A Traction Substation State Estimator for Integrating Smart Loads in Transportation Grids Without the Need for Additional Sensors

Abstract

Public electric transport grids tend to be oversized and underutilized. Therefore, they can become sustainable and multi-functional backbones to the city AC grid by integrating smart grid elements into their infrastructures. However, integrating smart grid loads and renewables requires a large array of wirelessly communicating sensors across the traction substations, the smart grid components, and each vehicle of the transport fleet. This can be both costly and technically complex. This paper proposes an analytical state estimator that can predict vehicle traffic count and spare power capacity under a traction substation without the use of any additional sensors. The estimator uses existing, locally available measurements at any power node on the traction section to inform the decision-making of the power management scheme at that node. Validating the results with up to 100000 stochastic test simulations of a verified traction grid model, up to 76% of the monitored conditions were detected, with no false positives, and without the need for additional sensors and wireless communication.