Multi-Agent System Based Cooperative Control for Speed Convergence of Virtually Coupled Train Formation.

Area of Science:

• Control Systems Engineering
• Transportation Systems
• Applied Mathematics

Background:

• Maintaining consistent spacing in train formations is crucial for operational safety and efficiency.
• Existing control methods face challenges due to communication limitations (distance, bandwidth) and unknown external disturbances.
• Ensuring adherence to safety hard constraints for inter-train spacing is a significant operational requirement.

Purpose of the Study:

• To propose a distributed cooperative control algorithm for train formations that ensures speed convergence and maintains safe spacing.
• To address practical operational constraints, including limited communication range and bandwidth.
• To develop a robust control strategy that compensates for unknown external disturbances and model uncertainties.

Main Methods:

• A distributed train-formation speed-convergence cooperative-control algorithm based on barrier Lyapunov function (BLF) is proposed.
• A distributed observer is designed for each train to estimate unknown reference trajectories and disturbances using adjacent train states.
• Nonlinear adaptive control theory is integrated with BLF to handle model parameter uncertainties and enforce spacing hard constraints.

Main Results:

• The proposed algorithm enables all trains to track the reference trajectory accurately.
• It ensures that inter-train spacing remains within a predefined safe interval, satisfying hard constraints.
• The closed-loop system demonstrates asymptotic stability, validated through a practical case study on Guangzhou Metro Line 22.

Conclusions:

• The developed barrier Lyapunov function-based adaptive control method effectively manages train formations under practical constraints.
• The distributed observer successfully estimates and compensates for unknown factors, enhancing control robustness.
• The algorithm provides a reliable solution for safe and stable train spacing control, confirmed by real-world data.