Liyan Jia

Malicious Data Attacks on the Smart Grid

Malicious attacks against power systems are investigated, in which an adversary controls a set of meters and is able to alter the measurements from those meters. Two regimes of attacks are considered. The strong attack regime is where the adversary attacks a sufficient number of meters so that the network state becomes unobservable by the control center. For attacks in this regime, the smallest set of attacked meters capable of causing network unobservability is characterized using a graph theoretic approach. By casting the problem as one of minimizing a supermodular graph functional, the problem of identifying the smallest set of vulnerable meters is shown to have polynomial complexity. For the weak attack regime where the adversary controls only a small number of meters, the problem is examined from a decision theoretic perspective for both the control center and the adversary. For the control center, a generalized likelihood ratio detector is proposed that incorporates historical data. For the adversary, the trade-off between maximizing estimation error at the control center and minimizing detection probability of the launched attack is examined. An optimal attack based on minimum energy leakage is proposed.

Malicious Data Attacks on Smart Grid State Estimation: Attack Strategies and Countermeasures

The problem of constructing malicious data attack of smart grid state estimation is considered together with countermeasures that detect the presence of such attacks. For the adversary, using a graph theoretic approach, an efficient algorithm with polynomial-time complexity is obtained to find the minimum size unobservable malicious data attacks. When the unobservable attack does not exist due to restrictions of meter access, attacks are constructed to minimize the residue energy of attack while guaranteeing a certain level of increase of mean square error. For the control center, a computationally efficient algorithm is derived to detect and localize attacks using the generalized likelihood ratio test regularized by an L_1 norm penalty on the strength of attack.

Limiting false data attacks on power system state estimation

Malicious attacks against power system state estimation are considered. It has been recently observed that if an adversary is able to manipulate the measurements taken at several meters in a power system, it can sometimes change the state estimate at the control center in a way that will never be detected by classical bad data detectors. However, in cases when the adversary is not able to perform this attack, it was not clear what attacks might look like. An easily computable heuristic is developed to find bad adversarial attacks in all cases. This heuristic recovers the undetectable attacks, but it will also find the most damaging attack in all cases. In addition, a Bayesian formulation of the bad data problem is introduced, which captures the prior information that a control center has about the likely state of the power system. This formulation softens the impact of undetectable attacks. Finally, a new L∞ norm detector is introduced, and it is demonstrated that it outperforms more standard L2 norm based detectors by taking advantage of the inherent sparsity of the false data injection.

Modeling and Stochastic Control for Home Energy Management

The problem of modeling and stochastic optimization for home energy management is considered. Several different types of load classes are discussed, including heating, ventilation, and air conditioning unit, plug-in hybrid electric vehicle, and deferrable loads such as washer and dryer. A first-order thermal dynamic model is extracted and validated using real measurements collected over an eight months time span. A mixed integer multi-time scale stochastic optimization is formulated for the scheduling of loads of different characteristics. A model predictive control based heuristic is proposed. Numerical simulations coupled with real data measurements are used for performance evaluation and comparison studies.

Impact of Data Quality on Real-Time Locational Marginal Price

The problem of characterizing impacts of data quality on real-time locational marginal price (LMP) is considered. Because the real-time LMP is computed from the estimated network topology and system state, bad data that cause errors in topology processing and state estimation affect real-time LMP. It is shown that the power system state space is partitioned into price regions of convex polytopes. Under different bad data models, the worst case impacts of bad data on real-time LMP are analyzed. Numerical simulations are used to illustrate worst case performance for IEEE-14 and IEEE-118 networks.

Malicious data attack on real-time electricity market

Malicious data attacks to the real-time electricity market are studied. In particular, an adversary launches an attack by manipulating data from a set of meters with the goal of influencing revenues of a real-time market. The adversary must deal with the tradeoff between avoiding being detected by the control center and making maximum profit from the real time market. Optimal attacking strategy is obtained through an optimization of a quasi-concave objective function. It is shown that the probability of detection of optimal attack will always be less than 0.5. Attack performance is evaluated using simulations on the IEEE 14-bus system.

Modeling and stochastic control for Home Energy Management

The problem of modeling and control for Home Energy Management (HEM) is considered. A first order thermal dynamic model is considered and its parameters are extracted using real measurements over a period of three summer months. The identified model is validated using separate data sets. The extracted model shows certain nonstationarity and non-Gaussianity. However, local approximations using a stationary model are shown to have relatively small modeling and prediction errors. The extracted model is then used for developing a multi-scale multi-stage stochastic optimization framework for the control of the Heating, Ventilation, and Air Conditioning (HVAC) unit, the charging of Plug-in Hybrid Electric Vehicle (PHEV), and the scheduling of deferrable load such as washer/dryer operations. A two time scale Model Predictive Control (MPC) strategy is proposed that minimizes the discomfort level subject to power and budget constraints: at the slow time scale, a power budget is allocated across different appliances at the hourly level; at the fast time scale, sensor measurements are used for the scheduling and control of different loads. Using parameters extracted from the real data, the proposed approach is compared with the simple rule based control strategy typically used in HVAC controllers.

Dynamic Pricing and Distributed Energy Management for Demand Response

The problem of dynamic pricing of electricity in a retail market is considered. A Stackelberg game is used to model interactions between a retailer and its customers; the retailer sets the day-ahead hourly price of electricity and consumers adjust real-time consumptions to maximize individual consumer surplus. For thermostatic demands, the optimal aggregated demand is shown to be an affine function of the day-ahead hourly price. A complete characterization of the tradeoffs between consumer surplus and retail profit is obtained. The Pareto front of achievable tradeoffs is shown to be concave, and each point on the Pareto front is achieved by an optimal day-ahead hourly price. Effects of integrating renewables and local storage are analyzed. It is shown that benefits of renewable integration all go to the retailer when the capacity of renewable is relatively small. As the capacity increases beyond a certain threshold, the benefit from renewable that goes to consumers increases.

Impacts of Malicious Data on Real-Time Price of Electricity Market Operations

Impacts of malicious data data attack on the real-time electricity market are studied. It is assumed that an adversary has access to a limited number of meters and has the ability to construct data attack based on what it observes. Different observation models are considered. A geometric framework is introduced based on which upper and lower bounds on the optimal data attack are obtained and evaluated in simulations.

On the nonlinearity effects on malicious data attack on power system

There has been a growing literature on the malicious data attack (or data injection attack) on power systems. Most existing work focuses on the DC (linear) model with linear state estimators. This paper examines the effects of nonlinearity in the power systems on the effectiveness of malicious data attack on state estimation and real-time market. It is demonstrated that attack algorithms designed for the DC model may not be effective when they are applied to nonlinear system with nonlinear state estimators. Discussion and experiments results about nonlinearity are provided.