Distributed Holistic Framework for Smart City Infrastructures: Tale of Interdependent Electrified Transportation Network and Power Grid

M. Hadi Amini, Javad Mohammadi, Soummya KarIEEE Access 2019 · DOI

Abstract

Plug-in Electric Vehicles (PEVs) play a pivotal role in transportation electrification. The flexible nature of PEVs' charging demand can be utilized for reducing charging cost as well as optimizing the operating cost of power and transportation networks. Utilizing charging flexibility of geographically spread PEVs requires design and implementation of efficient optimization algorithms. In this paper, we introduce a holistic framework to model the interdependent nature of power systems and electrified transportation networks, enhance the operational performance of these systems as a network-of-networks, and explain the required information exchange via coupling agents (e.g., PEVs and charging stations). We develop a holistic framework that enables distributed coordination of interdependent networks through the IoT lens. To this end, we propose to use a fully distributed consensus+innovations approach. This iterative algorithm achieves a distributed solution of the decision making for each agent through local computations and limited communication with other neighboring agents. Fully distributed decision making enables scalability, plug-and-play capability, and increased information privacy.

Introduction: The Rise of e-Mobility

As the global shift toward electric mobility (e-Mobility) accelerates, the relationship between the power grid and transportation networks is undergoing a fundamental transformation. Electric Vehicles (EVs) are no longer merely passive loads on the electrical system; they have become mobile energy storage units with significant "spatiotemporal flexibility." This means they can strategically choose both when and where to draw power, creating a complex, two-way interdependence between two of our most critical infrastructures.

Global growth of EVs from 2013 to 2017, highlighting the increasing scale of the coordination challenge.

Managing this growth through traditional centralized control faces significant hurdles. Scalability becomes an issue when trying to manage millions of individual agents from a single hub. Furthermore, centralized systems often require users to sacrifice privacy by sharing exact locations and travel plans. Most importantly, these systems must now account for cross-network effects: traffic congestion dictates where people charge, and grid congestion can influence where people drive.

The HADI Framework: A Network-of-Networks

To address these challenges, this research introduces the HADI (Holistic Agent-based Distributed) framework. This approach moves away from viewing infrastructures in isolation, instead treating every component—from a single EV to a distribution transformer—as an "agent" within an Internet-of-Things (IoT) ecosystem.

Evolution of network modeling from independent silos to an interdependent "plain surface" through the IoT lens, with EVs acting as the primary coupling agents.

The framework identifies EVs and charging stations as "coupling agents" that glue the power and transportation layers together. By viewing the city through this "IoT lens," the framework leverages the mobility of the load to solve grid problems. For instance, if a specific transformer is overloaded, the system can encourage EVs to move to a different charging location or delay their session, rather than simply cutting power.

Distributed Optimization via Consensus+Innovations

The mathematical engine driving HADI is a Consensus+Innovations algorithm. This approach allows for global optimization through purely local interactions:

  • Consensus: Agents communicate with immediate neighbors to align on global system variables, such as the overall state of the grid.
  • Innovations: Agents incorporate their own private, local data (e.g., battery state-of-charge or local transformer thermal limits) to update their individual decisions.

This iterative process enables the system to converge on an optimal solution—such as the most cost-effective charging schedule that maintains grid stability—without requiring a "Big Brother" controller. This distributed nature ensures privacy, as EV owners only exchange limited information with immediate neighbors or their specific charging station.

Key Results and Performance

Validation of the HADI framework through simulations involving various PEV battery capacities (16kWh and 24kWh) demonstrated several key advantages:

  1. Accuracy: The distributed algorithm successfully reached the same optimal solutions as traditional centralized optimization methods.
  2. Real-Time Agility: In dynamic scenarios where vehicles suddenly departed a station, the system rescheduled the remaining agents in approximately 2 seconds, with individual mathematical iterations taking less than 2 milliseconds.
  3. Resilience: The study proved that by leveraging mobility patterns, the system could find feasible charging solutions even when local power grid limits were strictly reduced (e.g., from 25kW down to 10kW).
  4. Scalability: The "plug-and-play" capability allows new vehicles or stations to join the network seamlessly, as the local consensus simply adjusts to the new neighbor.

Conclusion

The HADI framework provides a scalable, privacy-preserving solution for the growing interdependence of smart city infrastructures. While the current model assumes a connected communication topology, future work will explore system performance in "dead zones" and investigate the competitive equilibrium between different commercial charging station operators across a city-wide scale.

Summary updated

BibTeX
@article{amini2019distributed,
  author = {M. Hadi Amini and Javad Mohammadi and Soummya Kar},
  title = {{Distributed Holistic Framework for Smart City Infrastructures: Tale of Interdependent Electrified Transportation Network and Power Grid}},
  journal = {IEEE Access},
  year = {2019},
  doi = {10.1109/access.2019.2950372},
}
Figure