Continuous-Domain Real-Time Distributed ADMM Algorithm for Aggregator Scheduling and Voltage Stability in Distribution Network

Abstract

With an increase in the number of electric vehicles (EVs) and battery storage systems (BSSs), there are major challenges in the distribution grid to maintain a scaling control structure. Also, with the vehicle-to-grid (V2G) technology, EVs can now inject power into the grid for voltage regulation. This article proposes a new distributed real-time alternating direction method of multipliers (ADMM) technique to control EVs and BSSs for voltage regulation while maximizing their utility function. More specifically, a continuous-domain real-time optimization and control algorithm is developed in closed form, which exchanges relevant information among the neighboring nodes through the communication network and optimizes a combined convex objective of EVs and BSSs welfare and voltage regulation with power flow equations as constraints. Convergence analysis is provided using the Lyapunov direct approach, and simulation results are included to illustrate the effectiveness of the proposed scheme. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —This article is motivated by the problems that an electric distribution system is facing today due to the penetration of electric vehicles and battery storage systems (BSSs). The presence of a large number of electric vehicles (EVs) and BSSs is causing a substantial degradation of the quality and reliability of the power grid. Despite the adverse effects, EVs and BSSs can be controlled and used as a power source to benefit the grid. Also, this control of EVs and BSS cannot be done by a centralized body since this would be nonscalable and would require huge communication bandwidth. To tackle this ever scaling problem, this article develops a continuous-domain multiagent distributed algorithm to control the EVs and BSSs and utilize them to maintain the grid voltage within the normal operating range while also satisfying the consumers by maximizing their welfare. The algorithm was developed in the continuous domain since, in most of the other algorithms that are iterative and in discrete time, the accuracy of the optimal solution greatly depends on the sampling time, and thus, it is not robust to changes that are prevalent with distributed energy resources such as EVs and BSSs.