A desirable property in fault-tolerant controllers is adaptability to system changes as they evolve during systems operations. An adaptive controller does not require optimal control policies to be enumerated for possible faults. Instead it can approximate one in real-time. We present two adaptive fault-tolerant control schemes for a discrete time system based on hierarchical reinforcement learning. We compare their performance against a model predictive controller in presence of sensor noise and persistent faults. The controllers are tested on a fuel tank model of a C-130 plane. Our experiments demonstrate that reinforcement learning-based controllers perform more robustly than model predictive controllers under faults, partially observable system models, and varying sensor noise levels.

We propose a novel adaptive reinforcement learning control approach for fault tolerant control of degrading systems that is not preceded by a fault detection and diagnosis step. Therefore, \textit{a priori} knowledge of faults that may occur in the system is not required. The adaptive scheme combines online and offline learning of the on-policy control method to improve exploration and sample efficiency, while guaranteeing stable learning. The offline learning phase is performed using a data-driven model of the system, which is frequently updated to track the system's operating conditions. We conduct experiments on an aircraft fuel transfer system to demonstrate the effectiveness of our approach.

Although recent work in reinforcement learning has shown that robots can learn complex individual skills such as grasping [2], locomotion [3, 4], and manipulation tasks [5], designing reinforcement learning algorithms that perform well in sparse reward scenarios is still an open challenge of artificial intelligence. Standard reinforcement learning algorithms struggle in the sparse reward scenario because they rely on simple exploration behavior such as random actions. As a result, learning complex tasks often requires manually collecting examples [6, 7] or running learning algorithms over a long period of time which may not be possible in real life scenarios. Designing better exploration schemes would help agents autonomously discover interesting features that can then be used to learn the long term objective. Developing efficient exploration algorithms would thus help create a more autonomous learning agent. Several approaches have been considered in order to improve the exploration performances of reinforcement learning algorithms. One approach is to reward the agent for discovering novel observations in the form of an intrinsic reward that is added to the original reward of the environment [8].

Automatic industrial scheduling, aiming at optimizing the sequence of jobs over limited resources, is widely needed in manufacturing industries. However, existing scheduling systems heavily rely on heuristic algorithms, which either generate ineffective solutions or compute inefficiently when job scale increases. Thus, it is of great importance to develop new large-scale algorithms that are not only efficient and effective, but also capable of satisfying complex constraints in practice. In this paper, we propose a Bilevel Deep reinforcement learning Scheduler, \textit{BDS}, in which the higher level is responsible for exploring an initial global sequence, whereas the lower level is aiming at exploitation for partial sequence refinements, and the two levels are connected by a sliding-window sampling mechanism. In the implementation, a Double Deep Q Network (DDQN) is used in the upper level and Graph Pointer Network (GPN) lies within the lower level. After the theoretical guarantee for the convergence of BDS, we evaluate it in an industrial automatic warehouse scenario, with job number up to $5000$ in each production line. It is shown that our proposed BDS significantly outperforms two most used heuristics, three strong deep networks, and another bilevel baseline approach. In particular, compared with the most used greedy-based heuristic algorithm in real world which takes nearly an hour, our BDS can decrease the makespan by 27.5\%, 28.6\% and 22.1\% for 3 largest datasets respectively, with computational time less than 200 seconds.

Learned dynamics models combined with both planning and policy learning algorithms have shown promise in enabling artificial agents to learn to perform many diverse tasks with limited supervision. However, one of the fundamental challenges in using a learned forward dynamics model is the mismatch between the objective of the learned model (future state reconstruction), and that of the downstream planner or policy (completing a specified task). This issue is exacerbated by vision-based control tasks in diverse real-world environments, where the complexity of the real world dwarfs model capacity. In this paper, we propose to direct prediction towards task relevant information, enabling the model to be aware of the current task and encouraging it to only model relevant quantities of the state space, resulting in a learning objective that more closely matches the downstream task. Further, we do so in an entirely self-supervised manner, without the need for a reward function or image labels. We find that our method more effectively models the relevant parts of the scene conditioned on the goal, and as a result outperforms standard task-agnostic dynamics models and model-free reinforcement learning.

Welcome to the Tonic deep reinforcement learning library. Modularity: Building blocks for creating RL agents, such as models, replays, or exploration strategies, are implemented as configurable modules. Readability: Agents are written in a simple way with an identical API and logs are nicely displayed on the terminal with a progress bar. Fair comparison: The training pipeline is unique and compatible with all Tonic agents and environments. Agents are defined by their core ideas while general tricks/improvements like non-terminal timeouts, observation normalization and action scaling are shared.

The task of searching certain people in videos has seen increasing potential in real-world applications, such as video organization and editing. Most existing approaches are devised to work in an offline manner, where identities can only be inferred after an entire video is examined. This working manner precludes such methods from being applied to online services or those applications that require real-time responses. In this paper, we propose an online person search framework, which can recognize people in a video on the fly. This framework maintains a multimodal memory bank at its heart as the basis for person recognition, and updates it dynamically with a policy obtained by reinforcement learning. Our experiments on a large movie dataset show that the proposed method is effective, not only achieving remarkable improvements over online schemes but also outperforming offline methods.

This work is inspired by recent advances in hierarchical reinforcement learning (HRL) (Barto and Mahadevan 2003;Hengst 2010), and improvements in learning efficiency with heuristic-based subgoal selection and hindsight experience replay (HER)(Andrychowicz et al. 2017; Levy et al. 2019). We propose a new method to integrate HRL, HER and effective subgoal selection based on human expertise to support sample-efficient learning and enhance interpretability of the agent's behavior. Human expertise remains indispensable in many areas such as medicine (Buch, Ahmed, and Maruthappu 2018) and law (Cath 2018), where interpretability, explainability and transparency are crucial in the decision making process, for ethical and legal reasons. Our method simplifies the complex task sets for achieving the overall objectives by decomposing into subgoals at different levels of abstraction. Incorporating relevant subjective knowledge also significantly reduces the computational resources spent in exploration for RL, especially in high speed, changing, and complex environments where the transition dynamics cannot be effectively learned and modelled in a short time. Experimental results in two StarCraft II (SC2) minigames demonstrate that our method can achieve better sample efficiency than flat and end-to-end RL methods, and provide an effective method for explaining the agent's performance.

The design of building heating, ventilation, and air conditioning (HVAC) system is critically important, as it accounts for around half of building energy consumption and directly affects occupant comfort, productivity, and health. Traditional HVAC control methods are typically based on creating explicit physical models for building thermal dynamics, which often require significant effort to develop and are difficult to achieve sufficient accuracy and efficiency for runtime building control and scalability for field implementations. Recently, deep reinforcement learning (DRL) has emerged as a promising data-driven method that provides good control performance without analyzing physical models at runtime. However, a major challenge to DRL (and many other data-driven learning methods) is the long training time it takes to reach the desired performance. In this work, we present a novel transfer learning based approach to overcome this challenge. Our approach can effectively transfer a DRL-based HVAC controller trained for the source building to a controller for the target building with minimal effort and improved performance, by decomposing the design of neural network controller into a transferable front-end network that captures building-agnostic behavior and a back-end network that can be efficiently trained for each specific building. We conducted experiments on a variety of transfer scenarios between buildings with different sizes, numbers of thermal zones, materials and layouts, air conditioner types, and ambient weather conditions. The experimental results demonstrated the effectiveness of our approach in significantly reducing the training time, energy cost, and temperature violations.

Imitation learning (IL, Schaal, 1999; Osa et al., 2018) and inverse reinforcement learning (IRL, Ng and Russell, 2000) are two related areas of research that aim to teach agents by providing demonstrations of the desired behavior. Whereas imitation learning aims to learn a policy that results in a similar behavior, inverse reinforcement learning focuses on inferring a reward function that might have been optimized by the demonstrator, aiming to better generalize to different environments. Both areas of research are often formalized as distribution-matching, that is, the learned policy (or the optimal policy for IRL) should induce a distribution over states and actions that is close to the expert's distribution with respect to a given (usually non-metric) distance. Commonly applied distances are the forward Kullback-Leibler (KL) divergence (e.g., Ziebart, 2010), which maximizes the likelihood of the demonstrated state-action pairs under the agent's distribution, and the reverse Kullback-Leibler (RKL) divergence (e.g., Arenz et al., 2016; Fu et al., 2018; Ghasemipour et al., 2020) which minimizes the expected discrimination information (Kullback and Leibler, 1951) of state-action pairs sampled from the agent's distribution. However, since the emergence of generative adversarial networks (GANs, Goodfellow et al., 2014) as a solution technique for both areas, other divergences have been investigated such as the Jensen-Shannon divergence (Ho and Ermon, 2016), the Wasserstein distance (Xiao et al., 2019) and general f-divergences (Ke et al., 2019; Ghasemipour et al., 2020).