Hongzong Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Hongzong Li has a total of 20 co-authors. Here are the top co-authors by number of shared publications:
- Jun Wang
Department of Physics, Southeast University, Nanjing 210096, China
8 shared publications - Luwei Liao
College of Automation Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing, China
2 shared publications - Quan Ouyang
College of Automation Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing 210016, China
2 shared publications - Zhong Yang
Nanjing University of Aeronautics and Astronautics
2 shared publications - Danguo Cheng
Nanjing University of Aeronautics and Astronautics
1 shared publication - Dapeng Oliver Wu
City University of Hong Kong
1 shared publication
Latest Publications
- 4 August 2026
- Published by Elsevier in Knowledge-Based Systems
- Vol. 351
HumanDiff: Leveraging body-part expert-assisted diffusion transformers for human image generation
- 4 August 2026
- Published by Elsevier in Knowledge-Based Systems
- Vol. 351
HumanDiff: Leveraging body-part expert-assisted diffusion transformers for human image generation
- 2 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Electronics
- Vol. PP (02780046) , 1-12
Decoupled Large-Pitch Maneuvers of a 2DOF Thrust-Vectoring Bicopter via Quaternion Finite-Time Control and Dynamic Control Allocation
Multirotors are evolving from passive observation platforms to active interaction tools. Therefore, the demand for multirotors omnidirectional maneuverability in complex, nonlevel environments has been increasing. Traditional under-actuated multirotors and single-degree-of-freedom thrust-vectoring bicopter suffer from inherent pitch-longitudinal coupling issues. To overcome these limitations, this article proposes a novel two-degree-of-freedom thrust-vectoring bicopter, abbreviated as 2DOF-TVB, where the design can realize the pitch-longitudinal decoupling of the airframe. A singularity-free quaternion-based dynamic model is established to accurately describe large-attitude maneuvers. To address actuator redundancy, inherent center-oestimates and compensates CoG uncertaintyf-gravity uncertainty, and disturbance issues, a hierarchical control architecture was developed, and finite-time stability was analyzed by Lyapunov analysis. This architecture incorporates a real-time constrained dynamic control allocation to ensure feasible actuation under limits, together with the nonsingular fast terminal sliding mode control using the quaternion-based higher order sliding mode observer that mitigates external disturbances and CoG offset uncertainties. Experimental results validated its stable $\boldsymbol{\pm 90^{\circ}}$ pitch hovering and vertical wall adhesion capabilities, demonstrating its exceptional maneuverability and robustness.
- 2 April 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Industrial Electronics
- Vol. PP (02780046) , 1-12
Decoupled Large-Pitch Maneuvers of a 2DOF Thrust-Vectoring Bicopter via Quaternion Finite-Time Control and Dynamic Control Allocation
Multirotors are evolving from passive observation platforms to active interaction tools. Therefore, the demand for multirotors omnidirectional maneuverability in complex, nonlevel environments has been increasing. Traditional under-actuated multirotors and single-degree-of-freedom thrust-vectoring bicopter suffer from inherent pitch-longitudinal coupling issues. To overcome these limitations, this article proposes a novel two-degree-of-freedom thrust-vectoring bicopter, abbreviated as 2DOF-TVB, where the design can realize the pitch-longitudinal decoupling of the airframe. A singularity-free quaternion-based dynamic model is established to accurately describe large-attitude maneuvers. To address actuator redundancy, inherent center-oestimates and compensates CoG uncertaintyf-gravity uncertainty, and disturbance issues, a hierarchical control architecture was developed, and finite-time stability was analyzed by Lyapunov analysis. This architecture incorporates a real-time constrained dynamic control allocation to ensure feasible actuation under limits, together with the nonsingular fast terminal sliding mode control using the quaternion-based higher order sliding mode observer that mitigates external disturbances and CoG offset uncertainties. Experimental results validated its stable $\boldsymbol{\pm 90^{\circ}}$ pitch hovering and vertical wall adhesion capabilities, demonstrating its exceptional maneuverability and robustness.
- 16 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Systems, Man, and Cybernetics: Systems
- Vol. 55 (12) , 9456-9467
Machine-Cell and Part-Family Formation via Neurodynamics-Driven Constrained Binary Matrix Factorization
The formation of part families and their corresponding machine cells is a critical phase in the design of a cellular manufacturing system. This article presents a constrained binary matrix factorization approach to machine-cell and part-family formation. A constrained binary matrix factorization problem is formulated for machine-cell and part-family formation to minimize the number of exceptional elements (EEs). The constrained binary matrix factorization is further reformulated to a quadratic unconstrained binary optimization (QUBO) problem by reducing the quartic objective function of the binary matrix factorization problem to a quadratic one and penalizing the violation of constraints. A neurodynamics-driven algorithm is proposed to solve the reformulated quadratic problem by leveraging several Boltzmann machines (BMs) for searching solutions and a particle swarm optimization rule to reinitialize the neuronal states upon their local convergence to escape from local solutions and move toward global optimal ones. Experimental results on 18 benchmark datasets are presented to showcase the superior performance of the proposed approach in terms of four criteria.
- 16 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Systems, Man, and Cybernetics: Systems
- Vol. 55 (12) , 9456-9467
Machine-Cell and Part-Family Formation via Neurodynamics-Driven Constrained Binary Matrix Factorization
The formation of part families and their corresponding machine cells is a critical phase in the design of a cellular manufacturing system. This article presents a constrained binary matrix factorization approach to machine-cell and part-family formation. A constrained binary matrix factorization problem is formulated for machine-cell and part-family formation to minimize the number of exceptional elements (EEs). The constrained binary matrix factorization is further reformulated to a quadratic unconstrained binary optimization (QUBO) problem by reducing the quartic objective function of the binary matrix factorization problem to a quadratic one and penalizing the violation of constraints. A neurodynamics-driven algorithm is proposed to solve the reformulated quadratic problem by leveraging several Boltzmann machines (BMs) for searching solutions and a particle swarm optimization rule to reinitialize the neuronal states upon their local convergence to escape from local solutions and move toward global optimal ones. Experimental results on 18 benchmark datasets are presented to showcase the superior performance of the proposed approach in terms of four criteria.
- Open Access
- 21 January 2025
- Published by Qeios Ltd
Review of: "An Approximated QUBO Formulation for Solving Practical SAT Problems"
This paper introduces a straightforward formalization method to convert SAT instances into QUBO instances. The approach is both interesting and meaningful, and the experimental analysis is thorough. Below are my suggestions, which may help improve th...
- 21 January 2025
- Published by Qeios Ltd
Open AccessReview of: "An Approximated QUBO Formulation for Solving Practical SAT Problems"
This paper introduces a straightforward formalization method to convert SAT instances into QUBO instances. The approach is both interesting and meaningful, and the experimental analysis is thorough. Below are my suggestions, which may help improve th...
- 3 June 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Emerging Topics in Computational Intelligence
- Vol. 9 (1) , 228-239
A Collaborative Neurodynamic Algorithm for Quadratic Unconstrained Binary Optimization
Quadratic unconstrained binary optimization (QUBO) is a typical combinatorial optimization problem with widespread applications in science, engineering, and business. As QUBO problems are usually NP-hard, conventional QUBO algorithms are very time-consuming for solving large-scale QUBO problems. In this paper, we present a collaborative neurodynamic optimization algorithm for QUBO. In the proposed algorithm, multiple discrete Hopfield networks, Boltzmann machines, or their variants are employed for scattered searches, and a particle swarm optimization rule is used to re-initialize neuronal states repeatedly toward global optima. With extensive experimental results on four classic combinatorial optimization problems, we demonstrate the efficacy and potency of the algorithm against several prevailing exact and meta-heuristic algorithms.
- 3 June 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Emerging Topics in Computational Intelligence
- Vol. 9 (1) , 228-239
A Collaborative Neurodynamic Algorithm for Quadratic Unconstrained Binary Optimization
Quadratic unconstrained binary optimization (QUBO) is a typical combinatorial optimization problem with widespread applications in science, engineering, and business. As QUBO problems are usually NP-hard, conventional QUBO algorithms are very time-consuming for solving large-scale QUBO problems. In this paper, we present a collaborative neurodynamic optimization algorithm for QUBO. In the proposed algorithm, multiple discrete Hopfield networks, Boltzmann machines, or their variants are employed for scattered searches, and a particle swarm optimization rule is used to re-initialize neuronal states repeatedly toward global optima. With extensive experimental results on four classic combinatorial optimization problems, we demonstrate the efficacy and potency of the algorithm against several prevailing exact and meta-heuristic algorithms.
- 30 April 2024
- Published by Elsevier in Neural Networks
- Vol. 176, 106348
Binary matrix factorization via collaborative neurodynamic optimization
- 30 April 2024
- Published by Elsevier in Neural Networks
- Vol. 176, 106348
Binary matrix factorization via collaborative neurodynamic optimization
- 27 September 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Fuzzy Systems
- Vol. 32 (3) , 1181-1194
From Soft Clustering to Hard Clustering: A Collaborative Annealing Fuzzy $c$-Means Algorithm
The fuzzy c-means clustering algorithm is the most widely used soft clustering algorithm. In contrast to hard clustering, the cluster membership of data generated using the fuzzy c-means algorithm is ambiguous. Similar to hard clustering algorithms, the clustering results of the fuzzy cmeans clustering algorithm are also sub-optimal with varied performance depending on initial solutions. In this paper, a collaborative annealing fuzzy c-means algorithm is presented. To address the issue of ambiguity, the proposed algorithm leverages an annealing procedure to phase out the fuzzy cluster membership degree toward a crispy one by reducing the exponent gradually according to a cooling schedule. To address the issue of sub-optimality, the proposed algorithm employs multiple fuzzy c-means modules to generate alternative clusters based on membership srepeatedly re-initialized using a meta-heuristic rule. Experimental results on eight benchmark datasets are elaborated to demonstrate the superiority of the proposed algorithm to thirteen prevailing hard and soft algorithms in terms of internal and external cluster validity indices.
- 27 September 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Fuzzy Systems
- Vol. 32 (3) , 1181-1194
From Soft Clustering to Hard Clustering: A Collaborative Annealing Fuzzy $c$-Means Algorithm
The fuzzy c-means clustering algorithm is the most widely used soft clustering algorithm. In contrast to hard clustering, the cluster membership of data generated using the fuzzy c-means algorithm is ambiguous. Similar to hard clustering algorithms, the clustering results of the fuzzy cmeans clustering algorithm are also sub-optimal with varied performance depending on initial solutions. In this paper, a collaborative annealing fuzzy c-means algorithm is presented. To address the issue of ambiguity, the proposed algorithm leverages an annealing procedure to phase out the fuzzy cluster membership degree toward a crispy one by reducing the exponent gradually according to a cooling schedule. To address the issue of sub-optimality, the proposed algorithm employs multiple fuzzy c-means modules to generate alternative clusters based on membership srepeatedly re-initialized using a meta-heuristic rule. Experimental results on eight benchmark datasets are elaborated to demonstrate the superiority of the proposed algorithm to thirteen prevailing hard and soft algorithms in terms of internal and external cluster validity indices.