The rapid development of smart grids has imposed a new requirement on residential appliances, i.e. being demand response-enabled (DR-enabled). One of the key features of DR-enabled appliances is to automatically respond to real-time electricity pricing (RTP) so as to effectively shift peak power demands from high-RTP to low-RTP periods and reduce electricity costs. Advanced DR control methods are essential to develop DR-enabled appliances. Residential air conditioners (ACs) are the major contributors to home electricity bills and electrical grids. Inverter AC becomes popular in today's homes due to the higher energy efficiency at part-load conditions. In this paper, we aim to apply the model predictive control (MPC) method to inverter AC to make it RTP-responsive. The MPC method can simultaneously consider multiple influential variables including weather condition, occupancy and RTP to achieve the optimization of energy consumption or electricity cost. Considering the computational efficiency for real-time online control, a simple-structured grey-box room thermal model is developed. We also develop a steady-state physical model of inverter AC and generate its performance maps for online applications. A TRNSYS-MATLAB co-simulation testbed is developed to test the performances of the MPC controller. Test results show that compared with PID control, the MPC-based DR controller helps to improve the thermal comfort at the beginning of occupancy, reduce peak power demands and total electricity costs.
展开▼
