Mitigating errors of predicted delays of a train at neighbouring stops

Abstract

Delay prediction is significant for the well-organised operation of trains. Meanwhile, the accurate prediction is interfered by comprehensive factors in practice, hence there are always quite a few predicted errors due to cascading effect in operation. This study newly proposes a method based on state-space equations to mitigate errors of predicted delays of a train at neighbouring stops. A linear quadratic regulator for reducing errors is developed from modern control theory. Moreover, necessary derivation and proof processes are made to support the method. An artificial fish swarm algorithm is designed to determine necessary parameters of the regulator, and the corresponding simulation is achieved by the SIMULINK architecture. The new method is able to be an effective reference for theoretical innovation and practical application in the operation of railway systems.