Real time electrical power estimation for the energy management of automatic metro lines

This paper intends to present a methodology to maximise reuse of regenerative braking energy in automatic metro lines for both offline and real time energy management. It first describes optimisation techniques for scheduling energy efficient timetables, while considering a no-fluctuation operating mode, as it corresponds to the most dominant operating case. Impact of headway and dwell time management on regenerative braking recovery are especially examined with a multi-criteria fitness function. Then, iterative solving techniques are introduced to precisely quantify energy transfers between trains. A neural estimator of trains power consumption is also proposed to meet real time requirements. Simulation results based on experiments conducted on Torino metro line are exposed to evaluate the performance of this estimator.