arXiv subjects

Dileep Kalathil

Publications and source records attributed to Dileep Kalathil.

At least 55 records · Page 3Linked to original sources

arXiv2022

Optimal Control of Material Micro-Structures

In this paper, we consider the optimal control of material micro-structures. Such material micro-structures are modeled by the so-called phase field model. We study the underlying physical structure of the model and propose a data based approach for its optimal control, along with a comparison to the control using a state of the art Reinforcement Learning (RL) algorithm. Simulation results show the feasibility of optimally controlling such micro-structures to attain desired material properties and complex target micro-structures.

eess.SY↗

arXiv2022

Enhanced Meta Reinforcement Learning using Demonstrations in Sparse Reward Environments

Meta reinforcement learning (Meta-RL) is an approach wherein the experience gained from solving a variety of tasks is distilled into a meta-policy. The meta-policy, when adapted over only a small (or just a single) number of steps, is able to perform near-optimally on a new, related task. However, a major challenge to adopting this approach to solve real-world problems is that they are often associated with sparse reward functions that only indicate whether a task is completed partially or fully. We consider the situation where some data, possibly generated by a sub-optimal agent, is available for each task. We then develop a class of algorithms entitled Enhanced Meta-RL using Demonstrations (EMRLD) that exploit this information even if sub-optimal to obtain guidance during training. We show how EMRLD jointly utilizes RL and supervised learning over the offline data to generate a meta-policy that demonstrates monotone performance improvements. We also develop a warm started variant called EMRLD-WS that is particularly efficient for sub-optimal demonstration data. Finally, we show that our EMRLD algorithms significantly outperform existing approaches in a variety of sparse reward environments, including that of a mobile robot.

cs.LG↗

arXiv2022

Intelligent Vision-based Autonomous Ship Landing of VTOL UAVs

The paper discusses an intelligent vision-based control solution for autonomous tracking and landing of Vertical Take-Off and Landing (VTOL) capable Unmanned Aerial Vehicles (UAVs) on ships without utilizing GPS signal. The central idea involves automating the Navy helicopter ship landing procedure where the pilot utilizes the ship as the visual reference for long-range tracking; however, refers to a standardized visual cue installed on most Navy ships called the "horizon bar" for the final approach and landing phases. This idea is implemented using a uniquely designed nonlinear controller integrated with machine vision. The vision system utilizes machine learning-based object detection for long-range ship tracking and classical computer vision for the estimation of aircraft relative position and orientation utilizing the horizon bar during the final approach and landing phases. The nonlinear controller operates based on the information estimated by the vision system and has demonstrated robust tracking performance even in the presence of uncertainties. The developed autonomous ship landing system was implemented on a quad-rotor UAV equipped with an onboard camera, and approach and landing were successfully demonstrated on a moving deck, which imitates realistic ship deck motions. Extensive simulations and flight tests were conducted to demonstrate vertical landing safety, tracking capability, and landing accuracy.

cs.RO↗

arXiv2022

Robust Reinforcement Learning Algorithm for Vision-based Ship Landing of UAVs

This paper addresses the problem of developing an algorithm for autonomous ship landing of vertical take-off and landing (VTOL) capable unmanned aerial vehicles (UAVs), using only a monocular camera in the UAV for tracking and localization. Ship landing is a challenging task due to the small landing space, six degrees of freedom ship deck motion, limited visual references for localization, and adversarial environmental conditions such as wind gusts. We first develop a computer vision algorithm which estimates the relative position of the UAV with respect to a horizon reference bar on the landing platform using the image stream from a monocular vision camera on the UAV. Our approach is motivated by the actual ship landing procedure followed by the Navy helicopter pilots in tracking the horizon reference bar as a visual cue. We then develop a robust reinforcement learning (RL) algorithm for controlling the UAV towards the landing platform even in the presence of adversarial environmental conditions such as wind gusts. We demonstrate the superior performance of our algorithm compared to a benchmark nonlinear PID control approach, both in the simulation experiments using the Gazebo environment and in the real-world setting using a Parrot ANAFI quad-rotor and sub-scale ship platform undergoing 6 degrees of freedom (DOF) deck motion.

cs.RO↗

arXiv2022

DETERRENT: Detecting Trojans using Reinforcement Learning

Insertion of hardware Trojans (HTs) in integrated circuits is a pernicious threat. Since HTs are activated under rare trigger conditions, detecting them using random logic simulations is infeasible. In this work, we design a reinforcement learning (RL) agent that circumvents the exponential search space and returns a minimal set of patterns that is most likely to detect HTs. Experimental results on a variety of benchmarks demonstrate the efficacy and scalability of our RL agent, which obtains a significant reduction ($169\times$) in the number of test patterns required while maintaining or improving coverage ($95.75\%$) compared to the state-of-the-art techniques.

cs.LG↗

arXiv2022

Meta-Learning Online Control for Linear Dynamical Systems

In this paper, we consider the problem of finding a meta-learning online control algorithm that can learn across the tasks when faced with a sequence of $N$ (similar) control tasks. Each task involves controlling a linear dynamical system for a finite horizon of $T$ time steps. The cost function and system noise at each time step are adversarial and unknown to the controller before taking the control action. Meta-learning is a broad approach where the goal is to prescribe an online policy for any new unseen task exploiting the information from other tasks and the similarity between the tasks. We propose a meta-learning online control algorithm for the control setting and characterize its performance by \textit{meta-regret}, the average cumulative regret across the tasks. We show that when the number of tasks are sufficiently large, our proposed approach achieves a meta-regret that is smaller by a factor $D/D^{*}$ compared to an independent-learning online control algorithm which does not perform learning across the tasks, where $D$ is a problem constant and $D^{*}$ is a scalar that decreases with increase in the similarity between tasks. Thus, when the sequence of tasks are similar the regret of the proposed meta-learning online control is significantly lower than that of the naive approaches without meta-learning. We also present experiment results to demonstrate the superior performance achieved by our meta-learning algorithm.

cs.LG↗

arXiv2022

Distributed Learning of Neural Lyapunov Functions for Large-Scale Networked Dissipative Systems

This paper considers the problem of characterizing the stability region of a large-scale networked system comprised of dissipative nonlinear subsystems, in a distributed and computationally tractable way. One standard approach to estimate the stability region of a general nonlinear system is to first find a Lyapunov function for the system and characterize its region of attraction as the stability region. However, classical approaches, such as sum-of-squares methods and quadratic approximation, for finding a Lyapunov function either do not scale to large systems or give very conservative estimates for the stability region. In this context, we propose a new distributed learning based approach by exploiting the dissipativity structure of the subsystems. Our approach has two parts: the first part is a distributed approach to learn the storage functions (similar to the Lyapunov functions) for all the subsystems, and the second part is a distributed optimization approach to find the Lyapunov function for the networked system using the learned storage functions of the subsystems. We demonstrate the superior performance of our proposed approach through extensive case studies in microgrid networks.

eess.SY↗

arXiv2022

Sample Complexity of Robust Reinforcement Learning with a Generative Model

The Robust Markov Decision Process (RMDP) framework focuses on designing control policies that are robust against the parameter uncertainties due to the mismatches between the simulator model and real-world settings. An RMDP problem is typically formulated as a max-min problem, where the objective is to find the policy that maximizes the value function for the worst possible model that lies in an uncertainty set around a nominal model. The standard robust dynamic programming approach requires the knowledge of the nominal model for computing the optimal robust policy. In this work, we propose a model-based reinforcement learning (RL) algorithm for learning an $ε$-optimal robust policy when the nominal model is unknown. We consider three different forms of uncertainty sets, characterized by the total variation distance, chi-square divergence, and KL divergence. For each of these uncertainty sets, we give a precise characterization of the sample complexity of our proposed algorithm. In addition to the sample complexity results, we also present a formal analytical argument on the benefit of using robust policies. Finally, we demonstrate the performance of our algorithm on two benchmark problems.

cs.LG↗

arXiv2022

The Impact of Heavy-Duty Vehicle Electrification on Large Power Grids: a Synthetic Texas Case Study

The electrification of heavy-duty vehicles (HDEVs) is a nascent and rapidly emerging avenue for decarbonization of the transportation sector. In this paper, we examine the impacts of increased vehicle electrification on the power grid infrastructure, with particular focus on HDEVs. We utilize a synthetic representation of the 2000-bus Texas transmission grid, and realistic representations of multiple distribution grids in Travis county, Texas, as well as transit data pertaining to HDEVs, to uncover the consequences of HDEV electrification, and expose the limitations imposed by existing electric grid infrastructure. Our analysis reveals that grid-wide voltage problems that are spatiotemporally correlated with the mobility of HDEVs may occur even at modest penetration levels. In fact, we find that as little as 11% of heavy duty vehicles in Texas charging simultaneously can lead to significant voltage violations on the transmission network that compromise grid reliability. Furthermore, we find that just a few dozen EVs charging simultaneously can lead to voltage violations at the distribution level.

eess.SY↗

arXiv2022

Reinforcement Learning with Sparse Rewards using Guidance from Offline Demonstration

A major challenge in real-world reinforcement learning (RL) is the sparsity of reward feedback. Often, what is available is an intuitive but sparse reward function that only indicates whether the task is completed partially or fully. However, the lack of carefully designed, fine grain feedback implies that most existing RL algorithms fail to learn an acceptable policy in a reasonable time frame. This is because of the large number of exploration actions that the policy has to perform before it gets any useful feedback that it can learn from. In this work, we address this challenging problem by developing an algorithm that exploits the offline demonstration data generated by a sub-optimal behavior policy for faster and efficient online RL in such sparse reward settings. The proposed algorithm, which we call the Learning Online with Guidance Offline (LOGO) algorithm, merges a policy improvement step with an additional policy guidance step by using the offline demonstration data. The key idea is that by obtaining guidance from - not imitating - the offline data, LOGO orients its policy in the manner of the sub-optimal policy, while yet being able to learn beyond and approach optimality. We provide a theoretical analysis of our algorithm, and provide a lower bound on the performance improvement in each learning episode. We also extend our algorithm to the even more challenging incomplete observation setting, where the demonstration data contains only a censored version of the true state observation. We demonstrate the superior performance of our algorithm over state-of-the-art approaches on a number of benchmark environments with sparse rewards and censored state. Further, we demonstrate the value of our approach via implementing LOGO on a mobile robot for trajectory tracking and obstacle avoidance, where it shows excellent performance.

cs.LG↗

arXiv2022

Policy Optimization for Constrained MDPs with Provable Fast Global Convergence

We address the problem of finding the optimal policy of a constrained Markov decision process (CMDP) using a gradient descent-based algorithm. Previous results have shown that a primal-dual approach can achieve an $\mathcal{O}(1/\sqrt{T})$ global convergence rate for both the optimality gap and the constraint violation. We propose a new algorithm called policy mirror descent-primal dual (PMD-PD) algorithm that can provably achieve a faster $\mathcal{O}(\log(T)/T)$ convergence rate for both the optimality gap and the constraint violation. For the primal (policy) update, the PMD-PD algorithm utilizes a modified value function and performs natural policy gradient steps, which is equivalent to a mirror descent step with appropriate regularization. For the dual update, the PMD-PD algorithm uses modified Lagrange multipliers to ensure a faster convergence rate. We also present two extensions of this approach to the settings with zero constraint violation and sample-based estimation. Experimental results demonstrate the faster convergence rate and the better performance of the PMD-PD algorithm compared with existing policy gradient-based algorithms.

cs.LG↗

arXiv2021

Safe Online Convex Optimization with Unknown Linear Safety Constraints

We study the problem of safe online convex optimization, where the action at each time step must satisfy a set of linear safety constraints. The goal is to select a sequence of actions to minimize the regret without violating the safety constraints at any time step (with high probability). The parameters that specify the linear safety constraints are unknown to the algorithm. The algorithm has access to only the noisy observations of constraints for the chosen actions. We propose an algorithm, called the {Safe Online Projected Gradient Descent} (SO-PGD) algorithm, to address this problem. We show that, under the assumption of the availability of a safe baseline action, the SO-PGD algorithm achieves a regret $O(T^{2/3})$. While there are many algorithms for online convex optimization (OCO) problems with safety constraints available in the literature, they allow constraint violations during learning/optimization, and the focus has been on characterizing the cumulative constraint violations. To the best of our knowledge, ours is the first work that provides an algorithm with provable guarantees on the regret, without violating the linear safety constraints (with high probability) at any time step.

cs.LG↗

arXiv2021

OTTR: Off-Road Trajectory Tracking using Reinforcement Learning

In this work, we present a novel Reinforcement Learning (RL) algorithm for the off-road trajectory tracking problem. Off-road environments involve varying terrain types and elevations, and it is difficult to model the interaction dynamics of specific off-road vehicles with such a diverse and complex environment. Standard RL policies trained on a simulator will fail to operate in such challenging real-world settings. Instead of using a naive domain randomization approach, we propose an innovative supervised-learning based approach for overcoming the sim-to-real gap problem. Our approach efficiently exploits the limited real-world data available to adapt the baseline RL policy obtained using a simple kinematics simulator. This avoids the need for modeling the diverse and complex interaction of the vehicle with off-road environments. We evaluate the performance of the proposed algorithm using two different off-road vehicles, Warthog and Moose. Compared to the standard ILQR approach, our proposed approach achieves a 30% and 50% reduction in cross track error in Warthog and Moose, respectively, by utilizing only 30 minutes of real-world driving data.

cs.RO↗

arXiv2021

PyProD: A Machine Learning-Friendly Platform for Protection Analytics in Distribution Systems

This paper introduces PyProD, a Python-based machine learning (ML)-compatible test-bed for evaluating the efficacy of protection schemes in electric distribution grids. This testbed is designed to bridge the gap between conventional power distribution grid analysis and growing capability of ML-based decision making algorithms, in particular in the context of protection system design and configuration. PyProD is shown to be capable of facilitating efficient design and evaluation of ML-based decision making algorithms for protection devices in the future electric distribution grid, in which many distributed energy resources and pro-sumers permeate the system.

eess.SY↗

arXiv2021

Deep Reinforcement Learning-BasedRobust Protection in DER-Rich Distribution Grids

This paper introduces the concept of Deep Reinforcement Learning based architecture for protective relay design in power distribution systems with many distributed energy resources (DERs). The performance of widely-used overcurrent protection scheme is hindered by the presence of distributed generation, power electronic interfaced devices and fault impedance. In this paper, a reinforcement learning-based approach is proposed to design and implement protective relays in the distribution grid. The particular algorithm used is an Long Short-Term Memory (LSTM) enhanced deep neural network that is highly accurate, communication-free and easy to implement. The proposed relay design is tested in OpenDSS simulation on the IEEE 34-node test feeder and demonstrated much more superior performance over traditional overcurrent protection from the aspect of failure rate, robustness and response speed.

eess.SY↗

arXiv2021

Fully Decentralized Reinforcement Learning-based Control of Photovoltaics in Distribution Grids for Joint Provision of Real and Reactive Power

In this paper, we introduce a new framework to address the problem of voltage regulation in unbalanced distribution grids with deep photovoltaic penetration. In this framework, both real and reactive power setpoints are explicitly controlled at each solar panel smart inverter, and the objective is to simultaneously minimize system-wide voltage deviation and maximize solar power output. We formulate the problem as a Markov decision process with continuous action spaces and use proximal policy optimization, a reinforcement learning-based approach, to solve it, without the need for any forecast or explicit knowledge of network topology or line parameters. By representing the system in a quasi-steady state manner, and by carefully formulating the Markov decision process, we reduce the complexity of the problem and allow for fully decentralized (communication-free) policies, all of which make the trained policies much more practical and interpretable. Numerical simulations on a 240-node unbalanced distribution grid, based on a real network in Midwest U.S., are used to validate the proposed framework and reinforcement learning approach.

eess.SY↗

arXiv2021

Learning with Safety Constraints: Sample Complexity of Reinforcement Learning for Constrained MDPs

Many physical systems have underlying safety considerations that require that the policy employed ensures the satisfaction of a set of constraints. The analytical formulation usually takes the form of a Constrained Markov Decision Process (CMDP). We focus on the case where the CMDP is unknown, and RL algorithms obtain samples to discover the model and compute an optimal constrained policy. Our goal is to characterize the relationship between safety constraints and the number of samples needed to ensure a desired level of accuracy -- both objective maximization and constraint satisfaction -- in a PAC sense. We explore two classes of RL algorithms, namely, (i) a generative model based approach, wherein samples are taken initially to estimate a model, and (ii) an online approach, wherein the model is updated as samples are obtained. Our main finding is that compared to the best known bounds of the unconstrained regime, the sample complexity of constrained RL algorithms are increased by a factor that is logarithmic in the number of constraints, which suggests that the approach may be easily utilized in real systems.

cs.LG↗

arXiv2021

Robust Reinforcement Learning using Least Squares Policy Iteration with Provable Performance Guarantees

This paper addresses the problem of model-free reinforcement learning for Robust Markov Decision Process (RMDP) with large state spaces. The goal of the RMDP framework is to find a policy that is robust against the parameter uncertainties due to the mismatch between the simulator model and real-world settings. We first propose the Robust Least Squares Policy Evaluation algorithm, which is a multi-step online model-free learning algorithm for policy evaluation. We prove the convergence of this algorithm using stochastic approximation techniques. We then propose Robust Least Squares Policy Iteration (RLSPI) algorithm for learning the optimal robust policy. We also give a general weighted Euclidean norm bound on the error (closeness to optimality) of the resulting policy. Finally, we demonstrate the performance of our RLSPI algorithm on some standard benchmark problems.

cs.LG↗