Sign in Subscribe
Algorithms

Optimal Consensus Problem for Discrete-Time Multi-Agent Systems Investigated

Strategic Insights

— 1 min read

Researchers at Beijing Wuzi University have made new findings in the field of information technology, shedding light on the optimal consensus problem for discrete-time multi-agent systems with input constraints. The study aims to investigate the optimal protocol that enables all followers to reach consensus with the leader while minimizing the performance index. The research, supported by the National Natural Science Foundation of China (NSFC), utilizes actor-critic neural networks and addresses the challenges of deriving analytical solutions for the Hamilton-Jacobi-Bellman equation.

Key Takeaways:

  • The researchers aim to investigate the optimal consensus problem for discrete-time multi-agent systems with input constraints using data-based reinforcement learning algorithms.
  • The study utilizes actor-critic neural networks to implement the algorithm, relying on system data rather than system models, to address the challenges of deriving analytical solutions for the Hamilton-Jacobi-Bellman equation.
  • The projection operator and Frank-Wolfe methods are introduced to address the input constraints, and two algorithms (projection reinforcement learning algorithm and projection-free reinforcement learning algorithm) are designed.
  • The convergence and stability of the algorithms are examined, and an comprehensive algorithm is introduced by integrating the benefits of the two algorithms.
  • The research concludes that the algorithm is effective, as demonstrated by numerical examples.
  • The study was peer-reviewed and published in the International Journal of Control, Automation and Systems.

Statistics:

  • The study was supported by the National Natural Science Foundation of China (NSFC).
  • The research was conducted by a team of researchers from Beijing Wuzi University, including Lipo Mo, Min Zuo, Ruoxun Ma, and Ke Guo.
  • The study was published in the International Journal of Control, Automation and Systems, Volume 23, Issue 5, Pages 1322-1336.
  • The research was completed in 2025 and has been validated through numerical examples.

Sources:

  • Data-based Optimal Consensus of Multi-agent Systems With Reinforcement Learning Algorithms. International Journal of Control, Automation and Systems, 2025;23(5):1322-1336.
  • International Journal of Control, Automation and Systems. Inst Control Robotics & Systems, Korean Inst Electrical Engineers, Suseo Hyundai-Ventureville 723, Bamgogae-Ro 1-Gil 10, Gangnam-Gu, Seoul, South Korea. Springer, www.springer.com.
  • National Natural Science Foundation of China (NSFC), for financial support.

Read more