Design of Optimal Coasting Speed for.pdf

2090 IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, VOL. 45, NO. 6, NOVEMBER/DECEMBER 2009 Design of Optimal Coasting Speed for MRT Systems Using ANN Models Hui-Jen Chuang, Member, IEEE , Chao-Shun Chen, Senior Member, IEEE, Chia-Hung Lin, Member, IEEE , Ching-Ho Hsieh, and Chin-Yin Ho Abstract—An artificial neural network (ANN) has been pro- been operated for more than ten years with daily ridership posed in this paper to determine the optimal coasting speed of over 1.3 million. Kaohsiung Mass Rapid Transit (KMRT) has train operation for the Kaohsiung Mass Rapid Transit (KMRT) been constructed since 1999 and has been commissioned in system to achieve the cost minimization of energy consumption and passenger traveling time. A train performance simulation early 2008. To achieve the optimization of system operation, (TPS) is applied to solve the energy consumption and the travel- KMRT has to design the proper coasting speed of train sets by ing time required to complete the journey between stations with considering the traction energy consumption and the passenger various riderships to create the data set for ANN training. The traveling time. For a typical MRT system, the energy cost will ANN model for the determination of the optimal coasting speed contribute more than 20% of the total operation cost, and 75% is then derived by performing the ANN training. To demonstrate the effectiveness of the proposed ANN model, the annual ridership of the total energy consumption will be consumed by train sets forecast of the KMRT system over the project concession period for propulsion purposes. Although th