Robust Adaptive Control for Virtually Coupled Trains

Virtual coupling (vc) technology can highly increase the capacity of existing infrastructure. It includes formation coupling, cruise operation and formation decoupling. However, there are few research on vc cruise control considering actual train operation scenarios. In this paper, a leader-following vc trains cruise control model is focused on. Trains exchange their state information by bidirectional leader communication topology. Considering existing train operating scenarios, the method takes speed constraint, collision avoidance and input saturation into consideration. RBF neural network is used to deal with the nonlinearity caused by the unknown basic resistance of train and external resistance of train. By designing the appropriate sliding manifold, the stability of virtually coupled train set (VCTS) can be guaranteed. Finally, some simulations demonstrate the effectiveness of the control method. The final results show that all the state information can be limited to the required range by designing the appropriate parameters.