Distributed Model Predictive Control for Multiple Trains with Coupled Constraints

Zixuan Zhang,Haifeng Song,Weiqi Bai,Hongwei Wang,Hairong Dong
DOI: https://doi.org/10.23919/CCC58697.2023.10240754
2023-01-01
Abstract:This paper presents a distributed model predictive control of multiple trains, aiming at overcoming coupled constraints and achieving cooperative operation while reducing the computational burden. According to the coupled constraints associated with maintaining safety distance between adjacent trains, trains are divided into sub-systems and distributed optimal control problem is then formulated. To solve the problem, an alternating direction method of multipliers (ADMM) is employed with a fully distributed scheme, decomposing the original optimal control problem into smaller sub-problems that can be computed in parallel and meet the real-time control requirements. The proposed method effectively guarantees the cooperative operation of trains under local and coupled safety constraints. Numerical simulations are provided to demonstrate the feasibility and effectiveness of the proposed train control model and methods.
What problem does this paper attempt to address?