Cyber-physical-social systems (CPSS) integrate computing, physical assets, and human networks. Divided into four application areas to the electric grid, this book describes state-of-the-art CPSS in electric power systems, including detailed approaches on social constructs which are a critical aspect of the end-user realm. The book covers: Grid stability and security: distributed controls and algorithms for stability, cyber-security of generator units, and dynamic contingency analysis in the grid. Distribution system controls and economics: aggregators and agents for demand response, electric vehicles integration, smart neighborhoods, and microgrids. Social aspects and implementation: sociological customer interactions with the grid, SCADA systems, and occupants' roles in commercial buildings. Testbeds for validation: cyber-physical security assessment and distributed real time digital simulation. This book will be invaluable to academics and professional engineers engaged in cyber-physical social system applications for power engineering. It will also be of interest to social scientists, economists, and business personnel working on engineering applications for the Smart Grid, and students of power engineering.
Inspec keywords: SCADA systems; renewable energy sources; electric vehicles; load dispatching; distributed power generation; power engineering computing; security of data
Other keywords: generator coherency; cyber-physical-social system security testbeds; hybrid electric vehicles; electric cyber-physical systems; DC microgrid; cyber-physical-social systems; local renewable energy use; cyber-physical power networks; malicious attack; smart electric distribution grids; SCADA systems; social-economic dispatch
Subjects: General and management topics; Data security; Distributed power generation; Power engineering computing; Energy resources; Power system management, operation and economics; General electrical engineering topics; Transportation
The concept of cyber-physical systems (CPS) was introduced in 2006 to fill this gap in knowledge [1]. A CPS may be described as a typically large networked system, made of tightly interconnected physical and computational components, operating in a networked fashion. The history of CPS may be traced to the seminal article titled “As we may think” by Vannevar Bush in 1945, where the author challenged physicists returning from the war to develop the “memex,” a computer device for storing and accessing large sets of information like books and encyclopedias by humans [2,3]. Significant advances in the fields of communications and network engineering; computation, control and systems theory and engineering; information systems; Internet engineering; and sensor systems have led to the progression of the human-machine experience, thus paving the way for the evolution of the theory and hardware of CPS. Kim and Kumar present a detailed description of the history of CPS in Reference 4.
The next generation of the electricity grid, also known as `Smart Grid', is one of the most complex cyber-physical systems (CPS) due to its extreme dimension, geographic reach, and high reliability requirements. One of the main concerns for secure and reliable operation of power systems is the small signal stability problems caused by inter-area oscillations. In the future grid, enhancing the transfer capability while maintaining system stability requires damping these oscillations. In this chapter, we proposed a distributed optimal control framework using group sparse regularization functions. The proposed control aims to optimize a standard cost criterion while penalizing the number of communication links. The group sparse regularization approach is used to induce a desired communication structure and encode prior information about the underlying system into the control design. We present two applications of the proposed algorithm for damping inter-area modes in power networks. Our results suggest that the proposed method provides flexibility in designing wide-area type controls by allowing for a predefined communication structure. It can also act as an alternative approach to modal analysis methods in finding effective measurement-control loops in the system. The ability to encode system constraints in the control design objective is another major advantage of the method.
This chapter develops a distributed algorithmic framework for executing critical transmission-level operations of power systems using Synchrophasor data. As the number of phasor measurement units continues to increase exponentially, it is rather intuitive that the current state-of-the-art centralized communication and information processing architecture of wide-area measurement system will no longer be sustainable under such data-explosion, and a completely distributed cyber-physical architecture will need to be developed. Very little attention has yet been paid to the most critical consequence of this envisioned distributed architecture - namely, distributed algorithms. Our primary task in this chapter is to describe how one can develop distributed optimization methods for solving real-time wide-area monitoring problems with investigation of their convergence, robustness, and implementation issues.
The modern electric grid is rapidly evolving into a diverse system composed of cyberphysical entities such as distributed energy sources (DERs), data concentrators, and phasor measurement units (PMUs) that allow for intelligent actuation and active monitoring. Cyber-physical integration not only enables enhanced grid security and adaptive resilience but also promotes sustainable grid operations. Security is paramount in a system that is as tightly coupled as the grid. Without appropriate security mechanisms in place, failures in the grid can rapidly propagate vertically and result a catastrophic system-wide collapse which will be very expensive to recover from. Although cyber-physical entities in the grid are used primarily to mitigate threats, if existing communication vulnerabilities in the cyber components are not appropriately administered, these can be leveraged by adversaries to gain unauthorized access to highly sensitive physical grid components which can then be maneuvered to inflict serious damage to the grid. As such, we present two attack-mitigation approaches inspired by biology and differential games that actively engage uncompromised cyber-physical entities to restore system coherency in the face of malicious cyber attacks.
In recent years, the drive to bring about technological and regulatory changes that concern energy, natural resources, and climate change has gathered significant momentum. Of the numerous changes that the power grid is undergoing, perhaps the most transformative is the increased use of communication and computing technologies. The deployment of new communication, computing, and control technologies has significantly augmented the capabilities of traditional Supervisory Control and Data Acquisition systems. These technologies characterize the “smart grid” and have transformed them into the largest and most complex cyber-physical systems ever built; they also hold the potential to revolutionize power system operation and control paradigms. The system can often be protected from widespread consequences of failures of system components and other types of faults through timely detection and remedial action. By evaluating the stability of power systems and executing a sequence of remedial actions in real time or faster than real time, power systems can be hardened against cascading failures and unfolding events which can be either initiated by failures of system components, faults, or malicious attacks. This chapter discusses the application of real-time transient stability assessment and remedial action tools in enhancing power systems against potential cascading failures utilizing the developments in communication and computing technologies. Remedial actions are classified into preventive and corrective actions based on the time available to an operator to respond. Direct methods in the form of the energy function are utilized in transient stability screening and generating remedial actions. Different methods of calculating the controlling unstable equilibrium points (an equilibrium point that the trajectory of the system will go to after a destabilizing contingency) such as a Boundary of stability region-based Controlling Unstable equilibrium point method and a homotopy-based method are presented. Explicit derivations of expressions that are needed for transient stability analysis and remedial actions as well as an illustrative example are also provided.
We describe and contrast different market mechanisms to incentivize residential electricity customers to perform demand response (DR) via load shifting of schedulable assets. A customer-incentive pricing (CIP) mechanism from our past research is discussed, and compared to flat-rate, time-of-use (TOU), and real-time pricing (RTP). The comparison is made using a for-profit aggregator-based residential DR approach to solve the “Smart Grid resource allocation” (SGRA) problem. The aggregator uses a heuristic framework to schedule customer assets and to determine the customer-incentive price to maximize profit. Different customer response models are proposed to emulate customer behavior in the aggregator DR program. A large-scale system consisting of 5,555 residential customer households and 56,588 schedulable assets using real pricing data over a period of 24 h is simulated and controlled using the aggregator. We show that the aggregator enacts a beneficial change on the load profile of the overall power system by reducing peak demand. Additionally, the customers who are more flexible with their loads, represented as a parameter in the proposed customer α-model, have a greater reduction on their electricity bill.1
Modern electrical infrastructure is being challenged by increasing uncertainty from the introduction of renewable energy sources (RES) and distributed energy resources (DER). To cope with the challenges such technologies introduce, a considerable amount of attention has recently been given to the concept of “demand flexibility.” It is presented as an alternative to complement the current control and operation methodologies of the power grid, and to assist the penetration process of new technologies. Conventionally, flexibility is harnessed from power generation units and used by SOs to maintain the balance between supply and demand of the power system. However, as the penetration of stochastic generation and new forms of demand increase, this capability of conventional generation units might not be sufficient to cope with increasing uncertainty in both supply and demand. In response, throughout the literature, the necessity for a new flexibility source is highlighted. Demand side becomes active component in the control and operation of the power system, with the advances in ICT and computational intelligence in a so called cyber-physical social system (CPSS). The flexibility offered by the end-users, through for instance building energy management systems (BEMS), has the potential to help not only resolve network and system problems, but also accommodate a higher amount of renewables, increase asset utilization, and reduce peak demand. Yet, this is not a straightforward transition toward a smart energy system or smart grid (SG). Advanced energy management systems are required to manage the flexible demand while integrating emerging technologies. This chapter introduces a SG-BEMS interoperation framework and demonstrates the capabilities and benefits of multi-agent systems (MASs) in enabling the correct operation of the emerging SG, while unlocking the flexibility potential of the built environment.
In this chapter, we will explore the challenges and opportunities of grid integration of electric and hybrid electric vehicles in cyber-physical-social systems (CPSS). Transportation systems that move people, goods, and services in societies worldwide pose unprecedented environmental, economic, and social challenges, particularly with the growing urgency to conserve energy, cut back on carbon emissions and pollution, avoid crashes, and relieve congestion. Advances in intelligent transportation systems and Smart Grid offer great promise to address these challenges and have the potential to revolutionize future transportation systems. In the last decade, the government around the world has spurred efforts to boost the utilization of transportation electrification technologies because of their low-pollution emissions, energy independence, and high fuel economy. An ever-increasing number of electric and hybrid electric vehicles will radically change the traditional view of the power industry, transportation industry, social environment, and business world. Research on grid integration of electric and hybrid electric vehicles typically addresses topics at the vehicle-grid boundary, such as peak load impacts and agent-based charging control. While researchers around the world are making significant advances in these areas, there is very little work addressing the coupled cyber-physical-social effects of electric and hybrid electric vehicle charging with the mobility-focused, transportation ecosystem to meet the dynamic needs of a changing society. It is important to recognize that today's critical infrastructure is an interdependent network of networks. A single network consists of millions of subnetworks and individual agents. The “tie point” is critical to its reliability, cost-effectiveness, and resiliency. As such “tie points,” the emerging deployment of electric and hybrid electric vehicle charging facilities would complicate the understanding and design of interdependent critical infrastructure systems. For example, as a transportation tool and electricity carrier, electric and hybrid electric vehicle can be charged at any charging facility and at any time, which brings more spatial and temporal uncertainty to the power grid's load forecast. The retail electricity price and parking fee may also have an impact on customer behavior, eventually leading to a change in traffic flow. Besides engineering considerations, the placement of electric and hybrid electric vehicle charging stations are constrained by applicable local policy and regulation, financial incentives, and public interests. We will discuss the state of the art of grid integration of electric and hybrid electric vehicles in a CPSS environment. Moreover, we will present a future perspective to enable the dramatic increase of electrified vehicles, and ultimately lead to (1) reduced fossil fuel consumption; (2) reduced carbon emissions and pollution; (3) increased customer satisfaction; (4) increased reliability and efficiency for moving people and goods; (5) improved efficiency of intelligent transportation systems; (6) accelerated adoption of Smart Grid technologies; and (7) increased use of infrastructure capacity.
As distributed generation (DG) gains in popularity throughout the world, neighborhoods are expected to turn into small microgrids that may be able to operate autonomously from the grid when needed. Such cases may arise to reduce energy costs or when facing an outage of the distribution system. This feature is only possible if such neighborhoods are constructed with an advanced metering infrastructure, and homes are equipped with smart meters and home energy management systems to make local resources (DG, loads, storage, and electric vehicles) accessible and controllable. This, in turn, enables self-consumption mechanisms, where smart homes consume their own generated energy, when coupled with efficient energy management strategies. As the possibilities offered by a typical smart home are limited, mainly due to cost and comfort constraints, more possibilities are offered when several homes can coordinate their actions with each other, i.e., by sharing their resources and scheduling their use appropriately. By communicating with each other, homes can form groups (based on grid topology and economic criterion), estimate the total available energy capacity of the neighborhood, and collaboratively allocate energy generation, consumption, and storage over time. Homes are modeled as rational decision-makers (agents) in the neighborhood, and can cooperate with each other to meet the needs of their occupants.
DC microgrid is a feasible and effective solution to integrate renewable energy resources, as well as to supply reliable electricity. The control objective of DC microgrids is to maintain the system's stable operation, low-voltage regulation, and proportional load sharing among the multiple distributed generators. Compared to the high-bandwidth communication-dependent master-slave control, droop control is an effective method to implement the control of DC microgrids without the requirement of communication. Droop control is an output impedance programming method, in which the output current decreases linearly with the decrease of output voltage. The load sharing is automatically achieved. However, in the real applications of lowvoltage DC microgrids, the nominal reference offsets and unequal cable resistances require trade-offs to be made between voltage regulation and load sharing. Thus some compensations need to be performed so as to solve this problem. This chapter discusses the methods to compensate the voltage error introduced by droop control as well as the unequal load sharing due to the transmission lines and the nominal voltage reference offsets. The compensation methods in the literature using low-bandwidth communication are reviewed and a unified compensation framework is proposed using the common current. In this scheme, the voltage deviation and the unequal load sharing are compensated separately. The common current is generated in each local controller by using the local module currents shared in a dedicated low-bandwidth communication line. The contents of this chapter are organized as follows: Section 10.1 overviews and compares the active current sharing control and the droop control in DC microgrids; Section 10.2 analyses the limits of the basic droop control under the condition of nominal voltage offsets and unequal connecting cable impedances; Section 10.3 reviews and classifies the different compensation methods from the literature; Section 10.4 analyses voltage and load sharing performance of the proposed method, and investigates the boundaries of the compensation parameters to maintain system stability. In Section 10.5, some simulations are conducted in the MATLAB/Simulink environment. In Section 10.6, experimental tests are performed on a laboratory-scale test bench to verify the previous proposed theoretical analysis.
Nowadays, plug-in electric vehicles (PEVs) are widely promoted to replace conventional internal combustion-based vehicles to reduce dependency on fossil fuels and decrease greenhouse gases emissions. This is likely to create a huge stress to the power system and create new problems in its operation due to uncontrolled power and energy transactions by PEVs. Nevertheless, the stress can be avoided by proper coordination and cooperation of responsive PEVs (REVs) with utilities and system operators. This chapter presents a study on the integration of power system generation assets, REVs, and behavior of REV owners as a cyber-physical-social power system (CPSPS). Sustainability criteria are applied in a CPSPS framework to formulate a social-economic dispatch with REVs (SED-REVs) problem. This involves the participation of REVs through modeling of the behavior of REV owners and the load demands of a given power system. Two types of modeling are carried out, namely, first, modeling the participation percentage of REVs as their responsiveness level with respect to incentive schemes in a SED-REVs problem and second, stochastic modeling of the presence of REVs in SmartParks (parking lots for PEVs with capability to carry out energy transactions). The variability and stochastic nature of the REVs behavior and load demand is addressed by applying a heuristic optimization methodology to solve the SED-REVs problem.
Inside a smart home, smart appliances operate automatically. Outside, a smart meter receives and transmits information to the utility. In an instant, the smart grid optimizes the use of distributed resources to ensure that an uninterrupted supply of power is delivered to the home. This is the vision that energy technology experts have held of the smart grid, but there is one key piece of this picture missing -people. However, the goals of the smart grid can only be achieved with the cooperation of the consumer. Moreover, one of the defining characteristics of the smart grid is that it enables informed participation by customers. Thus, the consumer is integral to the success of the future of the smart grid. This chapter provides a roadmap for interdisciplinary collaboration between technology experts and social scientists, and provides concrete suggestions for getting consumers on board with the future of the smart grid.
Industrial Control Systems are at the heart of our critical infrastructures. Supervisory Control and Data Acquisition Systems (SCADA) are providing engineers with visibility and control of large-scale distributed control system infrastructures. Whilst these systems have in the past been isolated, today's technological advancements and changes in business needs mean that they are increasingly integrated with other networks that are connected to the Internet and infrastructures. This chapter provides an introduction to SCADA systems and reviews and analyses the resulting risks, threats arising from cyber attacks. The remainder of the chapter then investigates strategies to mitigate some of the risks that result from the increasing exposure of SCADA systems through prevention, attribution and incident response.
Demand response (DR) refers to the active participation by retail customers in electricity markets, seeing and responding to prices as the prices change over time [1]. Occupants are excluded from traditional commercial building DR control loops. However, our experimental results suggest that motivated occupants can achieve significant load shaving with proper information. The analysis indicates more load shaving potentials with financial incentives. During the DR events, utility companies send DR requests to commercial building facility managers (FMs), who often send emails to occupants with generic instructions. Unless the buildings are equipped with advanced hardware, office building FMs have limited means to reduce significant peak load in the DR periods. We present a software-based, occupant-engaged fast DR system for commercial office buildings. Our collaborative DR (cDR) module is built upon our collaborative energy management and control (cEMC) platform. Through a web portal, occupants can submit a preferred temperature range and schedule. In this chapter, we present a novel occupant-engaged collaborative DR system from the perspective of cyber-physical-social systems (CPSSes). Traditional building automation systems (BASes) are designed based on thermal and mechanical requirements, without considering the impact of occupants' psychological motivations, such as peer pressure, social recognition, and gaming experiences. The cDR system is designed with cybernetics and social factor in mind. From the cybernetics perspective, we introduced a semantic building data model to enable micro-zoning scenarios. From the social behavior perspective, (1) we developed a zonal virtual energy meter algorithm to split the whole building energy consumption into individual occupants and (2) provide an embedded social network, in order to encourage an energy competition game. The effectiveness of the energy game is validated by real building experiments. In addition, we developed a game-theoretical optimal incentive design (OID) algorithm to allocate real financial incentives to individual occupants, and conducted a simulation-based study. During field experiments on a mid-size office building at Pittsburgh, in 2014 summer, we observed up to 1.6% load reduction for the traditional email-based DR method, and 11-15% load reduction using our cDR system without compromising on comfort. Up to 56.7% load was shed, with an acceptable loss of comfort.
To analyze the complex interdependencies of power, social, and cyber domains, it is necessary to develop a cohesive modeling and simulation architecture for cyberpower-social systems. A real-time cyber-physical testbed with abilities to perform unified simulation, emulation, and interface with hardware devices provides a great platform to analyze the impact of cyber events on the power grid. In this chapter, development of a real-time cyber-physical testbed and applications have been discussed. The developed testbed has four different layers: (1) power system layer, (2) sensor and control layer, (3) communication layer, and (4) application layer. In the power system layer, Real-Time Digital Simulator (RTDS) and related software RSCAD are used to simulate the power system. In the sensor and control layer, real physical hardware devices such as phasor measurement units (PMU) and phasor data concentrator (PDC) have been used as sensors and supporting infrastructure. RealTime Automation Controller (RTAC) and Synchrophasor Vector Processor (SVP) are used as controllers in this layer. Satellite-synchronized clock provides the accurate time signal to all the synchrophasor devices and simulator. In the communication layer, DeterLab have been used to emulate the communication network. With the unique features available in the DeterLab, cyber attack generation, cyber protection control, and communication traffic collection and analysis can be performed to analyze the impact of cyber events on the power grid. In the application layer, Real-Time Voltage Stability Monitoring and Control (RT-VSMAC) Tool is used to demonstrate the wide-area monitoring and close-loop control concepts.
With the ever increasing cyber security concerns for critical infrastructures such as the power grid against sophisticated and advanced persistent threats, the need for realistic cyber-physical-social system (CPSS) security testbed environments is compelling. CPSS security testbeds capture the complex interactions between the various domains and their interdependencies appropriately, while also providing necessary system abstractions for limiting experimental scale and yet maintaining higher fidelity than traditional simulation-based tools. In this chapter, we motivate the need for CPSS security testbeds and provide a quick overview of various objectives in the design of such testbeds. We then provide a comprehensive survey of state-of-the-art research in the area of CPSS security testbeds. We also identify key research areas enabled by CPSS security testbeds and highlight certain limitations of individual testbeds, which motivate the need to federate CPSS security testbeds. As part of experimental case studies, we present a brief proof-of-concept case study to illustrate the feasibility of CPSS security testbed federation. We also describe a detailed case study of coordinated cyber attacks on Wide-Area Monitoring, Protection, and Control (WAMPAC) applications for the Smart Grid using the PowerCyber CPSS security testbed at Iowa State University. We conclude this chapter with a brief overview of how CPSS security testbeds could be leveraged beyond experimental research for educational and outreach activities to enhance the overall cyber security awareness of the current and future workforce in this area.
Real-Time Simulations (RTS) are increasingly being used to understand the complex device and system level interactions in power grids. RTS provides the capability to create detailed, highly accurate, and diverse set of power and control system components at low time steps (order of microseconds) that are based on “real-world clock-time.” RT simulator is a unique architecture with specialized processors and communication boards that allow time synchronization of simulations and the clocktime. Lean operating systems, specialized processors, faster communications, etc. are the typical attributes of RT simulators. RT simulators provide a unique capability of interfacing with power and control components via analog and digital interfaces. However, RT simulators have limited computational capability that constrains the size of power and control systems that can be simulated. Multiple simulators connected locally is typically used to increase the computation capability, however this is not always economical. Additionally, RT simulators are used at facilities with unique test infrastructure in the form of grid emulators, inverters, photovoltaic, wind turbine, microgrids, etc. Performing distributed RTS via Internet can augment simulation capacity and leverage unique infrastructure that is dispersed in academia and research laboratories. Research related to distributed RTS and its application in electric power engineering is discussed in this chapter.