Gradient-augmented Supervised Learning of Optimal Feedback Laws Using State-dependent Riccati Equations
This addresses computational bottlenecks in control systems for real-time applications, though it appears incremental as it builds on existing State-dependent Riccati Equation methods.
The paper tackles large-scale nonlinear stabilization by training a neural network to replace real-time Algebraic Riccati Equation solves, achieving this through a supervised learning approach with gradient-augmented loss.
A supervised learning approach for the solution of large-scale nonlinear stabilization problems is presented. A stabilizing feedback law is trained from a dataset generated from State-dependent Riccati Equation solves. The training phase is enriched by the use gradient information in the loss function, which is weighted through the use of hyperparameters. High-dimensional nonlinear stabilization tests demonstrate that real-time sequential large-scale Algebraic Riccati Equation solves can be substituted by a suitably trained feedforward neural network.