Search

Model-based Reinforcement Learning (RL) has shown its high training efficiency and capability of handling high-dimensional tasks. Regarding safety issues, safe model-based RL can achieve nearly zero-cost performance and effectively manage the trade-off between performance and safety. Nevertheless, prior works still pose safety challenges due to the online exploration in real-world deployment. To address this, some offline RL methods have emerged as solutions, which learn from a static dataset in a safe way by avoiding interactions with the environment. In this paper, we aim to further enhance safety during the deployment stage for vision-based robotic tasks by fine-tuning an offline-trained policy. We incorporate in-sample optimization, model-based policy expansion, and reachability guidance to construct a safe offline-to-online framework. Moreover, our method proves to improve the generalization of offline policy in unseen safety-constrained scenarios. Finally, the efficiency of our method is validated on simulation benchmarks with five vision-only tasks and a real robot by solving some deployment problems using limited data., Comment: 21 pages

Implicit Q-learning (IQL) serves as a strong baseline for offline RL, which learns the value function using only dataset actions through quantile regression. However, it is unclear how to recover the implicit policy from the learned implicit Q-function and why IQL can utilize weighted regression for policy extraction. IDQL reinterprets IQL as an actor-critic method and gets weights of implicit policy, however, this weight only holds for the optimal value function. In this work, we introduce a different way to solve the implicit policy-finding problem (IPF) by formulating this problem as an optimization problem. Based on this optimization problem, we further propose two practical algorithms AlignIQL and AlignIQL-hard, which inherit the advantages of decoupling actor from critic in IQL and provide insights into why IQL can use weighted regression for policy extraction. Compared with IQL and IDQL, we find our method keeps the simplicity of IQL and solves the implicit policy-finding problem. Experimental results on D4RL datasets show that our method achieves competitive or superior results compared with other SOTA offline RL methods. Especially in complex sparse reward tasks like Antmaze and Adroit, our method outperforms IQL and IDQL by a significant margin., Comment: 19 pages, 3 figures, 4 tables

Offline goal-conditioned reinforcement learning (GCRL) aims at solving goal-reaching tasks with sparse rewards from an offline dataset. While prior work has demonstrated various approaches for agents to learn near-optimal policies, these methods encounter limitations when dealing with diverse constraints in complex environments, such as safety constraints. Some of these approaches prioritize goal attainment without considering safety, while others excessively focus on safety at the expense of training efficiency. In this paper, we study the problem of constrained offline GCRL and propose a new method called Recovery-based Supervised Learning (RbSL) to accomplish safety-critical tasks with various goals. To evaluate the method performance, we build a benchmark based on the robot-fetching environment with a randomly positioned obstacle and use expert or random policies to generate an offline dataset. We compare RbSL with three offline GCRL algorithms and one offline safe RL algorithm. As a result, our method outperforms the existing state-of-the-art methods to a large extent. Furthermore, we validate the practicality and effectiveness of RbSL by deploying it on a real Panda manipulator. Code is available at https://github.com/Sunlighted/RbSL.git., Comment: Accepted by ICRA24

Whether a PTAS (polynomial-time approximation scheme) exists for equilibriums of games has been an open question, which relates to questions in three fields, the practicality of methods in algorithmic game theory, the equation PPAD=FP about the two complexity classes in computational complexity theory, and non-stationarity and curse of multiagency in MARL (multi-agent reinforcement learning). This paper introduces our discovery of the sufficient and necessary conditions for iterations based on dynamic programming and line search to approximate perfect equilibriums of dynamic games, out of which we construct a method proved to be a FPTAS (fully PTAS) for non-singular perfect equilibriums of dynamic games, where for almost any given dynamic game, all its perfect equilibriums are non-singular, indicating that FP$\subseteq$PPAD$\subseteq$Almost-FP. Our discovery consists of cone interior dynamic programming and primal-dual unbiased regret minimization, which fit into existing theories by degeneration in a structure-preserving manner. The former enables a dynamic programming operator to iteratively converge to a perfect equilibrium based on a concept called policy cone. The latter enables an interior-point line search to approximate a Nash equilibrium based on two concepts called primal-dual bias and unbiased central variety, solving a subproblem of the former. Validity of our discovery is cross-corroborated by a combination of theorem proofs, graphs of the three main concepts, and experimental results., Comment: 23 pages, 7 figures, code and animation are available at https://github.com/shb20tsinghua/PTAS_Game/tree/main

Constrained policy search (CPS) is a fundamental problem in offline reinforcement learning, which is generally solved by advantage weighted regression (AWR). However, previous methods may still encounter out-of-distribution actions due to the limited expressivity of Gaussian-based policies. On the other hand, directly applying the state-of-the-art models with distribution expression capabilities (i.e., diffusion models) in the AWR framework is intractable since AWR requires exact policy probability densities, which is intractable in diffusion models. In this paper, we propose a novel approach, $\textbf{Diffusion-based Constrained Policy Search}$ (dubbed DiffCPS), which tackles the diffusion-based constrained policy search with the primal-dual method. The theoretical analysis reveals that strong duality holds for diffusion-based CPS problems, and upon introducing parameter approximation, an approximated solution can be obtained after $\mathcal{O}(1/\epsilon)$ number of dual iterations, where $\epsilon$ denotes the representation ability of the parametrized policy. Extensive experimental results based on the D4RL benchmark demonstrate the efficacy of our approach. We empirically show that DiffCPS achieves better or at least competitive performance compared to traditional AWR-based baselines as well as recent diffusion-based offline RL methods. The code is now available at https://github.com/felix-thu/DiffCPS., Comment: 22 pages, 9 figures, 6 tables. Submitted to ICML 2024. arXiv admin note: text overlap with arXiv:1910.13393 by other authors

Autonomous terrain traversal of articulated tracked robots can reduce operator cognitive load to enhance task efficiency and facilitate extensive deployment. We present a novel hybrid trajectory optimization method aimed at generating efficient, stable, and smooth traversal motions. To achieve this, we develop a planar robot-terrain contact model and divide the robot's motion into hybrid modes of driving and traversing. By using a generalized coordinate description, the configuration space dimension is reduced, which facilitates real-time planning. The hybrid trajectory optimization is transcribed into a nonlinear programming problem and divided into subproblems to be solved in a receding-horizon planning fashion. Mode switching is facilitated by associating optimized motion durations with a predefined traversal sequence. A multi-objective cost function is formulated to further improve the traversal performance. Additionally, map sampling, terrain simplification, and tracking controller modules are integrated into the autonomous terrain traversal system. Our approach is validated in simulation and real-world scenarios with the Searcher robotic platform. Comparative experiments with expert operator control and state-of-the-art methods show advantages in terms of time and energy efficiency, stability, and smoothness of motion., Comment: IEEE Robotics and Automation Letters (RA-L)

Manual oropharyngeal (OP) swab sampling is an intensive and risky task. In this article, a novel OP swab sampling device of low cost and high compliance is designed by combining the visuo-tactile sensor and the pneumatic actuator-based gripper. Here, a concave visuo-tactile sensor called CoTac is first proposed to address the problems of high cost and poor reliability of traditional multi-axis force sensors. Besides, by imitating the doctor's fingers, a soft pneumatic actuator with a rigid skeleton structure is designed, which is demonstrated to be reliable and safe via finite element modeling and experiments. Furthermore, we propose a sampling method that adopts a compliant control algorithm based on the adaptive virtual force to enhance the safety and compliance of the swab sampling process. The effectiveness of the device has been verified through sampling experiments as well as in vivo tests, indicating great application potential. The cost of the device is around 30 US dollars and the total weight of the functional part is less than 0.1 kg, allowing the device to be rapidly deployed on various robotic arms. Videos, hardware, and source code are available at: https://sites.google.com/view/swab-sampling/., Comment: 8 pages

Given the recent surge of interest in data-driven control, this paper proposes a two-step method to study robust data-driven control for a parameter-unknown linear time-invariant (LTI) system that is affected by energy-bounded noises. First, two data experiments are designed and corresponding data are collected, then the investigated system is equivalently written into a data-based descriptor system with structured parametric uncertainties. Second, combined with model-based control theory for descriptor systems, state feedback controllers are designed for such data-based descriptor system, which stabilize the original LTI system and guarantee the ${H_\infty}$ performance. Finally, a simulation example is provided to illustrate the effectiveness and merits of our method.

For a parameter-unknown linear descriptor system, this paper proposes data-driven methods to testify the system's type and controllability and then to stabilize it. First, a data-based condition is developed to identify whether this unknown system is a descriptor system or is equivalent to a normal system. Furthermore, various controllability concepts are testified by replacing the descriptor system's matrices with data. Finally, a data-based decomposing method is proposed to transfer the nominal system into its slow-fast subsystems' form, so that a state feedback controller for the slow subsystem can be obtained from persistently exciting input and state sequences. Meanwhile, due to the equivalent stabilizability between the nominal system and its slow subsystem, a state feedback controller which stabilizes the nominal system is also obtained. A simulation example is provided to illustrate the effectiveness of those methods.