A survey on policy search algorithms for learning robot controllers in a handful of trials
Inria（法国国家信息与自动化研究所）；CNRS（法国国家科学研究中心）；Université de Lorraine（洛林大学）；LORIA（洛林计算机科学及其应用研究所）；Research Centre on Interactive Media, Smart Systems and Emerging Technologies（互动媒体、智能系统与新兴技术研究中心）；German Aerospace Center (DLR)（德国航空航天中心）；Idiap Research Institute（Idiap研究所）
Most policy search algorithms require thousands of training episodes to find an effective policy, which is often infeasible with a physical robot. This survey article focuses on the extreme other end of the spectrum: how can a robot adapt with only a handful of trials (a dozen) and a few minutes? By analogy with the word "big-data", we refer to this challenge as "micro-data reinforcement learning". We show that a first strategy is to leverage prior knowledge on the policy structure (e.g., dynamic movement primitives), on the policy parameters (e.g., demonstrations), or on the dynamics (e.g., simulators). A second strategy is to create data-driven surrogate models of the expected reward (e.g., Bayesian optimization) or the dynamical model (e.g., model-based policy search), so that the policy optimizer queries the model instead of the real system. Overall, all successful micro-data algorithms combine these two strategies by varying the kind of model and prior knowledge. The current scientific challenges essentially revolve around scaling up to complex robots (e.g., humanoids), designing generic priors, and optimizing the computing time.
Berkeley Artificial Intelligence Research（伯克利人工智能研究院）；University of California, Berkeley（加州大学伯克利分校）
Model-based reinforcement learning (RL) algorithms can attain excellent sample efficiency, but often lag behind the best model-free algorithms in terms of asymptotic performance. This is especially true with high-capacity parametric function approximators, such as deep networks. In this paper, we study how to bridge this gap, by employing uncertainty-aware dynamics models. We propose a new algorithm called probabilistic ensembles with trajectory sampling (PETS) that combines uncertainty-aware deep network dynamics models with sampling-based uncertainty propagation. Our comparison to state-of-t…
There is broad interest in creating RL agents that can solve many (related) tasks and adapt to new tasks and environments after initial training. Model-based RL leverages learned surrogate models that describe dynamics and rewards of individual tasks, such that planning in a good surrogate can lead to good control of the true system. Rather than solving each task individually from scratch, hierarchical models can exploit the fact that tasks are often related by (unobserved) causal factors of variation in order to achieve efficient generalization, as in learning how the mass of an item affects …
Continuous direct yaw moment control systems such as torque-vectoring controller are an essential part for vehicle stabilization. This controller has been extensively researched with the central objective of maintaining the vehicle stability by providing consistent stable cornering response. The ability of careful tuning of the parameters in a torque-vectoring controller can significantly enhance vehicle's performance and stability. However, without any re-tuning of the parameters, especially in extreme driving conditions e.g. low friction surface or high velocity, the vehicle fails to maintai…
We consider an improper reinforcement learning setting where a learner is given $M$ base controllers for an unknown Markov decision process, and wishes to combine them optimally to produce a potentially new controller that can outperform each of the base ones. This can be useful in tuning across controllers, learnt possibly in mismatched or simulated environments, to obtain a good controller for a given target environment with relatively few trials. Towards this, we propose two algorithms: (1) a Policy Gradient-based approach; and (2) an algorithm that can switch between a simple Actor-Critic …
Indian Institute of Science（印度科学学院）；Technion（以色列理工学院）
We consider an improper reinforcement learning setting where a learner is given $M$ base controllers for an unknown Markov decision process, and wishes to combine them optimally to produce a potentially new controller that can outperform each of the base ones. This can be useful in tuning across controllers, learnt possibly in mismatched or simulated environments, to obtain a good controller for a given target environment with relatively few trials. \par We propose a gradient-based approach that operates over a class of improper mixtures of the controllers. We derive convergence rate guarantee…
Google Brain（谷歌大脑）；University of California, Berkeley（加州大学伯克利分校）
Deep reinforcement learning (deep RL) holds the promise of automating the acquisition of complex controllers that can map sensory inputs directly to low-level actions. In the domain of robotic locomotion, deep RL could enable learning locomotion skills with minimal engineering and without an explicit model of the robot dynamics. Unfortunately, applying deep RL to real-world robotic tasks is exceptionally difficult, primarily due to poor sample complexity and sensitivity to hyperparameters. While hyperparameters can be easily tuned in simulated domains, tuning may be prohibitively expensive on …
Centre for Perceptual and Interactive Intelligence（感知与交互智能研究中心）；The Chinese University of Hong Kong（香港中文大学）；University of California, Los Angeles（加州大学洛杉矶分校）
Human intervention is an effective way to inject human knowledge into the training loop of reinforcement learning, which can bring fast learning and ensured training safety. Given the very limited budget of human intervention, it remains challenging to design when and how human expert interacts with the learning agent in the training. In this work, we develop a novel human-in-the-loop learning method called Human-AI Copilot Optimization (HACO).To allow the agent's sufficient exploration in the risky environments while ensuring the training safety, the human expert can take over the control and…
The Alan Turing Institute（阿兰·图灵研究所）；University of Warwick（华威大学）；Tongji University（同济大学）
Shared e-mobility services have been widely tested and piloted in cities across the globe, and already woven into the fabric of modern urban planning. This paper studies a practical yet important problem in those systems: how to deploy and manage their infrastructure across space and time, so that the services are ubiquitous to the users while sustainable in profitability. However, in real-world systems evaluating the performance of different deployment strategies and then finding the optimal plan is prohibitively expensive, as it is often infeasible to conduct many iterations of trial-and-err…
Load serving entities with storage units reach sizes and performances that can significantly impact clearing prices in electricity markets. Nevertheless, price endogeneity is rarely considered in storage bidding strategies and modeling the electricity market is a challenging task. Meanwhile, model-free reinforcement learning such as the Actor-Critic are becoming increasingly popular for designing energy system controllers. Yet implementation frequently requires lengthy, data-intense, and unsafe trial-and-error training. To fill these gaps, we implement an online Supervised Actor-Critic (SAC) a…
Harbin Institute of Technology（哈尔滨工业大学）；Delft University of Technology（代尔夫特理工大学）；University College London（伦敦大学学院）
Reinforcement learning (RL) is promising for complicated stochastic nonlinear control problems. Without using a mathematical model, an optimal controller can be learned from data evaluated by certain performance criteria through trial-and-error. However, the data-based learning approach is notorious for not guaranteeing stability, which is the most fundamental property for any control system. In this paper, the classic Lyapunov's method is explored to analyze the uniformly ultimate boundedness stability (UUB) solely based on data without using a mathematical model. It is further shown how RL w…
When manipulating a novel object with complex dynamics, a state representation is not always available, for example for deformable objects. Learning both a representation and dynamics from observations requires large amounts of data. We propose Learned Visual Similarity Predictive Control (LVSPC), a novel method for data-efficient learning to control systems with complex dynamics and high-dimensional state spaces from images. LVSPC leverages a given simple model approximation from which image observations can be generated. We use these images to train a perception model that estimates the simp…
Model-free reinforcement learning based methods such as Proximal Policy Optimization, or Q-learning typically require thousands of interactions with the environment to approximate the optimum controller which may not always be feasible in robotics due to safety and time consumption. Model-based methods such as PILCO or BlackDrops, while data-efficient, provide solutions with limited robustness and complexity. To address this tradeoff, we introduce active uncertainty reduction-based virtual environments, which are formed through limited trials conducted in the original environment. We provide a…
In this paper, we point out a fundamental property of the objective in reinforcement learning, with which we can reformulate the policy gradient objective into a perceptron-like loss function, removing the need to distinguish between on and off policy training. Namely, we posit that it is sufficient to only update a policy $π$ for cases that satisfy the condition $A(\fracπμ-1)\leq0$, where $A$ is the advantage, and $μ$ is another policy. Furthermore, we show via theoretic derivation that a perceptron-like loss function matches the clipped surrogate objective for PPO. With our new formulation, …
Department of Mechanical Engineering（机械工程系）；Carnegie Mellon University（卡内基梅隆大学）；Department of Physical Education（体育系）；Seoul National University（首尔国立大学）；Lampe Joint Department of Biomedical Engineering（生物医学工程联合部门）；UNC-Chapel Hill and NC State University（北卡罗来纳大学教堂山分校和北卡罗来纳州立大学）
Hong Kong University of Science and Technology（香港科技大学）
While deep reinforcement learning (RL) methods have achieved unprecedented successes in a range of challenging problems, their applicability has been mainly limited to simulation or game domains due to the high sample complexity of the trial-and-error learning process. However, real-world robotic applications often need a data-efficient learning process with safety-critical constraints. In this paper, we consider the challenging problem of learning unmanned aerial vehicle (UAV) control for tracking a moving target. To acquire a strategy that combines perception and control, we represent the po…
Nowadays an accurate modeling of the system to be controlled is essential for reliable autopilot. This paper presents a non-linear model of the autonomous underwater vehicle 'CWolf'. Matrices and the corresponding coefficients generate a parameterized representation for added mass, Coriolis and centripetal forces, damping, gravity and buoyancy, using the equations of motion, for all six degrees of freedom. The determination of actuator behaviour by surge tests allows the conversion of propeller revolutions to the respective forces and moments. Based on geometric approximations, the coefficient…
A Simulation Study of Functional Electrical Stimulation for An Upper Limb Rehabilitation Robot using Iterative Learning Control (ILC) and Linear models
Previous work has shown that robot navigation systems that employ an architecture based upon the idiotypic network theory of the immune system have an advantage over control techniques that rely on reinforcement learning only. This is thought to be a result of intelligent behaviour selection on the part of the idiotypic robot. In this paper an attempt is made to imitate idiotypic dynamics by creating controllers that use reinforcement with a number of different probabilistic schemes to select robot behaviour. The aims are to show that the idiotypic system is not merely performing some kind of …
University of Oxford（牛津大学）；Heriot-Watt University（赫瑞-瓦特大学）
Deep Reinforcement Learning (DRL) has been applied successfully to many robotic applications. However, the large number of trials needed for training is a key issue. Most of existing techniques developed to improve training efficiency (e.g. imitation) target on general tasks rather than being tailored for robot applications, which have their specific context to benefit from. We propose a novel framework, Assisted Reinforcement Learning, where a classical controller (e.g. a PID controller) is used as an alternative, switchable policy to speed up training of DRL for local planning and navigation…
School of Computer Science, University of Nottingham（诺丁汉大学计算机科学学院）
Previous work has shown that robot navigation systems that employ an architecture based upon the idiotypic network theory of the immune system have an advantage over control techniques that rely on reinforcement learning only. This is thought to be a result of intelligent behaviour selection on the part of the idiotypic robot. In this paper an attempt is made to imitate idiotypic dynamics by creating controllers that use reinforcement with a number of different probabilistic schemes to select robot behaviour. The aims are to show that the idiotypic system is not merely performing some kind of …
DiffCoTune: Differentiable Co-Tuning for Cross-domain Robot Control
Department of Aerospace and Mechanical Engineering, USC Viterbi School of Engineering（航空航天与机械工程系，USC维特比工程学院）；Department of Mechanical Science and Engineering, University of Illinois Urbana-Champaign（机械科学与工程系，伊利诺伊大学厄巴纳-香槟分校）
Dynamic Properties and Motion Reproducibility of a Compact Pneumatically Actuated Humanoid Upper Body for Data-Driven Control
Symbiotic Intelligent Systems Research Center, Institute for Open and Transdisciplinary Research Initiatives, The University of Osaka（共生智能系统研究中心，开放与跨学科研究倡议研究所，大阪大学）；International Professional University of Technology in Osaka, Umeda, Kita-ku, Osaka, Japan（大阪国际专业技术大学，乌梅达，北区，大阪，日本）；Chubu University Academy of Emerging Sciences, Kasugai, Aichi, Japan（名古屋大学新兴科学学院， Kasugai，爱知，日本）
KTH Royal Institute of Technology（KTH皇家理工学院）；Facebook AI Research（脸书人工智能研究院）
Data-efficiency is crucial for autonomous robots to adapt to new tasks and environments. In this work we focus on robotics problems with a budget of only 10-20 trials. This is a very challenging setting even for data-efficient approaches like Bayesian optimization (BO), especially when optimizing higher-dimensional controllers. Simulated trajectories can be used to construct informed kernels for BO. However, previous work employed supervised ways of extracting low-dimensional features for these. We propose a model and architecture for a sequential variational autoencoder that embeds the space …
University of Cincinnati（辛辛那提大学）；Harbin Institute of Technology（哈尔滨工业大学）；Tsinghua University（清华大学）
Object transportation could be a challenging problem for a single robot due to the oversize and/or overweight issues. A multi-robot system can take the advantage of increased driving power and more flexible configuration to solve such a problem. However, increased number of individuals also changed the dynamics of the system which makes control of a multi-robot system more complicated. Even worse, if the whole system is sitting on a centralized decision making unit, the data flow could be easily overloaded due to the upscaling of the system. In this research, we propose a decentralized control…
Trajectory following is one of the complicated control problems when its dynamics are nonlinear, stochastic and include a large number of parameters. The problem has significant difficulties including a large number of trials required for data collection and a massive volume of computations required to find a closed-loop controller for high dimensional and stochastic domains. For solving this type of problem, if we have an appropriate reward function and dynamics model; finding an optimal control policy is possible by using model-based reinforcement learning and optimal control algorithms. How…
