Energy Data Analytics

The energy industry is seeing an influx of data from new ICT and IOT interventions, leading to several problem statements seeking data-driven solutions. The problems span a wide gamut - from forecasting to pattern recognition, from anomaly detection to predictive maintenance and from cost minimization to efficiency optimization to name a few. This course can provide an introductory platform to the evolving world of energy informatics to interested students. Moreover, this course is ideally suited to introduce project/activity-based learning in the Electrical Engineering curriculum. It is an excellent sequel to the newly introduced AI and data science course EE 353, allowing students to apply the techniques learnt in that course to industry inspired problems. The course will also bring in industry professionals to provide insights on real world problems and also to evaluate the innovative solutions designed during the course. This way, the course is designed to produce well rounded energy data analysts much needed today. Module 1 (Introduction to energy data analytics and energy systems):Introduction to energy data analytics, focusing on the importance of data in theenergy sector. Overview of the key sources of energy data (smart meters,sensors, satellite data, etc.) and the typical tools and programming languages(Python, R, SQL, etc.) used for energy data analytics. Brief overview of energysystems, including generation, transmission, and distribution networks;renewable/conventional energy sources; consumption patterns and gridoperations.Module 2 (Data Processing and Exploratory Analysis): Data acquisition fromIoT devices, SCADA, and APIs. Data cleaning and preprocessing techniques(handling missing values and outliers). Time series analysis in energy data.Correlation between different energy variables. Pattern identification andanomaly detection. Typical machine learning applications in energy domain(demand forecasting, anomaly detection, consumer clustering, etc).Module 3 (Overview of various data-driven applications in energy domain):Forecasting energy demand and supply using time series techniques (ARIMA,SARIMA, LSTMs); Impact of weather on energy demand; Considerations inlong-term/medium-term/short-term forecasting.Optimization and decision support systems for energy efficiency optimization,demand-side management, building automation and control, energy costminimization and system control.Use of data for energy market analysis and policy studies, including electricitypricing models, carbon footprint analysis, energy transition studies.Overview of phasor measurement units and wide-area monitoring and control forpower grids.Role of IoT in smart energy systems (buildings cities, grids etc), big datatechniques for handling large-scale energy datasets, Cloud and edge computingapplications.Case studies and industry applications (examples include energy theft detection,power purchase cost optimization, power management in data centers)