Yongjia Zhu
Yongjia Zhu has a total of 5 co-authors. Here are the top co-authors by number of shared publications:
- Xueli Wang
Qilu University of Technology (Shandong Academy of Sciences),Key Laboratory of Computing Power Network and Information Security, Ministry of Education, Shandong Computer Science Center,Jinan,China,250014
2 shared publications - Derui Ding
Department of Control Science and Engineering, University of Shanghai for Science and Technology, Shanghai, China
1 shared publication Tianzhen Wang
1 shared publication- Tianzhen Wang
Logistics Engineering College, Shanghai Maritime University, Shanghai 201306, China
1 shared publication - Yun Zhou
Key Laboratory of Control of Power Transmission and Conversion (Ministry of Education) Shanghai Jiao Tong University Shanghai China
1 shared publication
Latest Publications
- Open Access
- 15 June 2026
- Published by Elsevier in SSRN Electronic Journal
Composite Observer-based Control with Adaptive Dynamic Programming for USVs: The Pursuit–Evasion Games
This paper investigates the pursuit-evasion game (PEG) problem for two uncrewed surface vehicles (USVs) operating in complex marine environments with unknown external disturbances. For the addressed problem, an augmented PEG model is first formulated. Based on this model, a dual-neural-network-driven hybrid observer is proposed to achieve decoupled identification of inherent system nonlinearities and external ocean wave disturbances, enabling accurate state estimation under adverse conditions. Leveraging the estimated states and the optimal control derived from a zero-sum game formulation, a hybrid adaptive dynamic programming (ADP) control scheme is then developed. This scheme synergistically integrates the game algebraic Riccati equation (GARE) with a critic-only broad learning system (BLS), and a meticulously derived weight update law is provided to effectively execute pursuit-evasion maneuvers. Theoretically, using Lyapunov stability theory, it is rigorously proven that all estimation errors of the hybrid observer are uniformly ultimately bounded (UUB), and sufficient conditions are established to guarantee UUB stability of the overall closed-loop PEG system. The observer gain matrix is obtained by solving linear matrix inequalities (LMIs). Finally, simulation results based on the Cybership II model vessel validate the effectiveness of the proposed framework, demonstrating improved disturbance rejection and pursuit-evasion performance compared to traditional methods.
- 15 June 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessComposite Observer-based Control with Adaptive Dynamic Programming for USVs: The Pursuit–Evasion Games
This paper investigates the pursuit-evasion game (PEG) problem for two uncrewed surface vehicles (USVs) operating in complex marine environments with unknown external disturbances. For the addressed problem, an augmented PEG model is first formulated. Based on this model, a dual-neural-network-driven hybrid observer is proposed to achieve decoupled identification of inherent system nonlinearities and external ocean wave disturbances, enabling accurate state estimation under adverse conditions. Leveraging the estimated states and the optimal control derived from a zero-sum game formulation, a hybrid adaptive dynamic programming (ADP) control scheme is then developed. This scheme synergistically integrates the game algebraic Riccati equation (GARE) with a critic-only broad learning system (BLS), and a meticulously derived weight update law is provided to effectively execute pursuit-evasion maneuvers. Theoretically, using Lyapunov stability theory, it is rigorously proven that all estimation errors of the hybrid observer are uniformly ultimately bounded (UUB), and sufficient conditions are established to guarantee UUB stability of the overall closed-loop PEG system. The observer gain matrix is obtained by solving linear matrix inequalities (LMIs). Finally, simulation results based on the Cybership II model vessel validate the effectiveness of the proposed framework, demonstrating improved disturbance rejection and pursuit-evasion performance compared to traditional methods.
- 17 February 2026
- Published by Elsevier in Ocean Engineering
- Vol. 352
Joint state-disturbance observation with encoding-decoding mechanisms for uncrewed surface vehicles: A neural network-based approach
- 17 February 2026
- Published by Elsevier in Ocean Engineering
- Vol. 352
Joint state-disturbance observation with encoding-decoding mechanisms for uncrewed surface vehicles: A neural network-based approach