Disaggregated Bundle Methods for Distributed Market Clearing in Power Networks

TL;DR

This paper introduces a fast distributed market clearing method for power networks that efficiently incorporates demand response from residential aggregators using a disaggregated bundle approach, improving convergence speed.

Contribution

It develops a novel disaggregated bundle method for dual optimization in distributed market clearing, enhancing convergence speed and reducing communication overhead.

Findings

Faster convergence compared to traditional dual update algorithms

Effective integration of demand response from residential aggregators

Numerical results confirm improved efficiency and reduced communication

Abstract

A fast distributed approach is developed for the market clearing with large-scale demand response in electric power networks. In addition to conventional supply bids, demand offers from aggregators serving large numbers of residential smart appliances with different energy constraints are incorporated. Leveraging the Lagrangian relaxation based dual decomposition, the resulting optimization problem is decomposed into separate subproblems, and then solved in a distributed fashion by the market operator and each aggregator aided by the end-user smart meters. A disaggregated bundle method is adapted for solving the dual problem with a separable structure. Compared with the conventional dual update algorithms, the proposed approach exhibits faster convergence speed, which results in reduced communication overhead. Numerical results corroborate the effectiveness of the novel approach.