Reinforcement-Learning-Based Robust Controller Design for Continuous-Time Uncertain Nonlinear Systems Subject to Input Constraints

By Derong Liu, Xiong Yang, Ding Wang, Qinglai Wei. IEEE Transactions on Cybernetics, 2015

405 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computational Theory and Mathematics, Control and Systems Engineering

View Paper: https://www.semanticscholar.org/paper/b280e864f7ef3f610fe5ec982527a270c60a7925

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Shaocheng Tong
2. Derong Liu
3. Yong-ming Li
4. Haibo He
5. Qinglai Wei
6. Ding Wang
7. Hanquan Lin
8. C. D. de Visser

In Computational Theory and Mathematics

1. Shaocheng Tong
2. Derong Liu
3. Yong-ming Li
4. Qinglai Wei
5. Haibo He
6. Ding Wang
7. Hanquan Lin
8. Kangkang Sun

In Control and Systems Engineering

1. Shaocheng Tong
2. Yong-ming Li
3. Kangkang Sun
4. Derong Liu
5. Ding Wang
6. Haibo He
7. Qinglai Wei
8. Hanquan Lin

Citations in the corpus

• Safe Exploration Algorithms for Reinforcement Learning Controllers (2018, background): “In recent years, multiple successful applications [10]–[13] utilize RL as a framework for learning, in particular model-free RL [14]–[16].”
• Observer-Based Adaptive Fuzzy Fault-Tolerant Optimal Control for SISO Nonlinear Systems (2019, result): “learning methods have developed in [48]–[50].”
• Value Iteration Adaptive Dynamic Programming for Optimal Control of Discrete-Time Nonlinear Systems (2016, methodology): “Among these approximate methods, adaptive dynamic programming (ADP), proposed by Werbos [18], [19], as an effective data-based approach to solve optimal control problems forward-in-time, has gained much attention from researchers [20]–[29].”
• Adaptive Critic Nonlinear Robust Control: A Survey (2017, methodology): “This method was extended to deal with the robust stabilization of matched nonlinear systems with unknown dynamics [120] and with constrained inputs [121].”