Thermostatically Controlled Loads Research Articles

A new hybrid approach for cold load restoration using deep learning

After an extended outage in a power system, restoration of feeders with high penetration levels of thermostatically controlled loads (TCLs) will be a challenging problem. By restoring these feeders, the demand will increase several times the normal conditions which is called the cold load pick-up (CLPU) phenomenon. This can lead to overload of equipment such as power transformers in transmission substations. In order to increase the reliability of network load supply, in transmission substations, usually several transformers are used which can be operated in parallel by closing circuit breakers on the secondary side called bus section (BS). When the restoration process is long and the loads are cold, more loads can be restored by closing the BSs and paralleling the transformers. On the other hand, it is not possible to close all the BSs simultaneously because paralleling all the transformers at the same time increases the short circuit level on the secondary side. Furthermore, in power systems, some loads are voltage-dependent and by reducing the voltage, their demand can be decreased and therefore the restoration process is expedited. Due to the increase in demand caused by CLPU, additional stress is applied to the power system, which can lead to operational constraints violation and even long-term voltage instability. Therefore, at the time of restoration of each feeder, the settings of voltage control devices are updated to avoid the violation of operational constraints and the occurrence of wider events. In this paper, in order to reduce cold load restoration time, closing the bus sections and voltage reduction are used simultaneously. Also, short circuit and operational constraints are considered. The cold load restoration problem is formulated as a mixed integer nonlinear programming and is solved using the pattern search algorithm. Since there are many decision variables and solving the problem is time-consuming, a combination of deep learning networks, pattern search algorithm and parallel computing is used to accelerate the computations.

Decentralized control strategy of thermostatically controlled loads considering the energy efficiency ratio and measurement error correction

The elimination of traditional thermal generators and the increase of renewable energy source access reduced the inertia of the power system, and thermostatically controlled loads have the potential to provide ancillary services to the power system. This paper introduces a decentralized control strategy for thermostatically controlled loads (TCLs), which focuses on assisting large-scale TCLs in providing frequency support to power systems while considering energy efficiency ratio and measurement error correction. Initially, the nonlinear relationship between heat flow and TCLs’ power is modeled using a polynomial relationship, and the coefficients are determined through a data-driven method to enhance accuracy in TCLs modeling and assess their regulation potential precisely. Subsequently, a decentralized control strategy tailored for power system frequency regulation is developed. To account for measurement errors in TCLs controllers, a measurement error back-correction method is proposed. Simulation examples demonstrate the effectiveness of the proposed control strategy in achieving accurate modeling and control of TCLs and reducing power system frequency fluctuations.

Encouraging thermostatically controlled loads to provide frequency regulation for smart grid: A comfort-level trading mechanism

This paper presents a comfort-level trading mechanism for thermostatically controlled loads (TCLs), aiming to incentivize customers to provide more frequency regulation service for smart grid. This mechanism firstly divides aggregators composed of TCLs into the leader layer and the follower layer according to comfort adjustment level produced by the fuzzy neural network controller (FNC). The aggregator at leader layer has a low regulation level, while aggregators at follower layer are high. Then we establish a Stackelberg game between leader layer and follower layer, which can optimize the customer utility. Meanwhile, we build a noncooperative game model among aggregators at follower layer, which can maximize the customer utility while supplying regulation capacity for leader layer. Finally, simulation results validate the effectiveness of FNC and the feasibility of trading mechanism. The FNC can improve tracking accuracy to 2.05%, reduce absolute tracking error to 208.4 kW, and decrease average switching times to 9 per hour. The trading mechanism can make the temperature setpoints of aggregators within their comfort range and improve the customer satisfaction.

Two-stage decision-dependent demand response driven by TCLs for distribution system resilience enhancement

Service restoration is a critical function of the self-healing distribution system to cope with extreme events. Cold load pickup (CLPU) caused by thermostatically controlled loads (TCLs) requires extra generation power to retore loads, slowing down the distribution system restoration process. Meanwhile, increasing renewable energy brings the supply-demand imbalance because of the power variability and may cause voltage or current violations. Accordingly, a two-stage decision-dependent demand response (DR) methodology is proposed to address the two barriers to distribution system resilience enhancement. Specifically, two load control strategies are developed in the first-stage DR to regulate TCLs for CLPU mitigation, which reduces the additional power requirement when re-energized. In the second-stage DR, the restored TCLs that can be controlled to provide flexibility are characterized as virtual energy storage (VES) for managing renewable energy variability. In this way, the required power generation is reduced, and the power fluctuation problem caused by renewable energy can be well addressed. It helps to speed up the load restoration and keep the distribution system operating securely. Moreover, the load profiles with CLPU mitigation control, VES parameters and VES availability time, depend on the re-energization time. In other words, load restoration decisions determine the two-stage DR. This is the most important finding of this paper and explicitly formulated in the load restoration model for accurate load characterization. The load restoration model is described by a mixed integer second-order cone programming (MISOCP) problem and solved using MATLAB 2021 with Gurobi solver. Case studies show that the proposed methodology can significantly mitigate CLPU and restore more loads. The decision-dependent behaviors of the two-stage DR are also validated.

Aggregate Modeling of Thermostatically Controlled Loads for Microgrid Energy Management Systems

Second-to-second renewable power fluctuations can severely hinder the frequency regulation performance of modern isolated microgrids, as these typically have a low inertia and significant renewable energy integration. In this context, the present paper studies the coordinated control of Thermostatically Controlled Loads (TCLs) for managing short-term power imbalances, and their integration in microgrid operations through the use of aggregate TCL models. In particular, two computationally efficient and accurate aggregate TCL models are developed: a virtual battery model representing the aggregate flexibility of TCLs considering solar irradiance heat gains and wall/floor heat transfers, and a frequency transient model representing the aggregate dynamics of a TCL collection considering communication delays and the presence of model uncertainty and time-variability. The proposed aggregate TCL models are then used to design a practical Energy Management System (EMS) integrating TCL flexibility, and study the impact of TCL integration on microgrid operation and frequency control. Computational experiments using detailed frequency transient and thermal dynamic models are presented, demonstrating the accuracy of the proposed aggregate TCL models, as well as the economic and reliability benefits resulting from using these aggregate models to integrate TCLs in microgrid operations.

Robust Scheduling of Thermostatically Controlled Loads With Statistically Feasible Guarantees

Flexible resources are increasingly significant for the reliable operation of power grids due to the high penetration of renewable energy. Thermostatically controlled loads (TCLs) are one of the common flexible resources, whose control has been extensively studied. Yet, much can be improved. We investigate the scheduling of TCLs facing uncertain temperatures and dynamic prices. Classical approaches often employ the chance-constrained program or robust optimization to handle such uncertainties. However, these approaches either require specific distribution knowledge or yield too conservative solutions. The distribution knowledge can be rather challenging to obtain especially when the uncertainties from different sources are correlated. To this end, we adopt the notion of statistical feasibility and propose a robust sample-based scheduling scheme for TCLs. Such a sample-based scheme relieves the reliance on the distribution knowledge and is able to characterize the coupling effects of the two uncertainty sources. Besides, through integrating the real-time domain knowledge into uncertainty set reconstruction, we relax the solutions’ conservation by exploring the consumers’ tolerance to the room temperature. Numerical studies highlight the remarkable performance of our proposed scheme. Specifically, our approach is able to simultaneously effectively reduce the electricity bills of consumers and satisfy the consumers’ tolerance to the room temperature.

Artificial-Neural-Network-Based Autonomous Demand Response Controller for Thermostatically Controlled Loads

Thermostatically controlled loads (TCLs) possess an inherent potential for demand-side management (DSM). Space heating (SH) demands offer added flexibility due to their heat-storing capability and fast response times that may be unnoticed, but it has significant value for DSM. In this article, a communication-free demand response (DR) controller has been presented that harnesses the heat-storing capacity of residential buildings by utilizing TCLs. The proposed controller addresses customer voltage violations, comfort constraints, and energy savings. The DR controller attempts to control the TCLs by relying on voltage and voltage-to-load sensitivity. By deploying an artificial-neural-network-based model, the proposed control scheme ensures that the SH loads connected at spatially distributed nodes in the power system participate fairly to mitigate the residential grid voltage violations. Simulation results verify the proposed controller's performance to revamp the system voltage profiles while maintaining the desired comfort of end users. The controller is successfully tested with distributed generation integration, system reconfiguration, severe ambient temperature, and noise in the measured signals.

Customer-Centered Pricing Strategy Based on Privacy-Preserving Load Disaggregation

Demand response (DR) is a demand reduction or shift of electricity use by customers to make electricity systems flexible and reliable, which is beneficial under increasing shares of intermittent renewable energy. For residential loads, thermostatically controlled loads (TCLs) are considered as major DR resources. In a price-based DR program, an aggregation agent, such as a retailer, formulates price signals to stimulate the customers to change electricity usage patterns. The conventional DR management methods fully rely on mathematical models to describe the customer’s price responsiveness. However, it is difficult to fully master the customers’ detailed demand elasticities, cost functions, and utility functions in practice. Hence, in this paper, we proposed a data-driven non-intrusive load monitoring (NILM) approach to study the customers’ power consumption behaviors and usage characteristics. Based on NILM, the DR potential of the TCLs can be properly estimated, which assists the retailer in formulating a proper pricing strategy. To realize privacy protection, a privacy-preserving NILM algorithm is proposed. The proposed methodologies are verified in case studies. It is concluded that the proposed NILM algorithm not only reaches a better prediction performance than state-of-art works but also can protect privacy by slightly sacrificing accuracy. The DR pricing strategy with NILM integrated brings more profit and lower risks to the retailer, whose results are close to the fully model-based method with strong assumptions. Furthermore, a NILM algorithm with higher performance can help the retailer earn more benefits and help the grids better realize DR requirements.

Modelling and Simulation of Thermostatically Controlled Loads for Power System Regulation in Building-to-Grid Integration Framework

Since last few decades the inclusion of renewable energies in the grid power system is rapidly increasing with the increase in technologies to minimize the use of conventional fossil fuels for minimizing the emission of greenhouse gases. But the uncertain behaviour of the renewable energy sources lead to a challenge of maintaining the stability and reliability of the grid system and to optimize the supply and demand balance. It is identified that more than 40% energy consumption is made by the consumers of buildings. To minimize the power fluctuation, caused by inclusion of renewable systems, the thermostatically controlled loads (TCLs) like ACs, refrigerators, heaters of buildings can be considered as flexible loads for regulation purpose as TCLs share the major part of the total power consumption and also TCLs can be adjusted without compromising the comfort level. In this paper, a practical linear network model of the multi layered construction materials of building is developed with 3R2C electrical circuit model and corresponding state-space equations are formulated, and also a 2ndorder TCL model is proposed as an RC network equivalent where main focus is on the works, which develop specific control strategies by modelling the building energy systems.

Hierarchical control strategy of thermostatically controlled load considering multiple factors

In this article, a distributed hierarchical control strategy of thermostatically controlled loads (TCLs) is proposed, which can effectively utilize TCLs to provide power dispatching service in a microgrid composed of buildings. Firstly, the distributed consensus protocol of TCLs is studied, and the state space of a TCL is established. The optimal control performance is achieved by selecting coupling strength and control gain. Then, the dispatching power of each building is reasonably distributed through pinning control, and consensus coefficient and proportional coefficient are determined according to the heterogeneity of building parameters. Finally, the case studies were conducted to validate the proposed novel distributed hierarchical control scheme, which can provide effective power dispatching services and realize the fair distribution of the dispatching requirements among all TCLs. The proposed strategy also takes into account the multiple factors of each TCL such as heterogeneous parameters, communication burden and user's actual conditions, which improves the feasibility and expansibility of the dispatching process.

Hierarchical Control for Voltage Unbalance Mitigation Considering Load Management in Stand-Alone Microgrid

Unevenly distributed loads at different phases may lead to stability issues and incur high losses in a low inertial microgrid with a rise in Voltage Unbalance (VU). Thus, a suitable control mechanism is mandatory for stable microgrid operation and mitigating VU under peak and off-peak loading conditions. This paper presents a hierarchical control mechanism that includes a primary Droop-based Positive-Negative Sequence Compensation (PNSC) approach and a secondary controlling technique implemented using Demand Side Management (DSM). Besides mitigating VU, DSM also performs effective energy management by manipulating Thermostatically Controlled Loads (TCLs) and Electric Vehicle Charging stations (EVCs), with specific emphasis on the Thermal Comfort (TC) of customers. The primary controller’s stability is also examined to test and validate its potential to mitigate VU during off-peak loading. Furthermore, a comprehensive analysis of the proposed hierarchical controller is performed with several dynamics taken into account. Moreover, the control technique is designed to mitigate generation and demand mismatch while adhering to IEC-61000-3-13. The effectiveness of the proposed control framework is tested using a modified IEEE-13 bus system in a MATLAB/ Simulink and Digsilent PowerFactory environment. Further the Hardware in Loop (HIL) is done in OPAL-RT to validate the real time performance of the system.

District cooling system control for providing regulation services based on safe reinforcement learning with barrier functions

Thermostatically controlled loads (TCLs) in buildings are ideal resources to provide regulation services for power systems. As large-scale and centralized TCLs with high efficiency and large regulation capacity, district cooling systems (DCSs) have attracted great research attention for minimizing energy costs, but little on providing regulation services. However, controlling a DCS to provide high-quality regulation services is challenging due to its complex thermal dynamic model and uncertainties from regulation signals and cooling demands. To fill this research gap, we propose a novel safe deep reinforcement learning (DRL) control method for a DCS to provide regulation services. The objective is to adjust the DCS’s power consumption to follow real-time regulation signals subject to buildings’ temperature comfort constraints. The proposed method is model-free and adaptive to uncertainties from regulation signals and cooling demands. Furthermore, the barrier function is combined with traditional DRL to construct a safe DRL controller, which can not only avoid unsafe explorations during training (this may result in catastrophic control results) but also improve training efficiency. We conducted case studies based on a realistic DCS to evaluate the performance of the proposed control method compared to traditional methods, and the results demonstrate the increased effectiveness and superiority of the proposed control method.

Market-based coordination of price-responsive demand using Dantzig-Wolfe decomposition method

The increasing penetration of Distributed Generation (DG) and Demand Responsive (DR) loads in power systems has necessitated the development of novel approaches to address the coordination problem of Price Responsive Devices (PRD). These PRDs are treated as self-interested players aiming to optimize their consumption patterns based on prevailing prices. In this paper, we propose a new algorithm based on the Dantzig-Wolfe (DW) Decomposition method, which tackles the coordination problem of self-interested PRDs in a distributed manner. Leveraging the distributed nature of the DW approach, we model the self-interested algorithms of PRDs as sub-problems within the DW framework. The coordinator, or grid operator, responsible for collecting the energy consumption information (energy bids) of PRDs, solves the master problem of the DW and determines the price signal accordingly.The proposed algorithm exhibits fast convergence as the sub-problems within DW, which could involve a large number of PRDs (potentially millions), can be solved simultaneously. Additionally, based on the DW theory, if the PRDs' subproblems are convex, reaching the optimal point (equivalent to Nash Equilibrium) is guaranteed within a limited number of iterations. To evaluate the proposed model, we conducted a simulation involving 200 participant households, each equipped with two types of loads: Electric Vehicles (EVs) as examples of interruptible loads, and Electric Water Heaters (EWHs) as examples of Thermostatically Controlled Loads (TCLs). The results demonstrate that when the algorithm converges to the optimal point, both the generation cost and user payment (based on the marginal cost of generation) decrease. Furthermore, there is a significant reduction in the Peak to Average Ratio (PAR) of the aggregate load

Three-stage day-ahead scheduling strategy for regional thermostatically controlled load aggregators

Thermostatically controlled loads (TCLs) are regarded as having potential to participate in power grid regulation. This paper proposes a scheduling strategy with three-stage optimization for regional aggregators jointly participating in day-ahead scheduling to support demand response. The first stage is on the profit of aggregators and peak load of the grid. The line loss and voltage deviation of regulation are considered to ensure stable operation of the power grid at the second stage, which guarantees the fairness of the regulation and the comfort of users. A single temperature adjustment strategy is used to control TCLs to maximize the response potential in the third stage. Finally, digital simulation based on the IEEE 33-bus distribution network system proves that the proposed three-stage scheduling strategy can keep the voltage deviation within ± 5% in different situations. In addition, the Gini coefficient of distribution increases by 20% and the predicted percentage of dissatisfied is 48% lower than those without distribution.

Decomposition of Convex High Dimensional Aggregative Stochastic Control Problems

We consider the framework of convex high dimensional stochastic control problems, in which the controls are aggregated in the cost function. As first contribution, we introduce a modified problem, whose optimal control is under some reasonable assumptions an $\varepsilon$-optimal solution of the original problem. As second contribution, we present a decentralized algorithm whose convergence to the solution of the modified problem is established. Finally, we study the application of the developed tools in an engineering context, studying a coordination problem for large populations of domestic thermostatically controlled loads (TCLs)

A unified framework for coordination of thermostatically controlled loads

A collection of thermostatically controlled loads (TCLs) – such as air conditioners – can vary their power consumption within limits to help the balancing authority of a power grid maintain demand supply balance. Doing so requires loads to coordinate their on/off decisions so that the aggregate power consumption tracks a grid-supplied reference. At the same time, each consumer’s quality of service (QoS) must be maintained. While there is a large body of work on TCL coordination, they do not provide guarantees on the reference tracking performance or QoS maintenance, and they do not provide a means to compute a suitable reference signal for power demand of a collection of TCLs. In this work we provide a framework that addresses these two weaknesses. The framework enables coordination of an arbitrary number of TCLs that: (i) is computationally efficient, (ii) is implementable at the TCLs with local feedback and low communication, and (iii) enables reference tracking by the collection while ensuring that temperature and cycling constraints are satisfied at every TCL at all times. The framework is based on a Markov model obtained by discretizing a pair of Fokker–Planck equations derived in earlier work by Malhame and Chong (1985). We then use this model to design randomized policies for TCLs. The balancing authority broadcasts the same policy to all TCLs, and each TCL implements this policy which requires only local measurement to make on/off decisions. Simulation results are provided to support these claims. Matlab implementation is made publicly available.

Statistical mechanics of thermostatically controlled multizone buildings.

We study the collective phenomena and constraints associated with the aggregation of individual cooling units from a statistical mechanics perspective. These units are modeled as thermostatically controlled loads (TCLs) and represent zones in a large commercial or residential building. Their energy input is centralized and controlled by a collective unit-the air handling unit (AHU)-delivering cool air to all TCLs, thereby coupling them together. Aiming to identify representative qualitative features of the AHU-to-TCL coupling, we build a simple but realistic model and analyze it in two distinct regimes: the constant supply temperature (CST) and the constant power input (CPI) regimes. In both cases, we center our analysis on the relaxation dynamics of individual TCL temperatures to a statistical steady state. We observe that while the dynamics are relatively fast in the CST regime, resulting in all TCLs evolving around the control set point, the CPI regime reveals the emergence of a bimodal probability distribution and two, possibly strongly separated, timescales. We observe that the two modes in the CPI regime are associated with all TCLs being in the same low or high airflow states, with an occasional collective transition between the modes akin to Kramer's phenomenon in statistical physics. To the best of our knowledge, this phenomenon has been overlooked in building energy systems despite its direct operational implications. It highlights a trade-off between occupational comfort-related to zonal temperature variations-and energy consumption.

An Improved Deep Reinforcement Learning Method for Dispatch Optimization Strategy of Modern Power Systems

As a promising information theory, reinforcement learning has gained much attention. This paper researches a wind-storage cooperative decision-making strategy based on dueling double deep Q-network (D3QN). Firstly, a new wind-storage cooperative model is proposed. Besides wind farms, energy storage systems, and external power grids, demand response loads are also considered, including residential price response loads and thermostatically controlled loads (TCLs). Then, a novel wind-storage cooperative decision-making mechanism is proposed, which combines the direct control of TCLs with the indirect control of residential price response loads. In addition, a kind of deep reinforcement learning algorithm called D3QN is utilized to solve the wind-storage cooperative decision-making problem. Finally, the numerical results verify the effectiveness of D3QN for optimizing the decision-making strategy of a wind-storage cooperation system.

Thermostatically Controlled Loads in the Power System Under Cyberattacks

More and more thermostatic loads in power systems are transformed into thermostatically controlled loads (TCLs) through the heterogeneous Internet of Things (IoT) devices. The proportion of TCLs exposed to the Internet has also further increased. Considering TCLs’ response characteristics and the network security vulnerability of their associated IoT devices, it is possible to form a new threat to power systems, manipulation of demand (MAD) attacks. In this article, we reveal a potential attack scenario of this threat in the electricity market and propose a highly concealable and destructive cross-domain MAD attack strategy based on compromised TCLs, causing a tremendous economic loss for the power retailer (PR) and end-use consumers (EUCs). Finally, the effectiveness of the proposed MAD attack strategy is verified by simulation cases, highlighting the endanger of large-scale compromised TCLs in power systems and the importance of their associated IoT devices’ network security.

Provision of multiple services with controllable loads as multi-area thermal energy storage

Power systems are experiencing a decrease of synchronous generation along with increased penetration of inverter based renewable generation leading to reduced system inertia and a need for flexible resources. Non-generating resources such as thermostatically controlled loads (TCLs) are flexible due to their thermal energy storage capacity. When aggregated, TCLs can arbitrage energy prices and provide reserves to the power system. We approach the operational flexibility of the TCLs by modeling a risk-averse aggregator that controls decentralized TCLs and aims to maximize its own profit. The high number and low power rating of residential TCLs makes it difficult to model and assess their flexibility potential on national level. Thus, we make use of a high-level thermal energy storage model for aggregations of TCLs to quantify their flexibility potential. We present a method to aggregate temperature, TCL parameters, and building stock data into a thermal battery equivalent. We propose a multi-period multi-market multi-zonal two-stage chance constrained rolling horizon optimization problem formulation for the risk-averse day-ahead self-scheduling problem of a price-taker TCL aggregator bidding in energy and reserve markets under uncertainty and recast the problem as a linear program. We perform several case studies in the Swedish power system based on a survey of single- and two-family dwellings with electric heating and assess the flexibility potential. Additionally, a sensitivity analysis provides insights regarding market design and policy implications.