Emergency control, typically such as under-voltage load shedding (UVLS), is broadly used to grapple with low voltage and voltage instability issues in practical power systems under contingencies. However, existing emergency control schemes are rule-based and cannot be adaptively applied to uncertain and floating operating conditions. This paper proposes an adaptive UVLS algorithm for emergency control via deep reinforcement learning (DRL) and expert systems. We first construct dynamic components for picturing the power system operation as the environment. The transient voltage recovery criteria, which poses time-varying requirements to UVLS, is integrated into the states and reward function to advise the learning of deep neural networks. The proposed approach has no tuning issue of coefficients in reward functions, and this issue was regarded as a deficiency in the existing DRL-based algorithms. Extensive case studies illustrate that the proposed method outperforms the traditional UVLS relay in both the timeliness and efficacy for emergency control.
Department of Electrical and Computer Engineering The University of Texas at Austin（德克萨斯大学奥斯汀分校电气与计算机工程系）
Effective and timely responses to unexpected contingencies are crucial for enhancing the resilience of power grids. Given the fast, complex process of cascading propagation, corrective actions such as optimal load shedding (OLS) are difficult to attain in large-scale networks due to the computation complexity and communication latency issues. This work puts forth an innovative learning-for-OLS approach by constructing the optimal decision rules of load shedding under a variety of potential contingency scenarios through offline neural network (NN) training. Notably, the proposed NN-based OLS de…
Load shedding has been one of the most widely used and effective emergency control approaches against voltage instability. With increased uncertainties and rapidly changing operational conditions in power systems, existing methods have outstanding issues in terms of either speed, adaptiveness, or scalability. Deep reinforcement learning (DRL) was regarded and adopted as a promising approach for fast and adaptive grid stability control in recent years. However, existing DRL algorithms show two outstanding issues when being applied to power system control problems: 1) computational inefficiency …
Underfrequency (UF) load shedding schemes are traditionally implemented in two ways: One approach is based on manual load shedding, with system operators requesting loads to be shed ahead of anticipated stressful operating conditions. Manual load shedding is usually done through phone calls. The second method is automatic load shedding via underfrequency relays. Using static settings, these schemes can be designed to operate in stages and drop previously identified loads. The main limitation of traditional load shedding schemes is that they are reactive and leave little room for optimized corr…
Under voltage load shedding has been considered as a standard and effective measure to recover the voltage stability of the electric power grid under emergency and severe conditions. However, this scheme usually trips a massive amount of load which can be unnecessary and harmful to customers. Recently, deep reinforcement learning (RL) has been regarded and adopted as a promising approach that can significantly reduce the amount of load shedding. However, like most existing machine learning (ML)-based control techniques, RL control usually cannot guarantee the safety of the systems under contro…
This paper studies two fundamental problems in power systems: the economic dispatch problem (EDP) and load shedding. For the EDP, an extension of the problem considering the transmission losses is presented. Because the optimization problem is non-convex owing to quadratic equality constraints modeling the transmission losses, a convex relaxation method is presented. Under a particular assumption, it is shown that a distributed algorithm can be designed to solve the considered EDP. Furthermore, this work aims to handle an overloading problem by employing an optimal load shedding method. Emphas…
The work aims to improve the existing fast load shedding algorithm for industrial power system to increase performance, reliability, and scalability for future expansions. The paper illustrates the development of a scalable algorithm to compute the shedding matrix, and the test performed on a model of the electric grid of an offshore platform. From this model it is possible to study the impact on the transients of various parameters, such as spinning reserve and delay time. Subsequently, the code is converted into Structured Text and implemented on an ABB PLC. The scalability of the load shedd…
This paper introduces an advanced load shedding algorithm to improve the operability performance of a medium voltage direct current (MVDC) integrated shipboard power and energy system. Outcomes are compared to a baseline algorithm while considering power generation contingency scenarios. The case study is conducted with a real-time, embedded algorithm implementation using a control hardware-in-the-loop (CHIL) setup.
Distributed Load Shedding for Microgrid with Compensation Support via Wireless Network
Shanghai Jiao Tong University（上海交通大学）；Collaborative Innovation Center for Advanced Ship and Deep-Sea Exploration（先进船舶与深海探测协同创新中心）；Key Laboratory of System Control and Information Processing, Ministry of Education of China（教育部系统控制与信息处理重点实验室）
This paper proposes a cooperative game theory-based under-frequency load shedding (UFLS) approach for frequency stability and control in power systems. UFLS is a crucial factor for frequency stability and control especially in power grids with high penetration of renewable energy sources and restructured power systems. Conventional UFLS methods, most of which are off-line, usually shed fixed amounts of predetermined loads based on a predetermined schedule which can lead to over or under curtailment of load. This paper presents a co-operative game theory-based two-stage strategy to effectively …
The paradigm shift in the electric power grid necessitates a revisit of existing control methods to ensure the grid's security and resilience. In particular, the increased uncertainties and rapidly changing operational conditions in power systems have revealed outstanding issues in terms of either speed, adaptiveness, or scalability of the existing control methods for power systems. On the other hand, the availability of massive real-time data can provide a clearer picture of what is happening in the grid. Recently, deep reinforcement learning(RL) has been regarded and adopted as a promising a…
Enhancing restoration capabilities of distribution systems is one of the main strategies for resilient power systems to cope with extreme events. However, most of the existing studies assume the communication infrastructures are intact for distribution automation, which is unrealistic. Motivated by the applications of the emergency communication vehicles (ECVs) in quickly setting up wireless communication networks after disasters, in this paper, we propose an integrated distribution system restoration (DSR) framework and optimization models, which can coordinate the repair crews, the distribut…
In February 2021, an unprecedented winter storm swept across the U.S., severely affecting the Texas power grid, leading to more than 4.5 million customers' electricity service interruption. This paper assesses the load shedding experienced by customers under realistic scenarios in the actual power grid. It also conducts a preliminary study on using energy storage and load rationing to mitigate rotating blackout's adverse impact on the grid. It is estimated that utility-scale battery storage systems with a total installed capacity of 920 GWh would be required to fully offset the load shedding d…
Frequency Containment Reserves might be insufficient to provide an appropriate response in the presence of large disturbances and low inertia scenarios. As a solution, this work assesses the supplementary droop frequency-based Emergency Power Control (EPC) from HVDC interconnections, applied in the detailed Nordic Power System model. EPC distribution and factors that determine the EPC performance of an HVDC link are the focus of interest. The main criteria are the maximum Instantaneous Frequency Deviation and used EPC power. The presented methodology is motivated based on the theoretical obser…
Under voltage load shedding has been considered as a standard approach to recover the voltage stability of the electric power grid under emergency conditions, yet this scheme usually trips a massive amount of load inefficiently. Reinforcement learning (RL) has been adopted as a promising approach to circumvent the issues; however, RL approach usually cannot guarantee the safety of the systems under control. In this paper, we discuss a couple of novel safe RL approaches, namely constrained optimization approach and Barrier function-based approach, that can safely recover voltage under emergency…
This paper develops a community resilience optimization method subject to power flow constraints in the Cyber-Physical-Social Systems in Power Engineering, which is solved using a multi-agent-based algorithm. The tool that makes the nexus between electricity generation on the physical side and the consumers and the critical loads on the social side is the power flow algorithm. Specifically, the levels of emotion, empathy, cooperation, and the physical health of the consumers, prosumers are modeled in the proposed community resilience optimization approach while accounting for the electric powe…
The flexible power consumption feature of thermostatically controlled loads (TCLs) such as heating, ventilation, and air-conditioning (HVAC) systems makes them attractive targets for demand response (DR). TCLs possess a brief period where their power utilization can be altered without any significant impact on customer comfort level. This indicates TCLs are hidden potentials for providing ancillary services. This paper proposes a novel metric of demand response support time (DRST) for HVAC enabled demand response and a novel algorithm for the quantification of such HVAC-DR. The consumers' comf…
By manipulating tens of thousands of internet-of-things (IoT) enabled high-wattage electrical appliances (e.g., WiFi-controlled air-conditioners), large-scale load-altering attacks (LAAs) can cause severe disruptions to power grid operations. In this work, we present a rare-event sampling approach to identify LAAs that lead to critical network failure events (defined by the activation of a power grid emergency response (ER)). The proposed sampler is designed to "skip" over LAA instances that are of little interest (i.e., those that do not trigger network failure), thus significantly reducing t…
Developing effective strategies to rapidly support grid frequency while minimizing loss in case of severe contingencies is an important requirement in power systems. While distributed responsive load demands are commonly adopted for frequency regulation, it is difficult to achieve both rapid response and global accuracy in a practical and cost-effective manner. In this paper, the cyber-physical design of an Internet-of-Things (IoT) enabled system, called Grid Sense, is presented. Grid Sense utilizes a large number of distributed appliances for frequency emergency support. It features a local p…
The transmission grid is often comprised of several control areas that are connected by multiple tie lines in a mesh structure for reliability. It is also well-known that line failures can propagate non-locally and redundancy can exacerbate cascading. In this paper, we propose an integrated approach to grid reliability that (i) judiciously switches off a small number of tie lines so that the control areas are connected in a tree structure; and (ii) leverages a unified frequency control paradigm to provide congestion management in real time. Even though the proposed topology reduces redundancy,…
Voltage collapse is a type of blackout-inducing dynamic instability that occurs when power demand exceeds the maximum power that can be transferred through a network. The traditional (preventive) approach to avoid voltage collapse is based on ensuring that the network never reaches its maximum capacity. However, such an approach leads to inefficient use of network resources and does not account for unforeseen events. To overcome this limitation, this paper seeks to initiate the study of voltage collapse stabilization, i.e., the design of load controllers aimed at stabilizing the point of volta…
To enable power supply in rural areas and to exploit clean energy, fully renewable power systems consisting of cascaded run-of-the-river hydropower and volatile energies such as pv and wind are built around the world. In islanded operation mode, the primary and secondary frequency control, i.e., hydro governors and automatic generation control (AGC), are responsible for the frequency stability. However, due to limited water storage capacity of run-of-the-river hydropower and river dynamics constraints, without coordination between the cascaded plants, the traditional AGC with fixed participati…
Demand response (DR) for smart grids, which intends to balance the required power demand with the available supply resources, has been gaining widespread attention. The growing demand for electricity has presented new opportunities for residential load scheduling systems to improve energy consumption by shifting or curtailing the demand required with respect to price change or emergency cases. In this paper, a dynamic residential load scheduling system (DRLS) is proposed for optimal scheduling of household appliances on the basis of an adaptive consumption level (CL) pricing scheme (ACLPS). Th…
Load-generation balance and system inertia are essential for maintaining frequency in power systems. Power grids are equipped with Rate-of-Change-of Frequency (ROCOF) and Load Shedding (LS) relays in order to keep load-generation balance. With the increasing penetration of renewables, the inertia of the power grids is declining, which results in a faster drop in system frequency in case of load-generation imbalance. In this context, we analyze the feasibility of launching False Data Injection (FDI) in order to create False Relay Operations (FRO), which we refer to as FRO attack, in the power s…
Deep reinforcement learning (DRL) is a machine learning-based method suited for complex and high-dimensional control problems. In this study, a real-time control system based on DRL is developed for long-term voltage stability events. The possibility of using system services from demand response (DR) and energy storage systems (ESS) as control measures to stabilize the system is investigated. The performance of the DRL control is evaluated on a modified Nordic32 test system. The results show that the DRL control quickly learns an effective control policy that can handle the uncertainty involve…
