This work considers the problem of actively managing the power consumption of a large number of thermostically controlled loads (TCLs), namely a TCL population, and a case-study of household refrigerators. Control is performed using a new randomized actuation that consists of switching units on and off at given rates, while at the same time respecting the nominal constraints on each individual unit. Both the free and the controlled behavior of individual TCLs can be aggregated, making it possible to handle a TCL population as if it were a single system. The aggregation method uses the distribution of the TCLs individual states across the population. The distribution approach has two main advantages. It scales excellently since the computational requirements do not increase with the number of units, and it allows data from individual units to be used anonymously, which solves privacy concerns relevant for consumer adoption.

Abstract-Demand response is an important Smart Grid concept that aims at facilitating the integration of volatile energy resources into the electricity grid. This paper considers a residential demand response scenario and specifically looks into the problem of managing a large number thermostatbased appliances with on/off operation. The objective is to reduce the consumption peak of a group of loads composed of both flexible and inflexible units. The power flexible units are the thermostat-based appliances. We discuss a centralized, model predictive approach and a distributed structure with a randomized dispatch strategy.

Demand response is an important Smart Grid concept that aims at facilitating the integration of volatile energy resources into the electricity grid. This paper considers the problem of managing large populations of thermostat-based devices with on/off operation. The objective is to enable demand response capabilities within the intrinsic flexibility of the population. A temperature distribution model based on Fokker-Planck partial differential equations is used to capture the behavior of the population. To ensure probability conservation and high accuracy of the numerical solution, Finite Volume Method is used to spatially discretize these equations. Next, a broadcast strategy with two switching-fraction signals is proposed for actuating the population. This is applied in an open-loop scenario for tracking a power reference by running an optimization with a multilinear objective.

An increase in number of distributed generation (DG) units in power system allows the possibility of setting-up and operating micro-grids. In addition to a number of technical advantages, micro-grid operation can also reduce running costs by optimally scheduling the generation and/or storage systems under its administration. This paper presents an optimized scheduling of a micro-grid battery storage system that takes into account the next-day forecasted load and generation profiles and spot electricity prices. Simulation results show that the battery system can be scheduled close to optimal even with forecast errors. ICLOCS software has been used for the numerical implementation.

A smart grid is a complex system consisting of a wide range of electric grid components, entities controlling power distribution, generation and consumption, and a communication network supporting data exchange. This paper focuses on the influence of imperfect network conditions on smart grid controllers, and how this can be counteracted by utilizing Quality of Service (QoS) information from the communication network. Such an interface between grid controller and network QoS is particularly relevant for smart grid scenarios that use third party communication network infrastructure, where modification of networking and lower layer protocols are impossible. This paper defines a middleware solution for adaptation of smart grid control, which uses network QoS information and interacts with the smart grid controller to increase dependability. In order to verify the methodology, an example scenario of a low voltage grid controller is simulated under imperfect network conditions.