作者: Shuai Liu , Hans Hellendoorn , Bart De Schutter
DOI: 10.1109/TITS.2016.2573306
关键词:
摘要: The main aim of this paper is to use multi-class macroscopic traffic flow and emission models for Model Predictive Control (MPC) networks. In particular, we compare extended versions METANET, FASTLANE, VT-macro, VERSIT+. addition, end-point penalties based on these are also included in the objective function MPC account behavior system beyond prediction horizon. A simulation experiment implemented evaluate models. results show that approaches METANET (multi-class VT-macro or VERSIT+) can improve control performance total time spent emissions with respect non-control case, they more capable dealing queue length constraints than FASTLANE. Including further a small sacrifice computational efficiency but not