- Computer Science
- Geology & Geological
- Institute for Energy Studies
- Prospective Students
- Current Students
My research interest broadly lies in large scale Optimization and Smart Grid areas that relates fields of electrical engineering and computer science. My specific research interests and topics range from high performance computing, Operations Research, Smart Grid Modeling and Simulation, Big Data/green computing and distributed control.
I would like to explain the remaining research statement based on my current work under four main research focuses:
0. Data Analytics Research
Data analytics is a science that encompasses data mining, machine learning and statistical methods, and which focuses on cleaning, transforming, modeling and extracting actionable information from large, complex data sets. A smart grid generates a large amount of data from its various components, examples of which include renewable energy generators and smart meters; the potential value of this data is huge but exploiting this value will be almost impossible without the use of proper analytics. With the application of systematic analytics on the smart grid’s data, its goals of better economy, efficiency, reliability, and security can be achieved. A further consequence of this process is the steady growth in the complexity and connectedness of critical energy infrastructure. This trend, coupled with the rapid growth in computing power and an increasingly diverse threat landscape, has led to pressing concerns about the vulnerability of these installations to cybersecurity attacks from a range of state and non-state actors. It seems certain that intelligent algorithms and data analytics will be an important part of the solution if these problems are to be effectively countered. My research group is intersted in developing algorithms for next generation smart grid data sets.
1. Smart Grid and Synchrophasors (Big Data) Research
There is a growing concern and research issues on how electricity is transported from generator to end user in the current electric grid system which provides an insecure, unreliable and inefficient (only 43% efficient!) energy to its consumers with remaining 57% lost during transmission. Electric power systems are one of the Nation's 8 critical infrastructures ("Critical Infrastructure Protection" by the US President's Commission). Yet, much of the electricity supply and delivery infrastructure is nearing the end of its useful life ("2009 Electricity adequacy report" by the Electricity Advisory Committee, US DOE). The Energy Independence and Security Act of 2007, as well as the American Recovery and Reinvestment Act of 2009 officially recognize the need to modernize the grid.
I am interested to research the issues in the current electric grid and problems associated with it. My research will be dedicated to building a reliable, intelligent, sustainable and safe power grid of tomorrow, the Smart Grid. I am particularly interested to focus on the development of analytical and computational tools for the modeling, analysis, design and optimization of large-scale integrated systems such as power, control, and communication systems.
Specifically, one of the key sensing and measurement units in Smart Grid deployment for transmission lines is the use of Synchrophasors. Synchrophasor technology can help deliver better real-time tools that enhance system operators' situational awareness. A synchrophasor system with wide deployment of phasor measurement units and dedicated high-speed communications to collect and deliver synchronized high-speed grid condition data, along with analytics and other advanced on-line dynamic security assessment and control applications--will improve real-time situational awareness and decision support tools to enhance system reliability. This advanced grid monitoring technology enhances reliability by increasing the ability of grid operators to collect and analyze data on system conditions over a wide area at a much higher frequency. One of the problems in such measurement units is ability to interpret and disregard bad data from enormous samples.
PDC - Phasor Data Concentrator; PMU - Phasor Measurement Unit
DFR - Digital Fault Recorder; SCADA - Supervisory Control and Data Acquisition
The electric utilities are currently struggling to identify ways to mine and make the best use of such growing PMU data sets. This problem has been a stumbling block for utilities to solve.
One way to study the issues is through study current and developing new Simulation models with collaboration from Utility Industries.
2. Cyber Security Research and Attack Resilient Algorithms in U.S. Electric Grid
The United States has embarked on a major transformation of its electric power infrastructure, sometimes referred as the world's largest interconnected complex machine. The system is undergoing tremendous change that will unfold over a number of years. As the electric grid is modernized, it will become highly automated, leverage information technology more fully, and become more capable in managing energy from a variety of distributed sources. However, in this process of becoming increasingly "smarter," the grid will expand to contain more complex interconnections that may become portals for intrusions, error-caused disruptions, malicious attacks such as (Denial of Service, Sybil attacks), and other threats. Primary among them is devising effective strategies for securing the computing and communication networks that will be central to the performance and availability of the envisioned electric power infrastructure and for protecting the privacy of Smart Grid-related data.
3. Role of Software Engineering Research in Self-healing Smart Grid – Multi-Agent Systems, Requirements Modeling, Distributed Systems
Goal 1: Application of the methodology to the Smart Grid and adaptation to specific challenges towards a self-healing grid that accommodates generation options:
A direct application will be the protection of a micro grid, which is a local energy network that can operate in parallel with the grid or in a disconnected mode to provide a customized level of high reliability and resilience to grid disturbances. The main challenge that will be faced in protection will be the bi-directional power flow. I aim to develop adapted advanced protection schemes, local supervisory controls and algorithms. Another challenge related to a micro grid is how to decide on the boundaries of an island to disconnect from the grid in case of disturbances or power outages, in other words intentional islanding. I aim to adapt my study on fault tolerant topology design, combined with optimization methods to develop advanced algorithms to guide such a decision. I will apply my expertise in intelligent sensor placement to solve the optimization problem in minimizing the cost function and power losses in the electric grid.
Goal 2: Agent based Simulation Models and High performance computing environment
My other area of interest is Simulation models, computations, and Agent based learning. Agent-based systems technology has been hailed as a new paradigm for conceptualizing, designing, and implementing software systems. I am interested to research the issues in our current grid through a Multi Agent System (MAS) simulation model and approaches which could be developed in Java Agent Development Framework (JADE) framework. Here, I treat agents are sophisticated computer programs that can act autonomously on behalf of their users, across open and distributed environments, to solve a growing number of complex problems in the grid architecture. I plan to address how Agent-based approaches allow energy management to be automated by matching energy supply and demand at a marketplace, through Commercial trade optimization and Technical grid optimization. Adding, my research will support agent oriented architecture environment which utilizes many green renewable Distributed energy resources (DER) such as Wind, Solar, Biomass etc., in the event of power outage or any shortage of energy demand in our grid architecture. One of the recent works that I contributed on is a Smart Grid Simulation which uses four types of agents namely Device Agent, Control Agent, DER agent, and User agent to simulate and self-heal a grid environment and it has been presented to a 2010 IEEE Smart Grid Communication Conference.
Goal 3: Decision support mechanisms
Smart Grid should employ effective 'optimal' integrated decision agents who need to "know" when there is the need to quickly reduce load or redirect power and respond autonomously to adverse conditions. Optimal search models based on heuristic or probabilistic meta-heuristic algorithms such as Distribution Tree Problem (DTP), Minimum Spanning Tree(MST), Simulated Annealing (SIMA) and Tabu search (TS) techniques need to be analyzed, investigated or improved in their applicability to Smart Grid environment by the following ways: employing an objective function which minimizes the cost function of adding or eliminating a transmission line relating with linear power loss consideration; and selecting optimal alternative routing paths in the event of energy demand or outage etc., or to reduce the operating cost. The Smart Grid will also be able to "call for help," enlisting support from distributed energy resources to help balance system needs. Agents that receive real-time price feeds and other data from the utilities, should have a basic set of knowledge-based rules on control decisions, and makes the control decisions that need to be executed and integrated together in the environment to self-heal the grid in any uncertainty. I plan to focus my research on the above said areas in the immediate years to come.
4. Wireless Sensor network Research
My research interests cover a wide variety of topics in the intersection of sensor networks such as time synchronization, data aggregation, databases, and distributed systems. The distinguishing aspect of my research is that I seek efficient and theoretically sound techniques to qualitatively enhance the robustness of large-scale distributed systems and I validate the techniques by implementing them in real systems. The availability of low cost sensing devices and the ubiquity of network connectivity provide the opportunity to build Internet-scale sensing services on the information derived from live sensor feeds. An example of such services is a Grid Monitoring service that uses cameras and other sensors deployed in distributed and transmission network regions in order to study interesting events (e.g., oscillations, frequency and voltage deviations, phasor angle deviations).
Unfortunately, there are currently no suitable generic software tools to address different aspects of building a sensing service: sensor feed processing, distributed query processing, service deployment, load balancing, fault tolerance, etc. This made authoring and deploying sensing services an onerous task, as each service author needed to address all the above mentioned aspects.