The Smart Grid group is conducting research into informatics-driven scalable software architectures to address realtime power management in the domain of Smart Power Grids. This work is funded by the US Department of Energy as part of the five-year Los Angeles Smart Grid Demonstration project, to forecast and curtail power consumption by thousands of electricity consumers on-demand through large scale information processing and consumer pattern detection. We apply big data analytics and advance machine learning techniques to this emerging application area of critical importance to global sustainability.


This cyberphysical domain provides unique challenges to many existing computer science algorithms, approaches and frameworks due to the data complexity, dynamism, scale and need for realtime response.

Areas of Interest:

semantic information integration, graph analytics, stream processing, data analytics and machine learning, cyber physical security, and cloud computing

Research

Data Analytics

In a Smart Microgrid, both the utility and the consumer can benefit from analytics based on data for electricity consumption and related variables. There are several benefits and costs associated with performing analytics at different spatial and temporal granularities.


In our research, we perform analytics for the USC campus microgrid to develop reliable electricity consumption forecasting models that would work for different spatial (building-level and campus-level) and temporal granularities (15-min and daily intervals). Our goal is to provide the facility managers more insight into the consumption patterns on the campus and enable them to plan consumption and curtailment activities for Demand Response optimization in the campus microgrid.


The USC campus microgrid is a test-bed for the DOE- sponsored Los Angeles Smart Grid Demonstration Project (SGDP) [26]. The USC campus has over 170 buildings and 45,000 student and staff population with diverse demographics. THe Facilities Management Services (FMS) on campus collects data on power usage at 15-minute intervals from 170 smart power meters on campus. The goal of the SGDP is to demonstrate DR optimization in the campus microgrid.


Optimal Customer Selection

The objective of Demand Response (DR) in a Smart Grid is to reduce the aggregated grid power consumption by a targeted amount (targeted curtailment) to maintain demand supply balance during periods of high consumption. A customer can adopt one of the several available strategies: turning down the air conditioner, dimming the lights, etc. Each strategy is associated with a curtailment value which is obtained by the utility by performing Data Analytics on customer consumption data. For a successful Demand Response, the utility should determine the customer strategy pairs such that the targeted curtailment is achieved.


The problem of optimal customer selection for a Demand Response is NP-hard. Hence, in our research, we develop ILP based approximation algorithms for the same. Driven by our experience with the DR program in the USC campus microgrid, we identify challenges and their solutions. In the past, we have focused on evenly distributing the curtailment across the entire DR events to smoothen any spikes in the grid consumption profile during the DR. Currently, our focus is on ensuring the customers are selected in a fair manner (a few customers are not penalized heavily for the overall curtailment).


The goal of this research is to develop a Demand-Response suite which can be customized to develop end-to-end workflow for a Demand Response implementation in a Smart Grid. The utility will just need to determine the right models for consumption and curtailment prediction and customer-strategy selection from the plethora of models available in the suite which are best suited for its Smart Grid. Currently, we are interfacing with LADWP to implement a Demand Response program in the city of Los Angeles which has around 1 million customers.


Complex Event Processing

Complex Event Processing for Dynamic Demand Response Optimization Existing DR programs are typically based on static plans that either use higher power pricing or load curtailment schedules at pre-determined peak hours. Dynamic pricing in real-time, when presents, works on a broadcast mode to all consumers with unpredictable outcomes. These static DR models are becoming insufficient given dynamic power use activities monitored in real time in Smart Grid. An integrated view of Smart Grid information allows us to use patterns of information coming from different sources to describe dynamic DR situations. These transient patterns can be used to locate outliers missed by coarse grained model and allow opportunities curtailment.


We propose to apply complex event processing (CEP) for dynamic demand response optimization. Complex event processing (CEP) deals with detecting situations represented as event patterns from a cloud of data streams. CEP has been successfully applied in many application domains ranging from supply chain management to financial services. The requirement of timely response to the energy use activities in dynamic DR makes complex event processing an attractive solution. We abstracts incoming data from the information integration layer as events. We developed a semantic CEP engine that consists of semantic annotation, filtering and matching modules to perform continuous queries over events and domain ontologies to correlate power grid events for detecting peak load occurrences and load curtailment opportunities.


The consequence of this is easier to identify the most relevant events from the distributed and heterogeneous data sources, analyze their impact, and take subsequent action in real time. Since the event data conforms to a formal semantics, domain experts will be able to encode their knowledge of meaningful event patterns and define responses to events using declarative languages.


Security in Cyber Physical Systems

Cyber physical systems are becoming increasingly complex with the advancement of internet infrastructure, internet of things (IoT), and sensor networks. Smart grids are an emerging class of such systems. It contains an advanced metering infrastructure (AMI) at customers and sub-stations, distributed sensor networks to maintain the situational awareness, communication network to communicate these sensor readings and distributed power generation, transmission and storage units. Real time state estimation in smart grid is one key problem when it comes to smart grid security. Complex interactions happen between the interconnected cyber layer and the physical layer of these systems demands advanced analytics.


The goal of this research is to model the Cyber Physical Systems using dynamic graphs and develop fast graph theoretical algorithms to detect complex cyber attacks and maintain situational awareness. A system for real time threat detection in cyber physical systems will be developed.

Cloud Computing

Click here to read more about Cloud Computing Projects.

Recent Publications

  • Sanmukh R. Kuppannagari, Rajgopal Kannan, Charalampos Chelmis and Viktor K. Prasanna, Implementation of Learning-Based Dynamic Demand Response on a Campus Micro-grid, 25th International Joint Conference on Artificial Intelligence, IJCAI-16 Demo Track.

  • Michail Misyrlis, Charalambos Chelmis, Kannan Rajgopal, Viktor K. Prasanna, Sparse Causal Temporal Modeling to Inform Power System Defense, to appear in Procedia Computer Science, 2016.

  • Sanmukh R. Kuppannagari, Rajgopal Kannan, Charalampos Chelmis, Arash S. Tehrani, and Viktor K. Prasanna Optimal Customer Targeting for Sustainable Demand Response in Smart Grids, International Conference on Computational Science, 2016

  • Ranjan Pal and Viktor Prasanna, The STREAM Mechanism to Improve CPS Security The Case of the Smart Grid, IEEE Transactions on Computer-Aided Design for Integrated Circuits and Systems, 2016

  • Fabian Knirsch, Dominik Engel, Cristian Neureiter, Marc Frincu and Viktor Prasanna, Privacy Assessment of Data Flow Graphs for an Advanced Recommender System in the Smart Grid, Communications in Computer and Information Science, Springer, Vol 576, 2016

  • Ranjan Pal, Charalampos Chelmis, Marc Frincu and Viktor Prasanna, MATCH for the Prosumer Smart Grid The Algorithmics of Real-Time Power Balance, IEEE Transactions on Parallel and Distributed Systems (TPDS), 2016

  • Saima Aman,Charalampos Chelmis and Viktor Prasanna, Learning to REDUCE: A Reduced Electricity Consumption Prediction Ensemble, AAAI Workshop on Artificial Intelligence for Smart Grids and Smart Buildings (AAAI '16), February 2016

  • Charith Wickramaarachchi, Alok Kumbhare, Charalampos Chelmis and Viktor Prasanna, Hatrick: A System for Real-time Threat Detection in Cyber Physical Systems, Annual Computer Security Applications Conference (ACSAC '15), December 2015

  • Sanmukh R. Kuppannagari, Rajgopal Kannan and Viktor K. Prasanna, An ILP Based Algorithm for Optimal Customer Selection for Demand Response in Smartgrids, International Conference on Computational Science and Computational Intelligence (CSCI '15), December 2015

  • Saima Aman, Marc Frincu, Charalampos Chelmis, Mohammad Noor and Viktor Prasanna, Prediction Models for Dynamic Demand Response: Requirements, Challenges, and Insights, IEEE Conference on Smart Grid Communications (SmartGridComm '15), November 2015

  • Yogesh Simmhan, Neel Choudhury, Charith Wickramaarachchi, Alok Kumbhare, Marc Frincu, Cauligi Raghavendra and Viktor Prasanna, Distributed Programming over Time-series Graphs, IEEE International Parallel and Distributed Processing Symposium (IPDPS), May 2015.

  • Charalampos Chelmis, Jahanvi Kolte and Viktor K. Prasanna, Big Data Analytics for Demand Response: Clustering Over Space and Time, IEEE International Conference on Big Data Workshops, Big Data 2015

  • Ranjan Pal, Charalampos Chelmis, Chandra Tadepalli, Marc Frincu, Saima Aman and Viktor Prasanna, Online Time Series Clustering for Demand Response: A Theory to Break the 'Curse of Dimensionality', ACM e-Energy, July 2015 (challenge paper).

  • Marc Frincu, Charalampos Chelmis, Saima Aman, Muhammad Rizwan Saeed, Vasilis Zois, Viktor Prasanna, Carol Fern and Aras Akbari, Enabling Automated Dynamic Demand Response: From Theory to Practice, ACM e-Energy, July 2015 (poster).

  • Charalampos Chelmis, Muhammad Rizwan Saeed, Marc Frincu and Viktor Prasanna, Curtailment Estimation Methods for Demand Response: Lessons Learned by Comparing Apples to Oranges, ACM e-Energy, July 2015 (poster).

  • S. Aman, Y. Simmhan and V. K. Prasanna, Holistic Measures for Evaluating Prediction Models in Smart Grids, IEEE Transactions on Knowledge and Data Engineering (TKDE), Vol. 27, No. 2, pp: 475-488, 2015.

  • Fabian Knirsch, Dominik Engel, Marc Frincu, Viktor Prasanna, Model-based Assessment for Balancing Privacy Requirements and Operational Capabilities in the Smart Grid, IEEE Conference on Innovative Smart Grid Technologies (iSGT), February 2015.

  • Fabian Knirsch, Dominik Engel, Christian Neureiter, Marc Frincu, Viktor Prasanna, Model-driven Privacy Assessment in the Smart Grid, International Conference on Information Systems Security and Privacy (ICISSP)(Best Paper Award), February 2015.

  • S. Aman, C. Chelmis and V. K. Prasanna, Influence-driven Model for Time Series Prediction from Partial Observations, AAAI Conference on Artificial Intelligence, Austin, TX, 2015.

  • C. Chelmis, S. Aman, M. Saeed, M. Frincu and V. K. Prasanna, Estimating Reduced Consumption for Dynamic Demand Response, AAAI Workshop on Computational Sustainability, Austin, Texas, 2015.

  • S. Aman, C. Chelmis, and V. K. Prasanna Addressing data veracity in big data applications, IEEE International Conference on Big Data, 2014.

  • V. Zois, M. Frincu, C. Chelmis, M. Saeed and V. K. Prasanna, Efficient Customer Selection for Sustainable Demand Response in Smart Grids, International Green Computing Conference, Dallas, Texas, 2014.

  • V. Zois, M. Frincu and V. K. Prasanna, Integrated Platform for Automated Sustainable Demand Response in Smart Grids, International Workshop on Intelligent Energy Systems, San Diego, California, 2014.

  • Charith Wickramaarachchi, Marc Frincu, Viktor Prasanna, Enabling Real-time Pro-active Analytics on Streaming Graphs Technical Report, USC, 2014.

  • Qunzhi Zhou, Yogesh Simmhan and Viktor Prasanna Towards Hybrid Online On-Demand Querying of Realtime Data with Stateful Complex Event Processing, IEEE International Conference on Big Data (BigData), 2013.

  • Charith Wickramaarachchi and Yogesh Simmhan, Continuous Dataflow Update Strategies for Mission-Critical Applications IEEE Internatrional Conference on eScience (eScience), 2013.

  • Yogesh Simmhan and Muhammad Usman Noor, Scalable Prediction of Energy Consumption using Incremental Time Series Clustering, Workshop on Big Data and Smarter Cities,2013.

  • Yogesh Simmhan, Saima Aman, Alok Kumbhare, Rongyang Liu, Sam Stevens, Qunzhi Zhou and Viktor Prasanna, Cloud-Based Software Platform for Big Data Analytics in Smart Grids, Computing in Science and Engineering , 15(4), pp. 38-47, July-Aug. 2013, IEEE and AIP ([IF 1.422, CORE C]).

  • Saima Aman, Yogesh Simmhan and Viktor K. Prasanna, Energy Management Systems: State of the Art and Emerging Trends, IEEE Communications Magazine , 51 (1) , 2013 , pp. 114 - 119 , IEEE. ([IF 3.785]).

  • Alok Kumbhare, Yogesh Simmhan and Viktor Prasanna, Cryptonite: A Secure and Performant Data Repository on Public Clouds, International Cloud Computing Conference (CLOUD), 2012.

  • Yogesh Simmhan, Vaibhav Agarwal, Saima Aman, Alok Kumbhare, Sreedhar Natarajan, Nikhil Rajguru, Ian Robinson, Samuel Stevens, Wei Yin, Qunzhi Zhou and Viktor Prasanna, Adaptive Energy Forecasting and Information Diffusion for Smart Power Grids, IEEE International Scalable Computing Challenge (SCALE), 2012 (First Prize).

  • Wei Yin, Yogesh Simmhan and Viktor Prasanna, Scalable Regression Tree Learning on Hadoop using OpenPlanet, International Workshop on MapReduce and its Applications (MAPREDUCE), 2012.

  • Qunzhi Zhou, Yogesh Simmhan and Viktor Prasanna, Incorporating Semantic Knowledge into Dynamic Data Processing for Smart Power Grids, International Semantic Web Conference (ISWC), 2012.

  • Qunzhi Zhou, Sreedhar Natarajan, Yogesh Simmhany and Viktor Prasanna, Semantic Information Modeling for Emerging Applications in Smart Grid, International Conference on Information Technology : New Generations (ITNG), 2012.

  • Qunzhi Zhou, Yogesh Simmhan and Viktor Prasanna, Semantic Complex Event Processing over End-to-End Data Flows, 2012 , Computer Science Department, University of Southern California.

  • Yogesh Simmhan, Baohua Cao, Michail Giakkoupis and Viktor K. Prasanna, Adaptive rate stream processing for smart grid applications on clouds, International Workshop on Scientific Cloud Computing (ScienceCloud) , 2011 , pp. 33-38 , ACM.

  • Saima Aman, Yogesh Simmhan and Viktor K. Prasanna, Improving Energy Use Forecast for Campus Micro-grids using Indirect Indicators, International Workshop on Domain Driven Data Mining (DDDM) , 2011.

  • Saima Aman, Yogesh Simmhan and Viktor K. Prasanna, Smart Communication of Energy Use and Prediction in a Smart Grid Software Architecture, IEEE Coastal Los Angeles Section. (Poster).

  • Saima Aman, Wei Yin, Yogesh Simmhan and Viktor Prasanna, Machine Learning for Demand Forecasting in Smart Grid, Southern California Smart Grid Research Symposium (SoCalSGS) , 2011. (Poster).

  • Yogesh Simmhan, Alok Kumbhare, Baohua Cao and Viktor K. Prasanna, An Analysis of Security and Privacy Issues in Smart Grid Software Architectures on Clouds International Cloud Computing Conference (CLOUD), 2011, IEEE.

  • Alok Kumbhare, Yogesh Simmhan and Viktor Prasanna, Designing a Secure Storage Repository for Sharing Scientific Datasets using Public Clouds, International Workshop on Data Intensive Computing in the Clouds (DataCloud-SC11), 2011.

  • Yogesh Simmhan, Saima Aman, Baohua Cao, Michail Giakkoupis, Alok Kumbhare, Qunzhi Zhou, Karthik Gomadam and Viktor K. Prasanna, Scalable, Secure Energy Information Management for DR Analysis, Southern California Smart Grid Research Symposium (SoCalSGS), 2010 , University of Southern California. (Poster).

  • Yogesh Simmhan, Saima Aman, Baohua Cao, Mike Giakkoupis, Alok Kumbhare, Qunzhi Zhou, Donald Paul, Carol Fern, Aditya Sharma and Viktor K. Prasanna, An Informatics Approach to Demand Response Optimization in Smart Grids, 2011, University of Southern California.

  • Yogesh Simmhan, Viktor Prasanna, Saima Aman, Sreedhar Natarajan, Wei Yin and Qunzhi Zhou, Towards Data-driven Demand-Response Optimization in a Campus Microgrid, Workshop On Embedded Sensing Systems For Energy-Efficiency In Buildings (BuildSys) , 2011 , ACM. (Demo).

  • Yogesh Simmhan, Qunzhi Zhou and Viktor K. Prasanna, Chapter: Semantic Information Integration for Smart Grid Applications, 2011 , pp. 361-380 , Springer Berlin Heidelberg.

  • Qunzhi Zhou, Sreedhar Natarajan, Yogesh Simmhan and Viktor Prasanna, Semantic Information Integration and Processing for Demand Response Optimization, Southern California Smart Grid Research Symposium (SoCalSGS), 2011 (Poster).

  • Qunzhi Zhou, Yogesh Simmhan and Viktor K. Prasanna, Towards an inexact semantic complex event processing framework, International Conference on Distributed Event-Based System (DEBS), 2011 , pp. 401-402 , ACM. (Poster).

  • Yogesh Simmhan, Michail Giakkoupis, Baohua Cao and Viktor K. Prasanna, On Using Cloud Platforms in a Software Architecture for Smart Energy Grids, International Conference on Cloud Computing Technology and Science (CloudCom), 2010 , IEEE. (Poster).

  • Qunzhi Zhou, Yogesh Simmhan and Viktor K. Prasanna, Semantic Complex Event Processing for Smart Grid Information Integration and Management, Energy Forum: A System Approach Toward Green Energy Production and Adaptive Power Distribution, 2010 , IEEE Coastal Los Angeles Section. (Poster).