Dezong Zhao
@glasgow.ac.uk
University of Glasgow
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Scopus Publications
Li Zhang, Dezong Zhao, Chee Peng Lim, Houshyar Asadi, Haoqian Huang, Yonghong Yu, and Rong Gao
Elsevier BV
Wenjing Zhao, Siyuan Gong, Dezong Zhao, Fenglin Liu, N.N. Sze, Mohammed Quddus, and Helai Huang
Elsevier BV
Hongbo Yin, Daxin Tian, Chunmian Lin, Xuting Duan, Jianshan Zhou, Dezong Zhao, and Dongpu Cao
Institute of Electrical and Electronics Engineers (IEEE)
Jincheng Hu, Yang Lin, Jihao Li, Zhuoran Hou, Liang Chu, Dezong Zhao, Quan Zhou, Jingjing Jiang, and Yuanjian Zhang
Elsevier BV
Sifan Wu, Daxin Tian, Xuting Duan, Jianshan Zhou, Dezong Zhao, and Dongpu Cao
Institute of Electrical and Electronics Engineers (IEEE)
Peiyu Zhang, Daxin Tian, Jianshan Zhou, Mai Chang, Xuting Duan, Dezong Zhao, Dongpu Cao, and Vicor C.M. Leung
Institute of Electrical and Electronics Engineers (IEEE)
Peiyu Zhang, Daxin Tian, Jianshan Zhou, Xuting Duan, Dezong Zhao, and Dongpu Cao
Institute of Electrical and Electronics Engineers (IEEE)
Jiawei Li, Daxin Tian, Jianshan Zhou, Xuting Duan, Zhengguo Sheng, Dezong Zhao, and Dongpu Cao
Institute of Electrical and Electronics Engineers (IEEE)
Chunmian Lin, Daxin Tian, Xuting Duan, Jianshan Zhou, Dezong Zhao, and Dongpu Cao
Institute of Electrical and Electronics Engineers (IEEE)
Harikirshnan Vijayakumar, Dezong Zhao, Jianglin Lan, Wenjing Zhao, Daxin Tian, Dachuan Li, Quan Zhou, and Kang Song
Institute of Electrical and Electronics Engineers (IEEE)
Dachuan Li, Bowen Liu, Zijian Huang, Qi Hao, Dezong Zhao, and Bin Tian
Institute of Electrical and Electronics Engineers (IEEE)
Chunmian Lin, Daxin Tian, Xuting Duan, Jianshan Zhou, Dezong Zhao, and Dongpu Cao
Institute of Electrical and Electronics Engineers (IEEE)
Recent advances in cross-modal 3D object detection rely heavily on anchor-based methods, and however, intractable anchor parameter tuning and computationally expensive postprocessing severely impede an embedded system application, such as autonomous driving. In this work, we develop an anchor-free architecture for efficient camera-light detection and ranging (LiDAR) 3D object detection. To highlight the effect of foreground information from different modalities, we propose a dynamic fusion module (DFM) to adaptively interact images with point features via learnable filters. In addition, the 3D distance intersection-over-union (3D-DIoU) loss is explicitly formulated as a supervision signal for 3D-oriented box regression and optimization. We integrate these components into an end-to-end multimodal 3D detector termed 3D-DFM. Comprehensive experimental results on the widely used KITTI dataset demonstrate the superiority and universality of 3D-DFM architecture, with competitive detection accuracy and real-time inference speed. To the best of our knowledge, this is the first work that incorporates an anchor-free pipeline with multimodal 3D object detection.
Peiyu Zhang, Jianshan Zhou, Daxin Tian, Xuting Duan, Kaige Qu, Dezong Zhao, Zhengguo Sheng, Pinlong Cai, and Victor C.M. Leung
ACM
Vehicle platooning has gained significant attention due to its potential to enhance road safety and efficiency. Leveraging stochastic optimization methods, this paper presents a distributed Stochastic Model Predictive Control (SMPC) controller tailored for vehicle platooning systems to improve their safety and robustness. Uniquely, our methodology describes the vehicle's dynamic state and establishes the error equation for the platoon system founded on a mass-spring structure structural concept, a departure from existing models. Using this, we formulate an SMPC platoon control framework resilient to stochastic disturbances, effectively integrating desired objective and probabilistic chance constraints. Given the probabilistic information of the random perturbations, an equivalent, computationally efficient framework for the SMPC is deduced under a fixed distribution. Comprehensive simulation experiments serve to validate the efficacy of our proposed SMPC platoon controller.
Longbo Cheng, Zixuan Xu, Jianshan Zhou, Daxin Tian, Xuting Duan, Kaige Qu, and Dezong Zhao
MDPI AG
Anti-jamming communication technology is one of the most critical technologies for establishing secure and reliable communication between unmanned aerial vehicles (UAVs) and ground units. The current research on anti-jamming technology focuses primarily on the power and spatial domains and does not target the issue of intelligent jammer attacks on communication channels. We propose a game-theoretical center frequency selection method for UAV-enabled air-to-ground (A2G) networks to address this challenge. Specifically, we model the central frequency selection problem as a Stackelberg game between the UAV and the jammer, where the UAV is the leader and the jammer is the follower. We develop a formal matrix structure for characterizing the payoff of the UAV and the jammer and theoretically prove that the mixed Nash equilibrium of such a bimatrix Stackelberg game is equivalent to the optimal solution of a linear programming model. Then, we propose an efficient game algorithm via linear programming. Building on this foundation, we champion an efficacious algorithm, underpinned by our novel linear programming solution paradigm, ensuring computational feasibility with polynomial time complexity. Simulation experiments show that our game-theoretical approach can achieve Nash equilibrium and outperform traditional schemes, including the Frequency-Hopping Spread Spectrum (FHSS) and the Random Selection (RS) schemes, in terms of higher payoff and better stability.
Jianglin Lan and Dezong Zhao
Institute of Electrical and Electronics Engineers (IEEE)
Gledson Melotti, Weihao Lu, Pedro Conde, Dezong Zhao, Alireza Asvadi, Nuno Gonçalves, and Cristiano Premebida
Institute of Electrical and Electronics Engineers (IEEE)
Object detection in autonomous driving applications implies that the detection and tracking of semantic objects are commonly native to urban driving environments, as pedestrians and vehicles. One of the major challenges in state-of-the-art deep-learning based object detection are false positives which occur with overconfident scores. This is highly undesirable in autonomous driving and other critical robotic-perception domains because of safety concerns. This paper proposes an approach to alleviate the problem of overconfident predictions by introducing a novel probabilistic layer to deep object detection networks in testing. The suggested approach avoids the traditional Sigmoid or Softmax prediction layer which often produces overconfident predictions. It is demonstrated that the proposed technique reduces overconfidence in the false positives without degrading the performance on the true positives. The approach is validated on the 2D-KITTI objection detection through the YOLOV4 and SECOND (Lidar-based detector). The proposed approach enables interpretable probabilistic predictions without the requirement of re-training the network and therefore is very practical.
Harikirshnan Vijayakumar, Dezong Zhao, Jianglin Lan, Wenjing Zhao, Daxin Tian, and Yuanjian Zhang
Elsevier BV
Yuanjian Zhang, Bingzhao Gao, Jingjing Jiang, Chengyuan Liu, Dezong Zhao, Quan Zhou, Zheng Chen, and Zhenzhen Lei
Elsevier BV
Xiaoyan Hu, Yu Gong, Dezong Zhao, and Wen Gu
Institute of Electrical and Electronics Engineers (IEEE)
This article investigates the local linear model tree (LOLIMOT), a typical neural fuzzy model, in the multiple-input–multiple-output model predictive control (MPC). In the conventional LOLIMOT, the structural parameters including centers and variances of its Gaussian kernels are set based on equally dividing the input data space. In this article, after the structural parameters are initially obtained from the input space partition, they are optimized by the gradient descent search, from which the space partitions are further adjusted. This makes it better for the model structure to fit the input data statistics, leading to improved modeling performance with a small model size. The MPC based on the proposed structurally optimized LOLIMOT is then implemented and verified with both numerical and diesel engine plants. Validation results show that the proposed MPC has significantly a better controlling performance than the MPC based on the conventional LOLIMOT, making it an attractive solution in practice.
Peiyu Zhang, Daxin Tian, Jianshan Zhou, Xuting Duan, Zhengguo Sheng, Dezong Zhao, and Dongpu Cao
Institute of Electrical and Electronics Engineers (IEEE)
—Vehicle platooning technology is essential in achiev- ing group consensus, on-road safety, and fuel-saving. Meanwhile, Vehicle-to-Infrastructure (V2I) communication significantly fa- cilitates the development of connected vehicles. However, the coupled effects of the longitudinal vehicle’s mobility, platoon control and V2I communication may result in a low reliable communication network between the platoon vehicle and the roadside unit, there is a tradeoff between the platoon control and communication reliability. In this paper, we investigate a bi- objective joint optimization problem where the first objective is to maximize the success probability of data transmission (communication reliability) and the second objective function is to minimize the traffic oscillation flow. The vehicle’s mobility state of the platoon vehicle affects the channel capacity and transmission performance. In this context, we deeply explore the relationship between control signals and resource scheduling and theoretically deduce a closed-form expression of the optimal communication reliability objective. Through this closed expression, we transform the bi-objective model into a single objective MPC model by using ϵ -constraint method. We design an efficient algorithm for solving the joint optimization model and prove the convergence. To verify the effectiveness of the proposed method, we finally evaluate the spacing error, speed error, and resource scheduling of platooning vehicles through simulation experiments in two experimental scenarios. The results show that the proposed control-communication co-design can improve the platoon control performance while satisfying the high reliability of V2I communications.
Wenjing Zhao, Siyuan Gong, Dezong Zhao, Fenglin Liu, N.N. Sze, and Helai Huang
Elsevier BV
Jianglin Lan, Dezong Zhao, and Daxin Tian
Institute of Electrical and Electronics Engineers (IEEE)
Jianglin Lan, Dezong Zhao, and Daxin Tian
Wiley
This article considers mixed platoons consisting of both human-driven vehicles (HVs) and automated vehicles (AVs). The uncertainties and randomness in human driving behaviors highly affect the platoon safety and stability. However, most existing control strategies are either for platoons of pure AVs, or for special formations of mixed platoons with known HV models. This article addresses the control of mixed platoons with more general formations and unknown HV models. An innovative data-driven policy learning strategy is proposed to design the controllers for AVs based on vehicle-to-vehicle (V2V) communications. The policy learning strategy is embedded with the constraints of control input, inter-vehicular distance error and V2V communication topology. The strategy establishes a safe and robustly stable mixed platoon using prescribed communication topologies. The design efficacy is verified through simulations of a mixed platoon with different communication topologies and leader velocity profiles.
Weihao Lu, Dezong Zhao, Cristiano Premebida, Li Zhang, Wenjing Zhao, and Daxin Tian
Institute of Electrical and Electronics Engineers (IEEE)
Zihao Sheng, Lin Liu, Shibei Xue, Dezong Zhao, Min Jiang, and Dewei Li
Institute of Electrical and Electronics Engineers (IEEE)
Lane change for automated vehicles (AVs) is an important but challenging task in complex dynamic traffic environments. Due to difficulties in guaranteeing safety as well as a high efficiency, AVs are inclined to choose relatively conservative strategies for lane change. To avoid the conservatism, this paper presents a cooperation-aware lane change method utilizing interactions between vehicles. We first propose an interactive trajectory prediction method to explore possible cooperations between an AV and the others. Further, an evaluation on safety, efficiency and comfort is designed to make a decision on lane change. Thereafter, we propose a motion planning algorithm based on model predictive control (MPC), which incorporates AV’s decision and surrounding vehicles’ interactive behaviors into constraints so as to avoid collisions during lane change. Quantitative testing results show that compared with the methods without an interactive prediction, our method enhances driving efficiencies of the AV and other vehicles by 14.8% and 2.6%, respectively, which indicates that a proper utilization of vehicle interactions can effectively reduce the conservatism of the AV and promote the cooperation between the AV and others.