Thermostatically controlled loads have great potential to make a significant contribution to improving energy efficiency in the building sector, which is responsible for 40% of greenhouse gas emissions in the EU. This, in addition to the environmental damage, represents a huge expense in the electricity bill. Therefore, it is very important to train engineers on how to design energy management systems for TCLs. With this goal in mind, it would be very useful to have a simulation-based educational tool (SBET) to understand thermostatically controlled loads, their characteristics and possibilities in energy efficiency. In addition, it would be very useful if this tool could be introduced in engineering curricula to help students to be better trained and enter the labor market with more opportunities. Based on the shortcomings detected, this work develops a SBET specifically designed to teach about TCLs (SBET-TCLs), both about their intrinsic characteristics and their better management. To verify the developed SBET-TCL, it has been tested in a real scenario: A survey was carried out among the students of the subject 'Alternative Energy Sources' of the degrees of Industrial Engineering. The results show that the use of SBET-TCL has very positive effects on the learning process.

Demand response flexible loads can provide fast regulation and ancillary services as reserve capacity in power systems. This paper proposes a joint optimization dispatch control strategy for source-load system with stochastic renewable power injection and flexible thermostatically controlled loads (TCLs) and plug-in electric vehicles (PEVs). Specifically, the optimization model is characterized by a chance constraint look-ahead programming to maximal the social welfare of both units and load agents. By solving the chance constraint optimization with sample average approximation (SAA) method, the optimal power scheduling for units and TCL/PEV agents can be obtained. Secondly, two demand response control algorithms for TCLs and PEVs are proposed respectively based on the aggregate control models of the load agents. The TCLs are controlled by its temperature setpoints and PEVs are controlled by its charging power such that the DR control objective can be fulfilled. The effectiveness of the proposed dispatch and control algorithm has been demonstrated by the simulation studies on a modified IEEE 39 bus system with a wind farm, a photovoltaic power station, two TCL agents and two PEV agents.

In this paper, we study the performance of various deep reinforcement learning algorithms to enhance the energy management system of a microgrid. We propose a novel microgrid model that consists of a wind turbine generator, an energy storage system, a set of thermostatically controlled loads, a set of price-responsive loads, and a connection to the main grid. The proposed energy management system is designed to coordinate among the different flexible sources by defining the priority resources, direct demand control signals, and electricity prices. Seven deep reinforcement learning algorithms were implemented and are empirically compared in this paper. The numerical results show that the deep reinforcement learning algorithms differ widely in their ability to converge to optimal policies. By adding an experience replay and a semi-deterministic training phase to the well-known asynchronous advantage actor-critic algorithm, we achieved the highest model performance as well as convergence to near-optimal policies.