Search

Moving Target Defense (MTD) is a new defensive strategy to protect power system state estimation from cyber attacks. Using the distributed flexible AC transmission system (D-FACTS), MTD works by actively perturbing the branch parameters that are needed to construct the false data injection (FDI) attacks. Although there are many pioneer works on MTD, the relationship between the construction of MTD and detection of FDI attacks has not been revealed. In this paper, we reveal the correlation between the MTD design and FDI detection and optimize MTD's performance in terms of detecting FDI attacks. Moreover, with the aim to reduce the number of measurements that can be compromised by the attacker, we develop a heuristic algorithm to compute a near-optimal solution for the deployment of D-FACTS devices. Furthermore, we prove that the coordinated design of consecutive perturbation schemes within an MTD cycle can improve the performance of MTD in terms of detecting FDI attacks. Finally, we conduct extensive simulations with the IEEE standard test power systems to validate our findings.

The cybersecurity of the distributed AC optimal power flow (ACOPF) against false data injection attacks (FDIA) is investigated in this paper. A collaborative distributed ACOPF solver, rooted in the dual decomposition concept, is first formulated, based on which a theoretical framework is then developed to model the distributed ACOPF and its cybersecurity in the presence of FDIA. Under this proposed cybersecurity framework, a reputation-based peer-to-peer trust management system (TMS) is proposed to secure the resilience of the system against FDIA. Finally, the proposed TMS is validated on the IEEE 69-bus benchmark system. The primary contribution of this paper is the proposed holistic resilience framework, in which the FDIA is analytically assessed and effectively defended. In the analytical results, we have established two key propositions: the FDIA generally has one degree of freedom; and the FDIA can be effectively detected and mitigated using the proposed TMS with one-hop redundancy.

This article investigates the privacy issues in the distributed energy management system (EMS) of smart distribution systems and microgirds. A novel private collaborative distributed energy management system (P-CoDEMS) is proposed to solve the AC optimal power flow (ACOPF) problem in a distributed and private fashion. The proposed P-CoDEMS algorithm is based on an original primal dual subgradient distributed optimization technique and a state-of-the-art fully homomorphic encryption algorithm. The convergence and optimality of the proposed P-CoDEMS algorithm are evaluated on four representative systems. Simulation results indicate that the proposed P-CoDEMS algorithm can accurately solve the ACOPF problem in a fully distributed way while preserving individual agent privacy.

The uncertainties of renewable energy, loads, and electricity prices pose significant challenges to the economical and secure energy management of smart grids. In this article, a hybrid stochastic/robust (HSR) optimization method is developed to minimize the overall cost of all units. The proposed approach takes advantage of stochastic programming, robust optimization, and distributed optimization methods while considering various system constraints. First, stochastic electricity price scenarios are selected by the Latin hypercube sampling method. Second, the uncertainties of renewable energy generation and loads are managed by the proposed robust optimization method under each price scenario. Then, an improved distributed optimization method is proposed to solve the formulated HSR optimization problem, which considerably enhances the convergence with the accelerated gradient method. Numerical case studies of both small-scale and large-scale power systems demonstrate the accuracy, effectiveness, and scalability of the proposed distributed HSR approach. Additionally, the optimality and convergence of this proposed distributed algorithm are mathematically proven and analyzed.

Safe and reliable management of rechargeable lithium-ion batteries relies on accurate state-of-charge (SOC) estimation. SOC estimation algorithms developed based on the conventional state-space model can be inaccurate when the slope of the battery SOC to open-circuit voltage ( $V_{\text{OC}}$ ) profile is close to zero, where the observability matrix of the conventional battery state-space model is ill-conditioned. This ill-conditioned observability issue is usually overlooked and results in unreliable SOC estimations. This article proposes a robust state-space model using a set of carefully selected states, making the proposed model well-conditioned over the full SOC range and insesitive to the slope of the SOC- $V_{\text{OC}}$ profile. The proposed model, therefore, solves the observability issue and improves the SOC estimation accuracy. To examine the performance of the proposed state-space model over the full SOC range, the battery is discharged using a standard electric vehicle driving profile. Experiments are further carried out in various environmental temperature conditions to validate the robustness of the proposed model. The results show that the proposed state-space model is robust against the change of the slope of the SOC- $V_{\text{OC}}$ profile and significantly improves the SOC estimation accuracy, rendering more than 30% improvement in various temperature conditions.

Recently, collaborative distributed energy management systems (CoDEMS) have emerged as an effective solution to manage distributed energy resources in microgrid. In CoDEMS, devices collaborate in a distributive manner over communication networks to meet electrical loads and supply balance at minimum cost. However, mutual information exchanges among the devices in CoDEMS may leak important information about the devices states. In this paper, we investigate the challenging problem of how to achieve optimality while preserving the privacy of CoDEMS at relatively low cost. Unlike many previous works that preserve the privacy by using additive noises, we propose a novel random-weight privacy-preserving algorithm with error compensation, termed as REP-CoDEMS, for CoDEMS. In the proposal, each distributed device generates two random weights each time and it communicates with its neighbor conveying values based on the weights, incremental cost estimation and power imbalance estimation information along with a novel error compensation term to eliminate the error induced by the random weights. We theoretically prove that the proposed REP-CoDEMS algorithm converges and preserves the privacy of all devices. We also derive analytical expressions of the maximum privacy disclosure probability for initial and final states of the CoDEMS. In addition, we conduct extensive simulations and the results demonstrate the effectiveness of the proposed algorithm.

This paper proposes a novel dual decomposition based algorithm that solves the AC optimal power flow (ACOPF) problem in the radial distribution systems and microgrids in acollaborative and distributed manner. The proposed algorithm adopts the second-order cone program relaxed branch flow ACOPF model. In the proposed algorithm, bus-level agents collaboratively solve the global ACOPF problem by iteratively sharing partial variables with its 1-hop neighbors as well as carrying out local scalar computations that are derived using augmented Lagrangian and primal-dual subgradient methods. In addition to the algorithm innovation, this paper proposes two distributed computing platforms, i.e. high performance computing based and hardware-in-the-loop platforms, to validate and evaluate the proposed distributed algorithm. The computationand communication performances of the proposed algorithm are quantified and analyzed over typical IEEE test distribution systems. Experimental results indicate that 1) the proposed collaborative distributed ACOPF solver can be executed on a fully distributed computing structure and yield accurate ACOPF solution; and 2) the proposed distributed algorithm has a low communication overhead.

This article investigates the resilience of the consensus-based distributed economic dispatch (ED) in cyberphysical DC microgrid energy management system (EMS). A holistic framework, called resilient collaborative distributed EMS (R-CoDEMS), is proposed to model the system dynamics, adversaries, and the cybersecurity of the consensus-based distributed ED. Under the R-CoDEMS framework, a reputation-based distributed attack detection and mitigation algorithm is also proposed to improve the system’s resilience against typical non-colluding and colluding false data injection attacks. The effectiveness and scalability of the proposed R-CoDEMS countermeasure is evaluated on a typical microgrid system and large-scale networks via numerical simulations.

In an integrated energy system, the growing number of distributed heat and electric power generation units will bring new technical challenges to the existing centralized economic dispatch strategies. This paper proposes a distributed optimization approach for the economic system operation in a multienergy system by considering various equality and inequality constraints to accommodate the integration of intermittent renewable generations. The proposed distributed neurodynamic-based approach only requires the information exchange among neighboring units and offers flexibility, adaptivity, scalability, faster convergence, and lower communication burden compared with some traditional centralized methods. The simulation results of two integrated energy systems validate the effectiveness of the proposed distributed approach. Comparisons with other centralized and distributed optimization methods quantify the advantages of the proposed distributed approach in terms of convergence speed and computation complexity.

This paper investigates the issue of power management networked ac–dc microgrids (MGs) interconnected by interlinking converters with the consideration of unbalanced single-/three-phase ac MGs as well as power quality improvement. An integrated hierarchical distributed coordinated control approach is developed, which mainly consists of an up-layer event-triggered method of power sharing among MGs, and an event-triggered dynamic power flow routing approach to navigate the power flow among phases of the single-/three-phase ac MGs to balance the power of the MG. With the proposed control method, balanced output phase powers for the three-phase distributed generation (DGs) and enhanced voltage quality at the point of common coupling and DG terminals can be achieved besides proportional active power sharing among MGs and reduced communication. Simulation results are presented to demonstrate the proposed control method.

The well-developed consensus algorithm provides an elegant distributed way for solving the energy management problem. The convergence of the consensus-based distributed approach depends on the reliable exchange of the information with neighbors. This assumption might be violated in practice due to inevitable and random packet losses. In this paper, the impact of packet losses on the consensus-based distributed approach is analyzed. We show the iterative calculation converges to an incorrect solution in the presence of packet losses, where the power supply could not meet with the demand. More importantly, we find the error in the final result is accumulated over every packet loss event, and we identify the critical information that cause the error. Based on the analysis, a corrective method is proposed to assure convergence to the right schedule. The corrective method has two features: 1) it uses the original consensus network, and no communication reconfiguration is needed; and 2) no retransmissions for the lost packets are needed, instead, the errors are compensated by a new variable introduced in the update rule. We show that the proposed algorithm could achieve the optimal solution in the presence of packet losses. Numerical simulation results are used to validate the proposed algorithm.

Privacy is of great importance for microgrids and has gained much attention recently. By eavesdropping on the communications among the devices, attackers may infer sensitive system operation information and user behavior due the intimate interplay between communication and control in mircogrids. In this paper, in order to facilitate simulational evaluations of privacy preservation techniques for microgrids, we develop a versatile cyber-physical co-simulator which integrates both networked communication and power system control subsystems as a whole. The co-simulator is built upon MATLAB/Simulink and OMNeT++ along with a module that coordinates the two tools in real-time simulations. Based on the co-simulator, we evaluates three privacy-preserving algorithms proposed in the literature, and find that SFPA performs better than PEMA and REP-CoDEMS in aspect of protecting the privacy of controllable inputs, but REP-CoDEMS and PEMA has a better performance considering both controllable inputs and uncontrollable inputs.

Cyber-physical systems (CPS) refers to the integrative system consisting of interconnected computing and control devices interacting with the physical infrastructure via sensors and actuators. Recently, there is a swift growth of CPSs ranging from smart grids to smart buildings, robotics, and other industrial control systems. They have formed the keystone of the sustainable growth of the economy, manufacturing, and smart and connected communities. Due to extensive applications of CPSs, their resilience, reliability, and security are paramount. Many factors, however, pose significant threats to CPSs and lead to high economic losses and social impacts. Software defects also make CPSs vulnerable to security attacks and coordinated cyber and physical attacks. To address this issue, emerging technologies and methods for understanding and improving the resilience, reliability, and security of CPSs are needed. This Special Section aims to provide a platform to help define, understand, and quantify the resilience, reliability, and security of CPSs.

This paper addresses the vulnerability of consensus-based distributed energy scheduling algorithm to data integrity attacks. A reputation-based neighborhood-watch mechanism is introduced to detect the false information and achieve optimal operating point in the presence of misbehaving controllers. The reputation-based neighborhood watch algorithm has three major functions: 1) verifying the correctness of neighbors’ information based on two-hop shared information; 2) identifying the compromised controller based on reputation indexes; and 3) maintaining the correctness of local information estimation in the presence of false information. The effectiveness of the proposed method is illustrated through simulation analyses in the future renewable electric energy delivery and management system.

This paper presents an improved secondary voltage control (SVC) methodology incorporating compressive sensing (CS) for a multi-area power system. SVC minimizes the voltage deviation of the load buses while CS deals with the problem of the limited bandwidth capacity of the communication channel by reducing the size of massive data output from phasor measurement unit (PMU) based monitoring system. The proposed strategy further incorporates the application of a Morphological Median Filter (MMF) to reduce noise from the output of the PMUs. To keep the control area secure and protected locally, Mathematical Singular Entropy (MSE) based fault identification approach is utilized for fast discovery of faults in the control area. Simulation results with 27-bus and 486-bus power systems show that CS can reduce the data size up to 1/10th while the MSE based fault identification technique can accurately distinguish between fault and steady state conditions., Comment: 10 pages, 16 figures

This paper introduces a novel data integrity attack on the well-developed consensus-based energy management algorithm. In particular, we show that by sending out elaborately falsified information during the consensus iterations, attackers could manipulate the system operating point and gain extra economic benefits. Meanwhile, the system-level and device-level constraints are still satisfied, e.g., the power generation and demand are balanced, and the operation of individual device respects physical constraints. This data integrity attack has two major features: First, attackers rely only on local information to complete the attack; neither additional information about system topology nor additional colluders are required; second, the attacking effect is accumulative, which enables attackers to choose to finish in either single or multiple iterations. By revealing such vulnerability of consensus-based applications to data integrity attack, this paper conveys the message that besides the efforts of designing novel distributed energy management algorithms to address the renewable energy integration challenges, it is equally important to protect the distributed energy management algorithms from possible malicious attacks to avoid potential economic losses. The proposed attack is illustrated in the Future Renewable Electric Energy Delivery and Management system.

The transient stability control of power systems with growing penetration of renewable energy resources is challenging due to inherent small damping of generators and complicated operating conditions. To address the drawbacks of existing control approaches which need accurate systemwide network parameters, a model-free fuzzy controller is proposed to enhance the transient and frequency stability of power systems. Also, an adaptive parameter estimation scheme is developed to eliminate the fuzzy approximation errors and compensate the external disturbances. The proposed strategy is implemented based on the multiagent framework, which enables the sharing of communication and computation burdens among local controllers for fast and coordinated response. The convergence of the proposed distributed control approach is rigorously proved using the Graph theory and Lyapunov stability theory. Simulation studies validate the effectiveness of the proposed distributed control approach.

Lithium-ion batteries are widely used in many fields of modern life, e.g. wearable devices, electric vehicles and electric grids, etc. The safety and reliability of the lithium-ion battery are critical issues during the battery operation, where the battery management system (BMS) plays a key role. An accurate estimation of the state-of-charge (SOC) of the battery is essential for the BMS. However, due to the intrinsic nonlinearity of the lithium-ion battery, the accurate estimation of the SOC is technically challenging and has drawn lots of attention both from academic and industrial fields. In order to tackle this difficulty, many SOC estimation approaches have been proposed, in which an identification method for the parameters of the battery is normally implemented. However, the additional parameter identification approach greatly reduces the efficiency of SOC estimation and the bias from identification may significantly affect the accuracy of the SOC estimation. This paper proposes a novel data-driven SOC estimation approach based on the adaptive residual generator, which realizes integrating the parameter identification and the SOC estimation into a simultaneous procedure, where the convergences for both the parameter identification and SOC estimation are guaranteed. The proposed adaptive residual generator can estimate the SOC of the battery accurately due to real-time parameter identification that proactively minimizes the modeling error. The effectiveness and the performance of the proposed method are demonstrated through the case studies on a battery simulator. Also, owing to accurately identified parameters, the SOC of the battery is estimated accurately with almost 0% SOC estimation error.

Rechargeable lithium-ion batteries are now widely adopted in our life. To fulfil various energy and power requirements in real-world applications, battery cells are connected to form battery packs. The cell-to-cell difference exists in the battery pack after manufactured, and this difference will further deteriorates when the battery cells are exposed and used in various operating conditions. This unavoidable cell-to-cell difference results in early cut-off on the battery pack, which influences the performance of the battery pack and makes accurately estimating the battery pack SOC challenging. This paper proposes a novel real-time algorithm to effectively identify the most significant cells in a serial-connected battery pack in order to accurately estimate the SOC of the battery pack. A battery pack composed of ten serial-connected battery cells is carried out in this paper to evaluate the performance of the proposed algorithm. The results show that the most significant cells are successfully identified, and the SOC of the battery pack is estimated accurately based on the identified most significant cell.

Lithium-ion battery packs have been widely applied in many high-power applications which need battery management system (BMS), such as electric vehicles (EVs) and smart grids. Implementations of the BMS needs a combination between software and hardware, which includes battery state estimation, fault detection, monitoring and control tasks. This paper provides a comprehensive study on the state-of-the-art of machine learning approaches on BMS. It differentiates between these methods on the basis of principle, type, structure, and performance evaluation.

In collaborative distributed energy management system (CoDEMS) with distributed energy storage devices (DESDs), privacy is getting more and more attention, because customer and grid may suffer losses due to unintentional privacy disclosure. In this paper, we firstly introduce the networked model, distributed energy management problem and consensus-based CoDEMS algorithm. Then, we analyze the disclosure of true power supply and demand with adding-noise privacy-preserving consensus-based distributed (APCD) algorithm. To preserve the privacy, we proposed the random-weighted privacy-preserving consensus-based collaborative distributed energy management system (RP-CoDEMS) algorithm to preserve the confidentiality of the neighborhood communication. The effectiveness of RP-CoDEMS algorithm is demonstrated by simulation.

Consensus-based distributed microgrid energy management system is one of the most used distributed control strategies in the microgrid area. To improve its cybersecurity, the system needs to evaluate the trustworthiness of the participating agents in addition to the conventional cryptography efforts. This paper proposes a novel augmented reputation metric to evaluate the agents' trustworthiness in a distributed fashion. The proposed metric adopts a novel augmentation method to substantially improve the trust evaluation and attack detection performance under three typical difficult-to-detect attack patterns. The proposed metric is implemented and validated on a real-time HIL microgrid testbed.

Shape morphing is one of the most appealing applications of adaptive structures. Among the various means of achieving shape morphing, origami-inspired folding is particularly advantageous, because folding is a powerful approach to induce three-dimensional and sophisticated shape changes. However, attaining large-amplitude folding is still a challenge in origami engineering. While promising, the use of active materials as a folding activation strategy is limited due to the constant voltage supply that is required to maintain the desired configuration of the structure. One possible solution is to embed bi-stability into the structure. Bi-stability can play two significant roles here: first, it can significantly reduce the actuation requirement to induce shape morphing, and second, it can maintain the shape change without demanding sustained energy supply. In a previous study by the authors, a unique shape morphing (or self-folding) method using harmonic excitation has been proposed for a bi-stable water-bomb base. However, this approach has some drawbacks because the nonlinear dynamic behaviors of origami are quite sensitive to different design parameters, such as initial conditions, excitation parameters, and inaccuracies in manufacturing. In this study, via numerical simulations, we show that by harnessing the intra-well resonance of the water-bomb structure and incorporating a relatively simple feedback control strategy, one can achieve a rapid and robust morphing using relatively low actuation magnitude. The results of this study can lay the foundation of a new category of morphing origami mechanisms with efficient and reliable embedded actuation.

This article investigates the resilience of the consensus-based distributed microgrid energy management system (MG-EMS) against profit-driven attacks. A holistic framework, called reputation-based collaborative distributed EMS (R-CoDEMS), is proposed to model the system dynamics, profit-driven adversaries, and the cybersecurity of the distributed MG-EMS under profit-driven attacks. Simulation results on a real microgrid system indicate that the R-CoDEMS can effectively detect, mitigate, and recover from the attacks in a fully distributed fashion. The main technical contributions of this research work are: (1) systematically formulating and quantifying the system adversary, assets, threats, vulnerabilities, and risks of the consensus-based distributed MG-EMS; (2) proposing a reputation-based attack detection and mitigation strategy to overcome single and coordinated profit-driven attacks.

This paper investigates and addresses the vulnerability of the distributed DC optimal power flow (DC-OPF) algorithm to data integrity attacks. In particular, we first show that a compromised distributed controller on a single bus could manipulate the power dispatch result by sharing false information to neighboring buses. Two malicious scenarios of launching the data integrity attack are considered, namely economic-driven and infeasibility-driven attacks, respectively. These vulnerabilities demonstrate a growing need for anomaly detection and mitigation mechanisms that fit into the distributed power system framework to counteract highly skilled malicious cyber attackers. We then introduce a resilient distributed DC-OPF algorithm with an embedded attack-resilient control mechanism. It performs two major functions in a fully distributed way: 1) verifying the correctness of the shared information from neighboring buses while protecting each other’s privacy and 2) identifying the compromised distributed controllers and recovering the optimal power dispatch result from the impact of data integrity attacks. The effectiveness of the proposed attack-resilient mechanism is illustrated through case studies in the IEEE 14-bus system.

The State of Health (SOH) of the battery is often represented either using the decrease in the capacity of the battery or the increase in the internal resistance of the battery. While these indices are commonly used, they do not provide any insight on the reasons for the degradation of the health of the battery. Understanding battery aging and the impact it has on the working/performance of the battery is required to determine the State of Function (SOF) of the battery for that particular application. The SOF of the battery can provide information on the current applicability of the battery to the application. The Remaining Useful Life (RUL) of the battery is also highly dependent on the current and past operating conditions. Determining the reason behind the degradation and the impact on the health can also help determine the RUL or provide feedback to the user on alternate usage patterns to prolong the RUL. This paper uses a first principle based degradation model to determine the sensitivity of the terminal voltage and capacity of the battery to the degradation of the concentration of lithium ions in the anode/negative electrode.

In this paper, a coordinated adaptive droop control is addressed for DC microgrid to optimize its power distribution. The optimal solution for economical dispatch problem (EDP) of the microgrid is found through a fully distributed hierarchical control. The consensus-based economical regulator calculating the optimal solution for the generator is fully distributed. Thus, it eliminates the requirement of centralized coordinator. The droop controller receives the reference from the economical regulator and ensures the output power converging to the reference while maintaining the power balance of the system. Moreover, the economical regulator can estimate the load information of the system based on the characteristics of droop control. Thus, the information from load or renewable energy source is no longer required to solve the EDP that consequently decreases the number of communication nodes. This will reduce infrastructure cost, enhance the reliability, and fasten the converging speed of algorithm. The dynamic model is established and parameter selection guideline are given in this paper. A low-voltage dc-microgrid prototype platform is utilized to validate the effectiveness of the proposed control.

Distributed energy management algorithms are being developed for the smart grid to efficiently and economically allocate electric power among connected distributed generation units and loads. The use of such algorithms provides flexibility, robustness, and scalability, while it also increases the vulnerability of smart grid to unexpected faults and adversaries. The potential consequences of compromising the power system can be devastating to public safety and economy. Thus, it is important to maintain the acceptable performance of distributed energy management algorithms in a smart grid environment under malicious cyber-attacks. In this paper, a neighborhood-watch-based distributed energy management algorithm is proposed to guarantee the accurate control computation in solving the economic dispatch problem in the presence of compromised generation units. The proposed method achieves the system resilience by performing a reliable distributed control without a central coordinator and allowing all the well-behaving generation units to reach the optimal operating point asymptotically. The effectiveness of the proposed method is demonstrated through case studies under several different adversary scenarios.

Day-ahead scheduling of generation units and storage devices is essential for the economic and efficient operation of a power system. Conventionally, a control center calculates the dispatch schedule by gathering information from all of the devices. However, this centralized control structure makes the system vulnerable to single point of failure and communication failures, and raises privacy concerns. In this paper, a fully distributed algorithm is proposed to find the optimal dispatch schedule for a smart grid with renewable and energy storage integration. The algorithm considers modified dc power flow constraints, branch energy losses, and energy storage charging and discharging efficiencies. In this algorithm, each bus of the system is modeled as an agent. By solely exchanging information with its neighbors, the optimal dispatch schedule of the conventional generators and energy storage can be achieved in an iterative manner. The effectiveness of the algorithm is demonstrated through several representative case studies.

The papers in this special section provide a timely discussion on technical trends and challenges of some classical and emerging issues, such as networked control and filtering, fault detection and tolerant control, event-triggered control, security control, distributed control, sensor networks, and real-time network protocol design, over networks with resource constraints for industrial systems.

With the increasing trend in adoption of plug-in hybrid and plug-in electric vehicles, they will play a prominent role in the future electric energy market by acting as responsive loads to increase the grid stability and facilitate the integration of renewables. However, due to the large number of controllable devices in the future grid, central vehicle to grid (V2G) management would be challenging and vulnerable to single points of failure. This paper introduces a novel distributed approach for optimal management of unidirectional V2G considering multiple energy suppliers. Each charging station as well as each energy supplier is equipped with a local price regulator to control the price paid to the energy suppliers and the price paid by the vehicles through coordination with their neighbors. In response to the updated prices, the vehicles adjust their charging rates and energy suppliers adjust their production to maximize their benefit. The main advantages of the proposed approach are that it manages unidirectional V2G in a fully distributed way considering multiple energy suppliers and vehicles, and it converges to the global optimum despite the greedy behavior of the individuals.

Battery Energy Storage Systems (BESS) are key components in microgrids for reliable operations. To ensure the safety and reliability of the BESS, accurate State-of-Charge (SOC) estimation is crucial. Most battery SOC estimation algorithms are developed assuming the measured load current and terminal voltage data are trust-worthy. However, in the real-world BESS applications, the measured battery data contains abnormal data which result in poor SOC estimation accuracy and will be harmful to the safety of the BESS and the stability of the microgrid. In this paper, a practical abnormal data filtering framework for real-world BESS applications is proposed. This framework real-time detects and filters abnormal data in real-world applications. Three typical abnormal data filters are investigated and demonstrated in this paper to illustrate the effectiveness of using data filtering for accurate SOC estimation in real-world BESS applications.

Smart sensors can be used in nuclear power plants to detect operational vibrations in order to predict plant maintenance schedules. The sensor design process includes characterization of the sensing element and determination of the optimal sensor location for nuclear applications. The selection of appropriate amplifiers, filters, and control elements are analyzed using the sensor’s specifications such as output amplitude, frequency, etc. A mock-up experiment was designed to test the performance of the vibration sensor and the sensitivity of the sensing element in real time. The experimental tests showed the system tank experiencing vibrations due to various water flow rates. This research is used to validate the working of the smart sensor and highlight its capability to perform detailed calculations, make appropriate decisions, and communicate these decisions to a remote monitoring system.

The microgrid is envisioned to be the building block of the future smart grid, for its abilities to host distributed energy resources, to improve grid reliability, and to enhance system resiliency. One of the most studied research topics of the microgrid is the distributed microgrid energy management system. However, the algorithm prototyping and hardware validation still remain great challenges at the current stage. This paper proposes a highly scalable, customizable, and low-cost DC microgrid testbed framework that enables fast distributed MG-EMS prototyping and provides proof-of-concept validation. Moreover, an application example is presented to demonstrate the functionality and validity of the proposed microgrid testbed.

Lithium-ion batteries are becoming the main energy storage in electric vehicles and electric grids. To elevate the battery capacity and the voltage supply, the battery cells are stacked to form a battery pack. The state-of charge (SOC) of the battery pack requires continuous monitoring for the operation safety. The current developed SOC estimation algorithms shows decent estimation accuracy but they are designed for individual cells. These algorithms stay in the battery cell level because they cannot capture the cell-to-cell difference which exists after manufactured. This paper proposed a battery pack SOC Co-Estimation algorithm based on the estimated battery cell SOC. The proposed battery pack SOC Co-Estimation algorithm can accurately estimates the SOC of a battery pack with three serial connected battery cells but without cell balancing. This algorithm also has the potential to reduce the computation effort on the battery management system (BMS) because it does not need to monitor every single cell in the battery pack.

Batteries as a multi-disciplinary field have been analyzed from the electrical, material science and electrochemical engineering perspectives. The first principle-based four-dimensional degradation model (4DM) of the battery is used in the article to connect the interdisciplinary sciences that deal with batteries. The 4DM is utilized to identify the physical manifestation that electrolyte degradation has on the battery and the response observed in the terminal voltage. This paper relates the different kinds of side reactions in the electrolyte and the material properties affected due to these side reactions. It goes on to explain the impact the material property changes has on the electrochemical reactions in the battery. This paper discusses how these electrochemical reactions affect the voltage across the terminals of the battery. We determine the relationship the change in the terminal voltage has due to the change in the design properties of the electrolyte. We also determine the impact the changes in the electrolyte material property have on the terminal voltage. In this paper, the lithium ion concentration and the transference number of the electrolyte are analyzed and the impact of their degradation is studied.

The complex configuration and behavior of multisource hybrid energy systems (HESs) present challenges to their energy management. For a balanced solution, it is especially important to represent and take advantage of the characteristics of each device and the interactive relationship among them. In this paper, multi-agent modeling and a game theory-based control strategy are proposed and combined for the energy management of an example engine–generator/battery/ultracapacitor (UC) HES. The three devices such as engine–generator unit, battery and UC packs are modeled and controlled as independent but related agents, through which the performance and requirements of the individual devices are fully respected. The energy management problem is then formulated as a noncooperative current control (NCC) game. The Nash equilibrium is analytically derived as a balanced solution that compromises the different preferences of the independent devices. The following simulation and experimental results validate the game theory-based control and its real-time implementation. The proposed approach could be further extended to become a general solution for the energy management and control of networked energy systems, in which again fully representing and balancing the different preferences of the components are important.

This paper proposes a novel cooperative charging strategy for a smart charging station in the dynamic electricity pricing environment, which helps electric vehicles (EVs) to economically accomplish the charging task by the given deadlines. This strategy allows EVs to share their battery-stored energy with each other under the coordination of an aggregator, so that more flexibility is given to the aggregator for better scheduling. Mathematically, the scheduling problem is formulated as a constrained mixed-integer linear program (MILP) to capture the discrete nature of the battery states, i.e., charging, idle and discharging. Then, an efficient algorithm is proposed to solve the MILP by means of dual decomposition and Benders decomposition. At last, the algorithm can be implemented in a distributed fashion, which makes it scalable and thus suitable for large-scale scheduling problems. Numerical results validate our theoretical analysis.

Optimal dispatch of storage devices is crucial for the economic operation of smart grids with distributed energy resources. Through appropriate scheduling, storage devices can store the energy when the renewable production is high or electricity price is low, and support the demand when electricity is expensive. Conventionally, this scheduling requires a control centre to gather information from the entire system and find the optimal schedule in the required horizon for the controllable devices. This study proposes a fully distributed scheduling methodology based on discrete-time optimal control, primal-dual gradient descent, and consensus networks. In the proposed approach, the requirement for the control centre is eliminated and the optimal schedule for all the devices is found solely through iterative coordination of each device with its neighbours. The application of the algorithm is demonstrated in a 5-bus system and its convergence to the global optimum is validated through Monte Carlo simulations. Further, it is shown that the algorithm is robust against communication link failures provided that the communications topology remains connected or reconnects after being disconnected.

Power electronic-based components (PECs) are at the heart of the enabling technologies for the smart-grids. They improve the controllability of the power system and provide excellent features such as load regulation, high power factor, and transient performance. However, they can behave as negative impedance due to their capability to operate as constant power loads, and jeopardize the stability of the power systems. Therefore, stability analysis of the power electronic-based distribution systems is crucial for development of the future smart-grids. This paper provides two methods to analyze stability: real-time (online) and offline. In real-time approach, the system's small-signal stability is investigated based on $d - q$ impedance measurement and unit circle criterion and by calculating source and load impedances simultaneously and in a range of frequencies. In offline approach, the system dynamics are identified from the bode plots and then based on the eigenvalue analysis, the stability of the system under different loading conditions is analyzed. Furthermore, small-signal stability of a solid state transformer (SST) as an advanced PEC with power factor correction is investigated via the proposed methods. In addition, hardware experiment is developed though power hardware-in-the-loop experiment to assess stability of an SST in load variation and validate the real-time capability of the proposed technique.

The purpose of this “Special Section on Networked Control and Industrial Applications” of the IEEE Transactions on Industrial Electronics is to present some up-to-date analysis and synthesis approaches to NCSs, which can improve the efficiency, flexibility, and reliability so that the installation, reconfiguration and maintenance time, and costs can be reduced through constrained communication network and distributed control architecture. This Special Section consists of 12 papers, which can be divided into two parts. The fist part, which contains five papers, mainly focuses on the analysis and design issues of classical NCSs, and the second part, which contains seven papers, addresses the networked multi-agent systems with applications to multirobot systems and power systems. A brief overview of the papers included in the Special Section is provided as follows.

With an increasing number of controllable distributed energy resources deployed and integrated into the power system, how to economically manage these distributed resources will become a challenge for the future smart grid. To solve the issue, consensus based distributed economic dispatch algorithms have been introduced in the literature as computationally scalable approaches. However, in real-world applications with imperfect communications networks, the performance of consensus-based economic dispatch algorithms degrades when information losses occur. In this paper, a robust distributed system incremental cost estimation (RICE) algorithm is introduced to solve the Economic Dispatch Problem (EDP) in a smart grid environment in a distributed way considering communications information losses. Unlike the existing consensus-based algorithms to solve EDP, RICE algorithm has two updating layers running in parallel in each distributed controller: one layer uses the gossip updating rule to estimate the system's average power mismatch, while the other layer uses the consensus updating rule to update the system Incremental Cost (IC) estimation. In this approach, the vulnerability of consensus-based algorithms to communications information losses is eliminated. The convergence and optimality of the algorithm are guaranteed as long as the undirected communications topology among local controllers is connected. Several case studies are presented to illustrate the performance of the proposed algorithm, and show the robustness under different information loss scenarios with different communications topologies. Display Omitted The classic Economic Dispatch Problem is solved in a distributed manner.The proposed RICE algorithm is robust to communication information losses.The proposed RICE algorithm is highly scalable.Rigorous analysis is provided to prove the convergency and optimality.