An On-Line Optimal Controller for a Commuter Train
| Autor: | Dennis Yazhemsky, Muzamil Rashid, Shahin Sirouspour |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
050210 logistics & transportation
Optimization problem Computer science Mechanical Engineering 05 social sciences Solver Propulsion Optimal control 7. Clean energy Computer Science Applications Vehicle dynamics Control theory Robustness (computer science) 0502 economics and business Automotive Engineering Convex optimization Efficient energy use |
| Zdroj: | IEEE Transactions on Intelligent Transportation Systems. 20:1112-1125 |
| ISSN: | 1558-0016 1524-9050 |
| DOI: | 10.1109/tits.2018.2846480 |
| Popis: | This paper proposes an on-board optimal controller that drives a train between two stations in an optimal time efficient, energy efficient, or mixed-objective manner, while adhering to a set of system-specific constraints. To this end, at each step along the track, the train control problem is formulated and solved as a constrained optimization problem over the remainder of the trip, while utilizing the latest train sensor data. The optimization problem is a convex second-order cone program. It incorporates knowledge of the track profile and relevant velocity and propulsion/braking constraints in the computation of the optimal propulsion/braking commands. It features an option to enforce a safety buffer between the train and another leading train on the track. The resulting convex optimization problem can be efficiently solved using a simple numerical solver, ensuring global optimality and robustness of the solution. Real-time performance and simulated closed-loop control results are presented, for a realistic vehicle and advanced trip model on desktop and embedded computer architectures. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|