Enhancing generator load ramping & flexibility using dynamic AI and closed-loop optimization

Feb 21, 2023
Booth #2636, Exhibit Hall
Knowledge Hub- Digitalization and Optimization

As renewables put an increasing amount of power on the grid, there is a new value proposition for combustible fuel power generation: flexibility. 

Neural network optimization (D-NNO) presents a novel way to optimize power plants under ramping conditions. Using deep learning methods with recurrent neural networks--such as long short-term memory, gated recurrent units, and deep belief networks--the plant’s transient behavior can be captured in a machine-learning model that can then be used for online optimization. 

This session presents methodologies for both the machine learning model development and optimization strategy required for successful D-NNO implementation.

Jacob Tuttle, CTO - Taber International, LLC.
Jason Parrish, Performance Engineer - PacifiCorp