A Field Study on the Application of Extremum-Seeking Control for Energy Optimization of a VRF System

This paper presents the development and empirical validation of an Extremum Seeking Control (ESC) algorithm applied to a Variable Refrigerant Flow (VRF) system in an operational building.Conventional VRF control strategies often rely on fixed setpoints for refrigerant evaporating temperature, leading to suboptimal energy efficiency, especially under part-load conditions.To address this, a model-free ESC algorithm was implemented to dynamically adjust the evaporating temperature in real-time, minimizing the total power consumption of the outdoor unit.A field test was conducted to verify the algorithm's performance.The results demonstrated that the ESC successfully optimized the system's operation, significantly improving the Coefficient of Performance (COP) across all load ranges-more than doubling it at low loads.This led to a substantial reduction in energy consumption by up to 60% in mid-load ranges.Furthermore, the analysis of indoor temperature data confirmed that these energy efficiency gains were achieved while maintaining stable and comfortable indoor thermal conditions.This study validates the practical effectiveness of ESC as a robust solution for enhancing the energy performance of VRF systems in real-world building applications.