Big Data Application for Power Systems

This course is one in a series of several courses developed to address the evolving forces that will alter the fundamental operating characteristics of the electric grid, transforming it from a one-way central supply structure to one that has bidirectional power flows resulting from distributed energy resources (DER). Large numbers of self-generating consumers will alter the design requirements for the electric distribution system. New analytics techniques will be required to analyze, plan and operate the new grid. This 3-day course focuses on applications of big data analytics on smart electric power distribution systems and the use of Large Scale (Big) Data Analytical methods and their application to electric distribution system analysis and design. The basics of big data analytics and the electric power distribution system will be introduced. Four data-driven applications in electric power distribution systems will be studied closely. These include:

1. Distribution system topology identification using smart meters data;
2. Anomaly detection in power distribution systems;
3. Load and demand response forecasting; and
4. Predictive maintenance of transformers. This is a first course in developing techniques to analyze the distribution system in a Big Data format.

The course is intended for anyone interested in understanding how predictive analytics can be applied to electric power distribution systems. Students will include utility engineers and technicians, business & strategy staff, regulatory compliance staff, legal staff, and possibly regulators. Previous technical training is helpful but not necessary.

Topic 1: Brief Introduction of Data-Driven Analytics in Smart Electric Power Distribution Systems

• 1.1 Introduction to data-driven analytics
• 1.2 Data Volume, Variety, Velocity, and Value
• 1.3 Applications of big data analytics in smart electric power distribution systems

Topic 2: Singular Value Decomposition (SVD)

• 2.1 Application of SVD in Power System Anomaly Detection
• 2.2 Application of SVD in Bad Data Processing for State Estimation

Topic 3: Data-driven Application One: Distribution System Topology Identification

• 3.1 Introduction to machine learning algorithms
• 3.2 Unsupervised machine learning algorithm for phase identification
• 3.3 Supervised machine learning algorithm for phase identification
• 3.4 Transformer to customer association identification
• 3.5 Case study: Phase connectivity identification

Topic 4 Data-driven Application Two: Electricity Theft Detection

• 4.1 Development of modified linear model for distribution secondary
• 4.2 Physically inspired electricity theft detection method
• 4.3 Anomaly score and visualization of electricity theft

Topic 5 Data-driven Application Three: Evaluating the effectiveness of CVR

• 5.1 Introduction to robust regression
• 5.2 Quantify the load and voltage impacts of Conservation Voltage Reduction

Topic 6: Demand Response Forecast

• 6.1 Introduction to the mixture model
• 6.2 Forecast the load impact from demand response resources

Delegates who attend and complete the course would be issued a Certificate of Completion by Camaval International.

For further enquires kindly contact

Course Coordinator via email info@camaval.com

Register Print This Page Back to Course Page