A Cognitive Control Method for Cost Cost-Efficient Efficient CBTC Systems With Smart Grids
Abstract: Communication-based based train control (CBTC) systems use wireless local area networks for information transmission between trains and wayside equipment. Since inevitable packet delay and drop are introduced in train-wayside train communications, information uncertai uncertainties nties in trains' states will lead to unplanned traction/braking demands, as well as waste in electrical energy. Moreover, with the introduction of regenerative braking technology, power grids in CBTC systems are evolving to smart grids, and cost cost-aware powerr management should be employed to reduce the total financial cost of consumed electrical energy. In this paper, a cognitive control method for CBTC systems with smart grids is presented to enhance both train operation performance and cost efficiency. We formulate f a cognitive control system model for CBTC systems. The information gap in cognitive control is calculated to analyze how the train train-wayside wayside communications affect the operation of trains. The Q Q-learning learning algorithm is used in the proposed cognitive control ntrol method, and a joint objective function composed of the information gap and the total financial cost is a.pplied to generate optimal policy. The medium-access access control layer retry retry-limit limit adaption and traction strategy selection are adopted as cognitive actions. Extensive simulation results show that the cost efficiency and train operation performance of CBTC systems are substantially improved using our proposed cognitive control method.