in Drones
Verified
Journal Article
Open access
PDF file
LHLe Han
Le Han
ZHZhang Hui
Zhang Hui
NANan An
Nan An
In the field of unmanned aerial vehicle (UAV) path planning, the conventional deep Q-network (DQN) algorithm encounters the issue of action space discretization, which results in the generation of unsmooth and inefficient planned paths. To address this issue, we introduce the particle swarm optimization (PSO) algorithm into DQN to convert the discrete action space into a continuous one. This method divides the agent’s surrounding space into discrete and continuous action spaces. The PSO algorithm performs a global search in the continuous space to obtain a continuous candidate solution, while DQN learns a policy in the discrete space to obtain a discrete candidate solution. Then, the two candidate solutions are combined using a weighted vector method to determine a direction that balances global search and policy learning. Additionally, we introduce a novel feature matrix as the state space for DQN, providing more accurate environmental and positional representations. Furthermore, we incorporate a mechanism into the base prioritized experience replay (PER) and N-step updates, which combines the current temporal difference error (TD-error) with historical priorities and includes a policy entropy penalty term, thereby enhancing DQN’s ability to learn long-term dependencies. The performance of the PSO-DQN model is further improved through an enhanced ε-greedy policy and learning rate decay strategy. Simulation results and experiments using the Flightmare simulator demonstrate that the proposed method generates smoother and more efficient paths for drones, exhibiting strong robustness in complex environments.

8 January 2025 in Applied Sciences
Verified
Journal Article
Open access
PDF file
ZZZhikun Zhang
Zhikun Zhang
ZHZhang Hui
Zhang Hui
This paper investigates the attitude control problem of unmanned aerial vehicles (UAVs), especially in the presence of uncertainties and external disturbances. To address this challenge, a fractional-order reaching law sliding mode with active disturbance rejection controller (FOSM-ADRC) is proposed. The controller combines a fractional-order calculus operator and active disturbance rejection controller (ADRC) techniques to enhance the dynamic performance and robustness of the system. Through the inner and outer loop design, the jitter of the sliding mode controller (SMC) is effectively suppressed, and fast response and strong anti-jamming ability are achieved, which, in turn, improves the control accuracy. Firstly, the dynamic model of the UAV is established, and its nonlinear dynamic characteristics are analyzed in detail. On this basis, a fractional-order reaching law sliding mode controller (FO-SMC) is designed as the outer loop to achieve fast response. ADRC is employed in the inner loop to compensate for the internal and external disturbances of the system. The results show that the FOSM-ADRC can effectively suppress the jitter phenomenon and maintain good control performance.

7 August 2020 in Electronics
Verified
Journal Article
Open access
PDF file
YLYang Liu
Yang Liu
LLLei Liu
Lei Liu
JLJiacheng Liang
Jiacheng Liang
JCJin Chai
Jin Chai
Xuemei LeiXuemei Lei
Xuemei Lei
ZHZhang Hui
Zhang Hui
Long Range (LoRa) has become one of the most promising physical layer technologies for the Internet of Things (IoT) ecosystem. Although it manifests low-power consumption and long-distance communication, LoRa encounters a large number of collisions in the IoT environment, which severely affects the system’s throughput and delay performance. In this paper, a code division carrier sense multiple access (CD/CSMA) protocol that resolves the traditional channel collision problem and implements multi-channel transmission is proposed for the LoRa medium access control (MAC) layer. To reduce data transmission delay and maximize the throughput of the system, the adaptive p-persistent CSMA protocol divides the channel load into four states and dynamically adjusts the data transmission probability. Then, to reduce channel collisions significantly, the code division multiple access (CDMA) protocol is performed on different channel states. Moreover, the combination of the proposed adaptive p-persistent CSMA protocol and the CDMA successfully reduces the number of data retransmissions and makes LoRa more stable. The simulation results demonstrate that the proposed adaptive p-persistent CD/CSMA protocol can achieve near-optimal and occasionally even better performance than some conventional MAC protocols, especially in a heavy load channel.

6 November 2021 in Remote Sensing Letters
Verified
Journal Article
Closed access
LYLei Yang
Lei Yang
ZHZhang Hui
Zhang Hui
MZMingzhu Zhou
Mingzhu Zhou
High-frequency surface wave radar (HFSWR) has been widely used for ship detection to improve human marine activities. However, the azimuthal resolution is a major challenge for ship detection in HFSWR. Synthetic aperture technology can improve azimuthal resolution, which is different from increasing the size of the antenna array directly in practice. A new method to improve HFSWR azimuthal resolution for ships moving in formation is proposed. The proposed method extends inverse synthetic aperture radar (ISAR) cross-range imaging method to HFSWR, allowing ships in formation to be distinguished from each other in the cross-range domain. In the framework of HFSWR cross-range imaging, the motion of the ship formation is regarded as a rotating turntable by motion compensation. Based on the turntable configuration, the relationship between the cross-range domain and Doppler frequency can be built precisely. Then, the cross-range profile is obtained from the echo data by Fourier transform, and the cross-range spacing between ships can be estimated by the cross-range scaling. The simulated results demonstrate the effectiveness of the proposed method. The present work provides a new idea to improve the azimuthal resolution of HFSWR.

21 November 2024 in IEEE Transactions on Mobile Computing
Verified
Journal Article
Closed access
SGSiyang Guo
Siyang Guo
YGYaming Guo
Yaming Guo
ZHZhang Hui
Zhang Hui
JWJunbo Wang
Junbo Wang
The increasingly popular federated learning still faces the practical challenge of non-independent and identically distributed data. Most efforts to address this issue focus on limiting local updates or enhancing model aggregation. However, these methods either restrict the learning capacity of local models or overlook the negative knowledge transfer caused by local objective divergences. In contrast, we observe that the global update can be re-expressed as a weighted sum of the gradients of samples from different classes. Therefore, we hypothesize that the competition among local updates may arise from the conflict between the gradients of samples belonging to different classes. Inspired by this insight, we introduce the novel perspective of orthogonal class gradients, aimed at eliminating interference between updates from different classes without the aforementioned drawbacks. To this end, this paper presents FedOCF, which implements orthogonal class gradient constraints by encouraging orthogonality among features of different classes. Specifically, FedOCF maintains a generator to learn features that are orthogonal for different classes and utilizes it to regularize features learned in local learning. Theoretically, we also demonstrate that FedOCF can improve generalization performance through feature conditional distribution alignment during local learning. Extensive experiments validate the excellent performance of FedOCF in various heterogeneous scenarios.