In safe reinforcement learning (SRL) problems, an agent explores the environment to maximize an expected total reward and meanwhile avoids violation of certain constraints on a number of expected total costs. In general, such SRL problems have nonconvex objective functions subject to multiple nonconvex constraints, and hence are very challenging to solve, particularly to provide a globally optimal policy. Many popular SRL algorithms adopt a primal-dual structure which utilizes the updating of dual variables for satisfying the constraints. In contrast, we propose a primal approach, called constraint-rectified policy optimization (CRPO), which updates the policy alternatingly between objective improvement and constraint satisfaction. CRPO provides a primal-type algorithmic framework to solve SRL problems, where each policy update can take any variant of policy optimization step. To demonstrate the theoretical performance of CRPO, we adopt natural policy gradient (NPG) for each policy update step and show that CRPO achieves an $\mathcal{O}(1/\sqrt{T})$ convergence rate to the global optimal policy in the constrained policy set and an $\mathcal{O}(1/\sqrt{T})$ error bound on constraint satisfaction. This is the first finite-time analysis of primal SRL algorithms with global optimality guarantee. Our empirical results demonstrate that CRPO can outperform the existing primal-dual baseline algorithms significantly.
Technical University of Munich（慕尼黑工业大学）；Wayne State University（韦恩州立大学）；University of Illinois at Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）
Deep reinforcement learning (DRL) has achieved tremendous success in many complex decision-making tasks of autonomous systems with high-dimensional state and/or action spaces. However, the safety and stability still remain major concerns that hinder the applications of DRL to safety-critical autonomous systems. To address the concerns, we proposed the Phy-DRL: a physical deep reinforcement learning framework. The Phy-DRL is novel in two architectural designs: i) Lyapunov-like reward, and ii) residual control (i.e., integration of physics-model-based control and data-driven control). The concur…
Pacific Northwest National Laboratory（太平洋西北国家实验室）；Battelle Memorial Institute（巴特尔纪念研究所）
We develop a novel form of differentiable predictive control (DPC) with safety and robustness guarantees based on control barrier functions. DPC is an unsupervised learning-based method for obtaining approximate solutions to explicit model predictive control (MPC) problems. In DPC, the predictive control policy parametrized by a neural network is optimized offline via direct policy gradients obtained by automatic differentiation of the MPC problem. The proposed approach exploits a new form of sampled-data barrier function to enforce offline and online safety requirements in DPC settings while …
Neural networks serve as effective controllers in a variety of complex settings due to their ability to represent expressive policies. The complex nature of neural networks, however, makes their output difficult to verify and predict, which limits their use in safety-critical applications. While simulations provide insight into the performance of neural network controllers, they are not enough to guarantee that the controller will perform safely in all scenarios. To address this problem, recent work has focused on formal methods to verify properties of neural network outputs. For neural networ…
Robust Probabilistic Shielding for Safe Offline Reinforcement Learning
Radboud University Nijmegen The Netherlands；Ruhr University\,\&\,Radboud University Bochum Germany；Radboud University；Ruhr University\,\&\,Radboud University
In environments with uncertain dynamics exploration is necessary to learn how to perform well. Existing reinforcement learning algorithms provide strong exploration guarantees, but they tend to rely on an ergodicity assumption. The essence of ergodicity is that any state is eventually reachable from any other state by following a suitable policy. This assumption allows for exploration algorithms that operate by simply favoring states that have rarely been visited before. For most physical systems this assumption is impractical as the systems would break before any reasonable exploration has ta…
Several works have addressed the problem of incorporating constraints in the reinforcement learning (RL) framework, however majority of them can only guarantee the satisfaction of soft constraints. In this work, we address the problem of satisfying hard state constraints in a model-free RL setting with the deterministic system dynamics. The proposed algorithm is developed for the discrete state and action space and utilizes a multi-class support vector machine (SVM) to represent the policy. The state constraints are incorporated in the SVM optimization framework to derive an analytical solutio…
Safe Learning Control with Optimality and Stability Guarantees
Shenzhen Key Lab for Advanced Motion Control and Modern Automation Equipments, School of Intelligence Science and Engineering, Harbin Institute of Technology, Shenzhen, Guangdong 518055, China（深圳先进 motion 控制与现代自动化装备重点实验室，智能科学与工程学院，哈尔滨工业大学，深圳，广东 518055，中国）；School of Data Science, Lingnan University, Hong Kong（数据科学学院，岭南大学，香港）；School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore（电气与电子工程学院，南洋理工大学，新加坡）；MIT CSAIL
Safe exploration is a key to applying reinforcement learning (RL) in safety-critical systems. Existing safe exploration methods guaranteed safety under the assumption of regularity, and it has been difficult to apply them to large-scale real problems. We propose a novel algorithm, SPO-LF, that optimizes an agent's policy while learning the relation between a locally available feature obtained by sensors and environmental reward/safety using generalized linear function approximations. We provide theoretical guarantees on its safety and optimality. We experimentally show that our algorithm is 1)…
More for Less: Safe Policy Improvement With Stronger Performance Guarantees
Technische Universität Dresden（德累斯顿工业大学）；Centre for Tactile Internet with Human-in-the-Loop (CeTI)（人在环触觉互联网中心（CeTI））；Radboud University（拉德堡德大学）；Eindhoven University of Technology（埃因霍温理工大学）
University of Freiburg（弗赖堡大学）；ETH Zurich（苏黎世联邦理工学院）
Reinforcement learning has been successfully used to solve difficult tasks in complex unknown environments. However, these methods typically do not provide any safety guarantees during the learning process. This is particularly problematic, since reinforcement learning agent actively explore their environment. This prevents their use in safety-critical, real-world applications. In this paper, we present a learning-based model predictive control scheme that provides high-probability safety guarantees throughout the learning process. Based on a reliable statistical model, we construct provably a…
Predictive Safety Shield for Dyna-Q Reinforcement Learning
Departement U2IS, ENSTA Paris, Institut Polytechnique de Paris（ENSTA巴黎大学U2IS部门，巴黎理工学院）；Department of Electrical Engineering and Computer Science, University of California, Berkeley, USA（加州大学伯克利分校电气工程与计算机科学系）
Department of Computer Science, University of Manchester（曼彻斯特大学计算机科学系）；Faculty of Computer Science & DSME, RWTH Aachen University（亚琛工业大学计算机科学与DSME学院）；Department of Computer Science, University of Oxford（牛津大学计算机科学系）；Department of Computing, Imperial College London（伦敦帝国理工学院计算系）
Texas A&M University（得克萨斯农工大学）；University of Washington（华盛顿大学）
We present safe active incremental feature selection~(SAIF) to scale up the computation of LASSO solutions. SAIF does not require a solution from a heavier penalty parameter as in sequential screening or updating the full model for each iteration as in dynamic screening. Different from these existing screening methods, SAIF starts from a small number of features and incrementally recruits active features and updates the significantly reduced model. Hence, it is much more computationally efficient and scalable with the number of features. More critically, SAIF has the safe guarantee as it has t…
As artificial agents proliferate, it is becoming increasingly important to ensure that their interactions with one another are well-behaved. In this paper, we formalize a common-sense notion of when algorithms are well-behaved: an algorithm is safe if it does no harm. Motivated by recent progress in deep learning, we focus on the specific case where agents update their actions according to gradient descent. The paper shows that that gradient descent converges to a Nash equilibrium in safe games. The main contribution is to define strongly-typed agents and show they are guaranteed to interact s…
Autonomous systems like aircraft and assistive robots often operate in scenarios where guaranteeing safety is critical. Methods like Hamilton-Jacobi reachability can provide guaranteed safe sets and controllers for such systems. However, often these same scenarios have unknown or uncertain environments, system dynamics, or predictions of other agents. As the system is operating, it may learn new knowledge about these uncertainties and should therefore update its safety analysis accordingly. However, work to learn and update safety analysis is limited to small systems of about two dimensions du…
SafePredict is a novel meta-algorithm that works with any base prediction algorithm for online data to guarantee an arbitrarily chosen correctness rate, $1-ε$, by allowing refusals. Allowing refusals means that the meta-algorithm may refuse to emit a prediction produced by the base algorithm on occasion so that the error rate on non-refused predictions does not exceed $ε$. The SafePredict error bound does not rely on any assumptions on the data distribution or the base predictor. When the base predictor happens not to exceed the target error rate $ε$, SafePredict refuses only a finite number o…
University of Illinois at Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）；MIT-IBM Watson AI Lab（麻省理工学院-IBM沃森人工智能实验室）；IBM Research（IBM研究院）；SEAS Harvard University（哈佛大学工程与应用科学学院）
We study the adaptive control of an unknown linear system with a quadratic cost function subject to safety constraints on both the states and actions. The challenges of this problem arise from the tension among safety, exploration, performance, and computation. To address these challenges, we propose a polynomial-time algorithm that guarantees feasibility and constraint satisfaction with high probability under proper conditions. Our algorithm is implemented on a single trajectory and does not require system restarts. Further, we analyze the regret of our learning algorithm compared to the opti…
Typal LLC（Typal有限责任公司）；The University of Texas at Austin（德克萨斯大学奥斯汀分校）；Alibaba US, DAMO Academy（阿里巴巴美国达摩院）
Applications abound in which optimization problems must be repeatedly solved, each time with new (but similar) data. Analytic optimization algorithms can be hand-designed to provably solve these problems in an iterative fashion. On one hand, data-driven algorithms can "learn to optimize" (L2O) with much fewer iterations and similar cost per iteration as general-purpose optimization algorithms. On the other hand, unfortunately, many L2O algorithms lack converge guarantees. To fuse the advantages of these approaches, we present a Safe-L2O framework. Safe-L2O updates incorporate a safeguard to gu…
Safety and optimality in learning-based control at low computational cost
Aalto University（阿alto大学）；Uppsala University（乌普萨拉大学）；Wrocław University of Science and Technology（沃斯克拉大学科学与技术大学）；KTH Royal Institute of Technology（皇家理工学院）；Eindhoven University of Technology（埃因霍温理工大学）
