Decentralized multi-objective optimization of distribution network operation through flexibility of photovoltaic and split-HVAC systems

Abstract

Voltage regulation in distribution networks has become increasingly difficult as photovoltaic systems spread across the grid and summer cooling loads increase. This paper addresses the problem through a Decentralized Multi-Objective Optimization Problem (DMOOP) strategy that takes advantage of the flexibility available in split-system air conditioners (HVAC) and PV inverters. The thermal behavior of a split-system air conditioner is captured by a state-space model The comfort of occupants is managed through a Weibull-distributed set-point temperature, which defines the acceptable temperature band and the range over which the HVAC power consumption can be adjusted. PV systems provide the reactive power flexibility that the distribution system operator can rely on to improve grid operation, while compensating customers for this service. The overall optimization problem couples technical objectives, voltage deviations, with economic ones, and is decomposed into local sub-problems using the Alternating Direction Method of Multipliers (ADMM). The framework is tested on an IEEE MV benchmark network and compared against two solutions: a voltage optimization problem (VOP) using only PV reactive flexibility and without optimization.