More Like this

1. A Barrier Certificate-based Simplex Architecture with Application to Microgrids

   Damare, Amol; Roy, Shouvik; Smolka, Scott A.; Stoller, Scott D. (January 2022, 22nd International Conference on Runtime Verification (RV 2022))

   Dang, Thao; Stolz, Volker (Ed.)

   We present Barrier-based Simplex (Bb-Simplex), a new, provably correct design for runtime assurance of continuous dynamical systems. Bb-Simplex is centered around the Simplex Control Architecture, which consists of a high-performance advanced controller which is not guaranteed to maintain safety of the plant, a verified-safe baseline controller, and a decision module that switches control of the plant between the two controllers to ensure safety without sacrificing performance. In Bb-Simplex, Barrier certificates are used to prove that the baseline controller ensures safety. Furthermore, Bb-Simplex features a new automated method for deriving, from the barrier certificate, the conditions for switching between the controllers. Our method is based on the Taylor expansion of the barrier certificate and yields computationally inexpensive switching conditions. We consider a significant application of Bb-Simplex to a microgrid featuring an advanced controller in the form of a neural network trained using reinforcement learning. The microgrid is modeled in RTDS, an industry-standard high-fidelity, real-time power systems simulator. Our results demonstrate that Bb-Simplex can automatically derive switching conditions for complex systems, the switching conditions are not overly conservative, and Bb-Simplex ensures safety even in the presence of adversarial attacks on the neural controller. 

   more » « less
2. FI-ODE: Certified and Robust Forward Invariance in Neural ODEs

   Yujia Huang; Ivan Dario Jimenez Rodriguez; Huan Zhang; Yuanyuan Shi; Yisong Yue (January 2023, arXivorg)

   Forward invariance is a long-studied property in control theory that is used to certify that a dynamical system stays within some pre-specified set of states for all time, and also admits robustness guarantees (e.g., the certificate holds under perturbations). We propose a general framework for training and provably certifying robust forward invariance in Neural ODEs. We apply this framework in two settings: certified adversarial robustness for image classification, and certified safety in continuous control. Notably, our method empirically produces superior adversarial robustness guarantees compared to prior work on certifiably robust Neural ODEs (including implicit-depth models). 

   more » « less
3. Certified Control-Oriented Learning: A Coprime Factorization Approach

   https://doi.org/10.1109/CDC51059.2022.9992821  

   Singh, Rajiv; Sznaier, Mario (December 2022, IEEE)

   This paper considers the problem of learning models to be used for controller design. Using a simple example, it argues that in this scenario the objective should reflect the closed-loop, rather than open-loop distance between the learned model and the actual plant, a task that can be accomplished by using a gap metric motivated approach. This is particularly important when identifying open-loop unstable plants, since typically in this case the open-loop distance is unbounded. In this context, the paper proposes a convex optimization approach to learn its coprime factors. This approach has a dual advantage: (1) it can easily handle open-loop unstable plants, since the coprime factors are stable, and (2) it is "self certified", since a simple norm computation on the learned factors indicates whether or not a controller designed based on these factors will stabilize the actual (unknown) plant. If this test fails, it indicates that further learning is needed. 

   more » « less
4. Second-Order Provable Defenses against Adversarial Attacks

   Singla, Sahil; Feizi, Soheil (January 2020, International Conference on Machine Learning (ICML))

   null (Ed.)

   A robustness certificate is the minimum distance of a given input to the decision boundary of the classifier (or its lower bound). For {\it any} input perturbations with a magnitude smaller than the certificate value, the classification output will provably remain unchanged. Exactly computing the robustness certificates for neural networks is difficult since it requires solving a non-convex optimization. In this paper, we provide computationally-efficient robustness certificates for neural networks with differentiable activation functions in two steps. First, we show that if the eigenvalues of the Hessian of the network are bounded, we can compute a robustness certificate in the l2 norm efficiently using convex optimization. Second, we derive a computationally-efficient differentiable upper bound on the curvature of a deep network. We also use the curvature bound as a regularization term during the training of the network to boost its certified robustness. Putting these results together leads to our proposed {\bf C}urvature-based {\bf R}obustness {\bf C}ertificate (CRC) and {\bf C}urvature-based {\bf R}obust {\bf T}raining (CRT). Our numerical results show that CRT leads to significantly higher certified robust accuracy compared to interval-bound propagation (IBP) based training. We achieve certified robust accuracy 69.79\%, 57.78\% and 53.19\% while IBP-based methods achieve 44.96\%, 44.74\% and 44.66\% on 2,3 and 4 layer networks respectively on the MNIST-dataset. 

   more » « less
5. Towards Robust Data-Driven Control Synthesis for Nonlinear Systems with Actuation Uncertainty

   https://doi.org/10.1109/CDC45484.2021.9683511  

   Taylor, Andrew J.; Dorobantu, Victor D.; Dean, Sarah; Recht, Benjamin; Yue, Yisong; Ames, Aaron D. (December 2021, 2021 60th IEEE Conference on Decision and Control (CDC))

   Modern nonlinear control theory seeks to endow systems with properties such as stability and safety, and has been deployed successfully across various domains. Despite this success, model uncertainty remains a significant challenge in ensuring that model-based controllers transfer to real world systems. This paper develops a data-driven approach to robust control synthesis in the presence of model uncertainty using Control Certificate Functions (CCFs), resulting in a convex optimization based controller for achieving properties like stability and safety. An important benefit of our framework is nuanced data-dependent guarantees, which in principle can yield sample-efficient data collection approaches that need not fully determine the input-to-state relationship. This work serves as a starting point for addressing important questions at the intersection of nonlinear control theory and non-parametric learning, both theoretical and in application. We demonstrate the efficiency of the proposed method with respect to input data in simulation with an inverted pendulum in multiple experimental settings. 

   more » « less