ABSTRACT
This paper analyses the performance of motorway control strategies subject to real-time flow measurement and modeling uncertainties. The control strategies are derived and tested on the cell transmission model with which global optimal solutions can be derived through solving linear programs. In particular, we present an adaptive control strategy which incorporates prevailing variations in traffic flow through a rolling horizon optimization framework. This adaptive strategy is compared with a min–max robust control formulation on a Monte Carlo stochastic simulation test bed. The robust control delivers the best performance in terms of minimizing delay variability due to its underlying conservativeness, while it comes at the expense of overall delay reduction. In contrast, the adaptive controller outperforms the robust controller in terms of delay reduction. Nevertheless, the benefit gained from the adaptive control diminishes as the motorway system gets saturated with traffic. It is also found that the adaptive controller is not effective in improving travel time reliability, at least under recurrent conditions. The findings reveal the limitation of adaptive control and provide insight to control design and infrastructure planning concerning installation of an advanced traffic control system.
Acknowledgements
The authors would like to thank the anonymous referees for the constructive comments. The authors would like to acknowledge UK Highways Agency for providing the MIDAS traffic data, and IBM Academic Initiative for providing the software and license for using the CPLEX Optimization Studio.
Disclosure statement
No potential conflict of interest was reported by the authors.
ORCID
Andy H. F. Chow http://orcid.org/0000-0002-2877-357X