Robust Optimal Control for Train Operation of 2 Car AGT (Automated Guide-way Transit) in Automatic People Movers System

Abstract

The paper considers a train operation which minimizes energy consumption for leading the train from one station to the next station at the specified instant and stopping times. The minimization of the energy consumption is defined as an optimal trajectory used as a reference signal for the robust optimal control to regulate an AGT train in Automatic People Movers (APM) system. In this case, we consider the optimal control system where the performance index is of general form containing a final (terminal) cost function in addition to the integral cost function. Since the final time and the final state are fixed or specified, there are no extra boundary conditions to be used other than those given in the problem formulation. Full state feedback control law is selected to design Linear Quadratic Regulator (LQR) and PID controller. Here we also consider the parameter selection problems of the PID controller within the framework of robust control theory which provides an alternative method to optimize the setting of the PID controller. To handle uncertainty of the train parameter such as mass, robust design is applied to find a controller, for a given system, such that the closed loop system is robust. By considering the uncertain passenger arrival flow, the robust optimal control law is designed to suppress the effect of this uncertainty and the uncertain disturbances to the nominal mass of the train system. Simulations show the robustness of the proposed train control algorithms against disturbances occurring on a train parameter.