With new advances in machine learning and in particular powerful learning libraries, we illustrate some of the new possibilities they enable in terms of nonlinear system identification. For a large class of hybrid systems, we explain how these tools allow for identification of complex dynamics using neural networks. We illustrate the method by examining the performance on a quad-rotor example.
Paper
References (14)
10Neural ordinary differential equations, in 'Advances in neural information processing systems2018
11Neural ordinary differential equations, in ‘Advances in neural information processing2018
Scroll for more · 2 remaining