Reliably-stabilizing piecewise-affine neural network controllers
arXiv:2111.07183 · doi:10.1109/TAC.2022.3216978
Abstract
A common problem affecting neural network (NN) approximations of model predictive control (MPC) policies is the lack of analytical tools to assess the stability of the closed-loop system under the action of the NN-based controller. We present a general procedure to quantify the performance of such a controller, or to design minimum complexity NNs with rectified linear units (ReLUs) that preserve the desirable properties of a given MPC scheme. By quantifying the approximation error between NN-based and MPC-based state-to-input mappings, we first establish suitable conditions involving two key quantities, the worst-case error and the Lipschitz constant, guaranteeing the stability of the closed-loop system. We then develop an offline, mixed-integer optimization-based method to compute those quantities exactly. Together these techniques provide conditions sufficient to certify the stability and performance of a ReLU-based approximation of an MPC control law.
References in corpus (11)
- On the Number of Linear Regions of Deep Neural Networks
- Safe, Efficient, and Comfortable Velocity Control based on Reinforcement Learning for Autonomous Driving
- Fast inference of deep neural networks in FPGAs for particle physics
- Efficient representation and approximation of model predictive control laws via deep learning
- Real-time Power System State Estimation and Forecasting via Deep Neural Networks
- Learning an Approximate Model Predictive Controller with Guarantees
- Safe and Fast Tracking on a Robot Manipulator: Robust MPC and Neural Network Control
- A Unified View of Piecewise Linear Neural Network Verification
- Deep ReLU Networks Have Surprisingly Few Activation Patterns
- Piecewise Linear Regression via a Difference of Convex Functions
- Polytopic Input Constraints in Learning-Based Optimal Control Using Neural Networks