Extremum Seeking Control for Energy Optimization of Vapor Compression Systems

In this paper, an energy optimizing extremum seeking controller is developed for vapor compression systems (VCS) that automatically discovers sets of inputs that minimizes the energy consumption while the machine is in operation. This controller optimizes an input-output map (from VCS inputs to electrical energy consumed) in realtime, and without relying on a model of the dynamics of a vapor compression system. A detailed algorithm and rules for tuning the controller gains will be described. Experiments are performed on an inverter-driven room air conditioner that demonstrate convergence of inputs to their optimal values, resulting in an improvement in COP of 10-20% for some operating points.