In this webinar we will use machine learning in MATLAB and physical modeling in Simulink to demonstrate predictive maintenance concepts. Using data from a real world example, we will explore importing, pre-processing, and labeling data, as well as selecting features, and training and comparing multiple machine learning models. We will then consider a grid connected generation plant to demonstrate how a physical model of equipment can be used to complement machine learning techniques, by providing a platform to generate fault data for machine learning methods, and as an additional paradigm for monitoring system degradation.
Webinar Highlights:
- Developing larger scale physical simulations
- Using Big Data with simulations
- Estimating model parameters from measured data
- Prototyping, testing, and refining predictive models using machine learning methods
- Combining physical modeling and machine learning techniques for predictive maintenance
About the Presenter:
Graham Dudgeon, PhD Principal Industry Manager – Utilities & Energy MathWorks, Inc. Graham is Principal Industry Manager for Energy at MathWorks, and works closely with the Electric Power and Chemical & Petroleum industries worldwide. Before his role as industry manager, Graham was a Principal Technical Consultant at MathWorks and worked with a broad range of customers in the Electric Machinery, Aerospace, Defence, Automotive, Transport and Medical industries. Graham’s technical experience and expertise includes; electric grid simulation (transmission and distribution), renewable energy simulation (wind farm and solar farm operation and grid integration), control system design and analysis, data analytics, and power electronics.
