Co-Optimizing Power-Transportation Networks With Circulating Loads and Particle-Like Stochastic Motion

Coupling power-transportation systems may enhance the resilience of power grids by engaging energy-carrying mobile entities such as electric vehicles (EVs), truck-mounted energy storage systems, and Data Centers (DCs), which can shift the computing loads among their network. In practice, the co-optimization problem for power-transportation systems can be overly complicated due to a great deal of uncertainty and many decision variables rooted in the EV population and mobile energy storage. Another challenge is the heterogeneity in terms of size and supporting capability due to various types of such mobile entities. This work aims to facilitate power-transportation co-optimization by proposing and formalizing the concept of Circulating Loads (CirLoads) to generalize these spatial-temporal dispatchable entities. With the new concept, the stochastic process of CirLoads’ movement is introduced using Brownian particles for the first time. Such novel particle motion-based modeling for EVs can reflect their stochastic behaviors over time without requiring exact data of EVs. The distributions of CirLoads are further aggregated with Gaussian Mixture Models to reduce the dimensions. Based on this aggregated model, a co-optimization framework is proposed to coordinate the bulk of EVs while respecting data privacy between transportation and power systems. Simulation results demonstrate the effectiveness of the proposed framework.