There has been an increasing interest in using neural networks in closed-loop control systems to improve performance and reduce computational costs for on-line implementation. However, providing safety and stability guarantees for these systems is challenging due to the nonlinear and compositional structure of neural networks. In this paper, we propose a novel forward reachability analysis method for the safety verification of linear time-varying systems with neural networks in feedback interconnection. Our technical approach relies on abstracting the nonlinear activation functions by quadratic constraints, which leads to an outer-approximation of forward reachable sets of the closed-loop system. We show that we can compute these approximate reachable sets using semidefinite programming. We illustrate our method in a quadrotor example, in which we first approximate a nonlinear model predictive controller via a deep neural network and then apply our analysis tool to certify finite-time reachability and constraint satisfaction of the closed-loop system.
more »
« less
This content will become publicly available on June 1, 2025
Neural networks in feedback for flow analysis and control
This work presents a methodology for analysis and control of nonlinear fluid systems using neural networks. The approach is demonstrated in four different study cases: the Lorenz system, a modified version of the Kuramoto-Sivashinsky equation, a streamwise-periodic two-dimensional channel flow, and a confined cylinder flow. Neural networks are trained as models to capture the complex system dynamics and estimate equilibrium points through a Newton method, enabled by back-propagation. These neural network surrogate models (NNSMs) are leveraged to train a second neural network, which is designed to act as a stabilizing closed-loop controller. The training process employs a recurrent approach, whereby the NNSM and the neural network controller are chained in closed loop along a finite time horizon. By cycling through phases of combined random open-loop actuation and closed-loop control, an iterative training process is introduced to overcome the lack of data near equilibrium points. This approach improves the accuracy of the models in the most critical region for achieving stabilization. Through the use of L1 regularization within loss functions, the NNSMs can also guide optimal sensor placement, reducing the number of sensors from an initial candidate set. The data sets produced during the iterative training process are also leveraged for conducting a linear stability analysis through a modified dynamic mode decomposition approach. The results demonstrate the effectiveness of computationally inexpensive neural networks in modeling, controlling, and enabling stability analysis of nonlinear systems, providing insights into the system behavior and offering potential for stabilization of complex fluid systems.
more »
« less
- Award ID(s):
- 2238770
- PAR ID:
- 10557735
- Publisher / Repository:
- Physical Review Fluids
- Date Published:
- Journal Name:
- Physical Review Fluids
- Volume:
- 9
- Issue:
- 6
- ISSN:
- 2469-990X
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
We present an approach for the synthesis and verification of neural network controllers for closed loop dynamical systems, modelled as an ordinary differential equation. Feedforward neural networks are ubiquitous when it comes to approximating functions, especially in the machine learning literature. The proposed verification technique tries to construct an over-approximation of the system trajectories using a combination of tools, such as, Sherlock and Flow*. In addition to computing reach sets, we incorporate counter examples or bad traces into the synthesis phase of the controller as well. We go back and forth between verification and counter example generation until the system outputs a fully verified controller, or the training fails to terminate in a neural network which is compliant with the desired specifications. We demonstrate the effectiveness of our approach over a suite of benchmarks ranging from 2 to 17 variables.more » « less
-
Neural networks have become increasingly effective at many difficult machine learning tasks. However, the nonlinear and large-scale nature of neural networks makes them hard to analyze, and, therefore, they are mostly used as blackbox models without formal guarantees. This issue becomes even more complicated when neural networks are used in learning-enabled closed-loop systems, where a small perturbation can substantially impact the system being controlled. Therefore, it is of utmost importance to develop tools that can provide useful certificates of stability, safety, and robustness for neural network-driven systems.In this overview, we present a convex optimization framework for the analysis of neural networks. The main idea is to abstract hard-to-analyze components of a neural network (e.g., the nonlinear activation functions) with the formalism of quadratic constraints. This abstraction allows us to reason about various properties of neural networks (safety, robustness, generalization, stability in closed-loop settings, etc.) via semidefinite programming.more » « less
-
Neural networks present a useful framework for learning complex dynamics, and are increasingly being considered as components to closed loop predictive control algorithms. However, if they are to be utilized in such safety-critical advisory settings, they must be provably "conformant" to the governing scientific (biological, chemical, physical) laws which underlie the modeled process. Unfortunately, this is not easily guaranteed as neural network models are prone to learn patterns which are artifacts of the conditions under which the training data is collected, which may not necessarily conform to underlying physiological laws. In this work, we utilize a formal range-propagation based approach for checking whether neural network models for predicting future blood glucose levels of individuals with type-1 diabetes are monotonic in terms of their insulin inputs. These networks are increasingly part of closed loop predictive control algorithms for "artificial pancreas" devices which automate control of insulin delivery for individuals with type-1 diabetes. Our approach considers a key property that blood glucose levels must be monotonically decreasing with increasing insulin inputs to the model. Multiple representative neural network models for blood glucose prediction are trained and tested on real patient data, and conformance is tested through our verification approach. We observe that standard approaches to training networks result in models which violate the core relationship between insulin inputs and glucose levels, despite having high prediction accuracy. We propose an approach that can learn conformant models without much loss in accuracy.more » « less
-
N. Matni, M. Morari (Ed.)This paper proposes a computationally efficient framework, based on interval analysis, for rigorous verification of nonlinear continuous-time dynamical systems with neural network controllers. Given a neural network, we use an existing verification algorithm to construct inclusion functions for its input-output behavior. Inspired by mixed monotone theory, we embed the closed-loop dynamics into a larger system using an inclusion function of the neural network and a decomposition function of the open-loop system. This embedding provides a scalable approach for safety analysis of the neural control loop while preserving the nonlinear structure of the system. We show that one can efficiently compute hyper-rectangular over-approximations of the reachable sets using a single trajectory of the embedding system. We design an algorithm to leverage this computational advantage through partitioning strategies, improving our reachable set estimates while balancing its runtime with tunable parameters. We demonstrate the performance of this algorithm through two case studies. First, we demonstrate this method’s strength in complex nonlinear environments. Then, we show that our approach matches the performance of the state-of-the art verification algorithm for linear discretized systems.more » « less
