Communication-Aware Scheduling of Heterogeneous Thermostatic Loads for Fast Flexibility Support

Authors:

Abstract
Power dynamics, temperature control, and response latency are crucial factors to take into account while using Thermostatically Controlled Loads (TCLs) for Demand Response(DR) or Frequency Response (FR) for sustainable power system. Comfort and efficiency are controlled by temperature, responsiveness is improved by reducing communication delays. This paper proposes a novel Normalized Weighted-Greedy (NWG) scheduling framework for the grid connected TCLs exhibiting diversity in power ratings, operating temperatures, and communication delays, reflecting heterogeneity of 2 million TCLs. The IEEE 69-bus distribution system is used to assess the feasibility of the suggested NWG algorithm while taking N-1 contingency scenarios into account. the proposed NWG algorithm is compared against $\epsilon$-greedy, batch reinforcement learning (RL), and Monte Carlo RL methods. The NWG algorithm mitigates the imposed power deficit and confirms its efficacy by fully satisfying the demand response requirement, as shown by simulation results. The proposed approach, experimentally validated on a real-time hardware platform, demonstrates high efficiency, scalability, and expedience, confirming the effectiveness of scheduling TCLs for practical Fast Load Curtailment(FLC) deployment across millions of devices.

This publication has 27 references indexed in Scilit: