[ { "id": "https://authors.library.caltech.edu/records/fbfpc-exs54", "eprint_id": 120117, "eprint_status": "archive", "datestamp": "2023-08-20 08:41:24", "lastmod": "2023-10-25 16:53:07", "type": "monograph", "metadata_visibility": "show", "creators": { "items": [ { "id": "Donitz-Benjamin-P-S", "name": { "family": "Donitz", "given": "Benjamin P. S." } }, { "id": "Mages-Declan", "name": { "family": "Mages", "given": "Declan" }, "orcid": "0000-0002-2783-2144" }, { "id": "Tsukamoto-Hiroyasu", "name": { "family": "Tsukamoto", "given": "Hiroyasu" }, "orcid": "0000-0002-6337-2667" }, { "id": "Dixon-Peter", "name": { "family": "Dixon", "given": "Peter" } }, { "id": "Landau-Damon", "name": { "family": "Landau", "given": "Damon" } }, { "id": "Chung-Soon-Jo", "name": { "family": "Chung", "given": "Soon-Jo" }, "orcid": "0000-0002-6657-3907" }, { "id": "Bufanda-Erica", "name": { "family": "Bufanda", "given": "Erica" }, "orcid": "0000-0002-0406-8518" }, { "id": "Ingham-Michel-D", "name": { "family": "Ingham", "given": "Michel" }, "orcid": "0000-0001-5893-543X" }, { "id": "Castillo-Rogez-Julie-C", "name": { "family": "Castillo-Rogez", "given": "Julie" }, "orcid": "0000-0003-0400-1038" } ] }, "title": "Interstellar Object Accessibility and Mission Design", "ispublished": "unpub", "full_text_status": "public", "note": "\u00a9 2023 California Institute of Technology. \n\nThis work is being carried out in part at the Jet Propulsion Laboratory, California Institute of Technology, under contract to NASA. Government sponsorship acknowledged.\n\n
Accepted Version - 2210.14980.pdf
", "abstract": "Interstellar objects (ISOs) are fascinating and under-explored celestial objects, providing physical laboratories to understand the formation of our solar system and probe the composition and properties of material formed in exoplanetary systems. This paper will discuss the accessibility of and mission design to ISOs with varying characteristics, including a discussion of state covariance estimation over the course of a cruise, handoffs from traditional navigation approaches to novel autonomous navigation for fast flyby regimes, and overall recommendations about preparing for the future in situ exploration of these targets. The lessons learned also apply to the fast flyby of other small bodies including long-period comets and potentially hazardous asteroids, which also require a tactical response with similar characteristics.", "date": "2023-03-17", "date_type": "published", "publisher": "arXiv", "id_number": "CaltechAUTHORS:20230316-225910601", "official_url": "https://resolver.caltech.edu/CaltechAUTHORS:20230316-225910601", "rights": "No commercial reproduction, distribution, display or performance rights in this work are provided.", "funders": { "items": [ { "agency": "NASA/JPL/Caltech" } ] }, "local_group": { "items": [ { "id": "GALCIT" } ] }, "primary_object": { "basename": "2210.14980.pdf", "url": "https://authors.library.caltech.edu/records/fbfpc-exs54/files/2210.14980.pdf" }, "resource_type": "monograph", "pub_year": "2023", "author_list": "Donitz, Benjamin P. S.; Mages, Declan; et el." }, { "id": "https://authors.library.caltech.edu/records/a9fpd-t3v08", "eprint_id": 120118, "eprint_status": "archive", "datestamp": "2023-08-20 08:38:00", "lastmod": "2023-10-25 16:53:09", "type": "monograph", "metadata_visibility": "show", "creators": { "items": [ { "name": { "family": "Gan", "given": "Lu" } }, { "name": { "family": "Lee", "given": "Connor" } }, { "id": "Chung-Soon-Jo", "name": { "family": "Chung", "given": "Soon-Jo" }, "orcid": "0000-0002-6657-3907" } ] }, "title": "Unsupervised RGB-to-Thermal Domain Adaptation via Multi-Domain Attention Network", "ispublished": "unpub", "full_text_status": "public", "note": "This work is funded by Ford Motor Company and in part by the Office of Naval Research.\n\nSubmitted - 2210.04367.pdf
", "abstract": "This work presents a new method for unsupervised thermal image classification and semantic segmentation by transferring knowledge from the RGB domain using a multi-domain attention network. Our method does not require any thermal annotations or co-registered RGB-thermal pairs, enabling robots to perform visual tasks at night and in adverse weather conditions without incurring additional costs of data labeling and registration. Current unsupervised domain adaptation methods look to align global images or features across domains. However, when the domain shift is significantly larger for cross-modal data, not all features can be transferred. We solve this problem by using a shared backbone network that promotes generalization, and domain-specific attention that reduces negative transfer by attending to domain-invariant and easily-transferable features. Our approach outperforms the state-of-the-art RGB-to-thermal adaptation method in classification benchmarks, and is successfully applied to thermal river scene segmentation using only synthetic RGB images. Our code is made publicly available at https://github.com/ganlumomo/thermal-uda-attention.", "date": "2023-03-17", "date_type": "published", "publisher": "arXiv", "id_number": "CaltechAUTHORS:20230316-225914015", "official_url": "https://resolver.caltech.edu/CaltechAUTHORS:20230316-225914015", "rights": "No commercial reproduction, distribution, display or performance rights in this work are provided.", "funders": { "items": [ { "agency": "Ford Motor Company" }, { "agency": "Office of Naval Research (ONR)" } ] }, "local_group": { "items": [ { "id": "GALCIT" } ] }, "primary_object": { "basename": "2210.04367.pdf", "url": "https://authors.library.caltech.edu/records/a9fpd-t3v08/files/2210.04367.pdf" }, "resource_type": "monograph", "pub_year": "2023", "author_list": "Gan, Lu; Lee, Connor; et el." }, { "id": "https://authors.library.caltech.edu/records/df5rt-vae03", "eprint_id": 120119, "eprint_status": "archive", "datestamp": "2023-08-20 08:22:50", "lastmod": "2023-10-25 16:53:12", "type": "monograph", "metadata_visibility": "show", "creators": { "items": [ { "id": "Tsukamoto-Hiroyasu", "name": { "family": "Tsukamoto", "given": "Hiroyasu" }, "orcid": "0000-0002-6337-2667" }, { "id": "Chung-Soon-Jo", "name": { "family": "Chung", "given": "Soon-Jo" }, "orcid": "0000-0002-6657-3907" }, { "id": "Donitz-Benjamin-P-S", "name": { "family": "Donitz", "given": "Benjamin" } }, { "id": "Ingham-Michel-D", "name": { "family": "Ingham", "given": "Michel" }, "orcid": "0000-0001-5893-543X" }, { "id": "Mages-Declan", "name": { "family": "Mages", "given": "Declan" }, "orcid": "0000-0002-2783-2144" }, { "id": "Nakka-Yashwanth-Kumar-K", "name": { "family": "Nakka", "given": "Yashwanth Kumar" }, "orcid": "0000-0001-7897-3644" } ] }, "title": "Neural-Rendezvous: Learning-based Robust Guidance and Control to Encounter Interstellar Objects", "ispublished": "unpub", "full_text_status": "public", "note": "We thank Stefano Campagnola (NASA JPL) for providing his useful simulation codes utilized in Sec. VII, and thank Julie Castillo-Rogez (NASA JPL), Fred Y. Hadaegh (NASA JPL), and Karen Meech (University of Hawaii), and Robert Jedicke (University of Hawaii) for their insightful inputs and technical discussions. \n\nThis work is funded by the NASA Jet Propulsion Laboratory, California Institute of Technology.\n\nSubmitted - 2208.04883.pdf
", "abstract": "Interstellar objects (ISOs), astronomical objects not gravitationally bound to the Sun, are likely representatives of primitive materials invaluable in understanding exoplanetary star systems. Due to their poorly constrained orbits with generally high inclinations and relative velocities, however, exploring ISOs with conventional human-in-the-loop approaches is significantly challenging. This paper presents Neural-Rendezvous -- a deep learning-based guidance and control framework for encountering any fast-moving objects, including ISOs, robustly, accurately, and autonomously in real-time. It uses pointwise minimum norm tracking control on top of a guidance policy modeled by a spectrally-normalized deep neural network, where its hyperparameters are tuned with a newly introduced loss function directly penalizing the state trajectory tracking error. We rigorously show that, even in the challenging case of ISO exploration, Neural-Rendezvous provides 1) a high probability exponential bound on the expected spacecraft delivery error; and 2) a finite optimality gap with respect to the solution of model predictive control, both of which are indispensable especially for such a critical space mission. In numerical simulations, Neural-Rendezvous is demonstrated to achieve a terminal-time delivery error of less than 0.2 km for 99% of the ISO candidates with realistic state uncertainty, whilst retaining computational efficiency sufficient for real-time implementation.", "date": "2022-08-09", "date_type": "published", "publisher": "arXiv", "id_number": "CaltechAUTHORS:20230316-225917431", "official_url": "https://resolver.caltech.edu/CaltechAUTHORS:20230316-225917431", "rights": "No commercial reproduction, distribution, display or performance rights in this work are provided.", "funders": { "items": [ { "agency": "NASA/JPL/Caltech" } ] }, "local_group": { "items": [ { "id": "GALCIT" } ] }, "primary_object": { "basename": "2208.04883.pdf", "url": "https://authors.library.caltech.edu/records/df5rt-vae03/files/2208.04883.pdf" }, "resource_type": "monograph", "pub_year": "2022", "author_list": "Tsukamoto, Hiroyasu; Chung, Soon-Jo; et el." }, { "id": "https://authors.library.caltech.edu/records/d50xx-egb52", "eprint_id": 109917, "eprint_status": "archive", "datestamp": "2023-08-20 03:37:18", "lastmod": "2023-10-23 18:12:50", "type": "monograph", "metadata_visibility": "show", "creators": { "items": [ { "id": "Nakka-Yashwanth-Kumar-K", "name": { "family": "Nakka", "given": "Yashwanth Kumar" }, "orcid": "0000-0001-7897-3644" }, { "id": "Chung-Soon-Jo", "name": { "family": "Chung", "given": "Soon-Jo" }, "orcid": "0000-0002-6657-3907" } ] }, "title": "Trajectory Optimization of Chance-Constrained Nonlinear Stochastic Systems for Motion Planning and Control", "ispublished": "unpub", "full_text_status": "public", "note": "Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) \n\nThis work supported in part by Jet Propulsion Laboratory. \n\nThe authors are thankful to Amir Rahmani, Fred Y. Hadaegh, Joel Burdick, Richard Murray and Yisong Yue for stimulating discussions and technical help.\n\nSubmitted - 2106.02801.pdf
", "abstract": "We present gPC-SCP: Generalized Polynomial Chaos-based Sequential Convex Programming method to compute a sub-optimal solution for a continuous-time chance-constrained stochastic nonlinear optimal control problem (SNOC) problem. The approach enables motion planning and control of robotic systems under uncertainty. The proposed method involves two steps. The first step is to derive a deterministic nonlinear optimal control problem (DNOC) with convex constraints that are surrogate to the SNOC by using gPC expansion and the distributionally-robust convex subset of the chance constraints. The second step is to solve the DNOC problem using sequential convex programming (SCP) for trajectory generation and control. We prove that in the unconstrained case, the optimal value of the DNOC converges to that of SNOC asymptotically and that any feasible solution of the constrained DNOC is a feasible solution of the chance-constrained SNOC. We derive a stable stochastic model predictive controller using the gPC-SCP for tracking a trajectory in the presence of uncertainty. We empirically demonstrate the efficacy of the gPC-SCP method for the following three test cases: 1) collision checking under uncertainty in actuation, 2) collision checking with stochastic obstacle model, and 3) safe trajectory tracking under uncertainty in the dynamics and obstacle location by using a receding horizon control approach. We validate the effectiveness of the gPC-SCP method on the robotic spacecraft testbed.", "date": "2021-06-05", "date_type": "published", "publisher": "arXiv", "id_number": "CaltechAUTHORS:20210719-210132481", "official_url": "https://resolver.caltech.edu/CaltechAUTHORS:20210719-210132481", "rights": "No commercial reproduction, distribution, display or performance rights in this work are provided.", "funders": { "items": [ { "agency": "JPL" } ] }, "local_group": { "items": [ { "id": "GALCIT" } ] }, "doi": "10.48550/arXiv.2106.02801", "primary_object": { "basename": "2106.02801.pdf", "url": "https://authors.library.caltech.edu/records/d50xx-egb52/files/2106.02801.pdf" }, "resource_type": "monograph", "pub_year": "2021", "author_list": "Nakka, Yashwanth Kumar and Chung, Soon-Jo" }, { "id": "https://authors.library.caltech.edu/records/rhcjt-68j20", "eprint_id": 107617, "eprint_status": "archive", "datestamp": "2023-08-20 00:29:45", "lastmod": "2023-12-13 17:03:01", "type": "monograph", "metadata_visibility": "show", "creators": { "items": [ { "id": "Cai-Karena-X", "name": { "family": "Cai", "given": "Karena X." } }, { "id": "Phan-Minh-Tung", "name": { "family": "Phan-Minh", "given": "Tung" } }, { "id": "Chung-Soon-Jo", "name": { "family": "Chung", "given": "Soon-Jo" }, "orcid": "0000-0002-6657-3907" }, { "id": "Murray-R-M", "name": { "family": "Murray", "given": "Richard M." }, "orcid": "0000-0002-5785-7481" } ] }, "title": "Rules of the Road: Towards Safety and Liveness Guarantees for Autonomous Vehicles", "ispublished": "unpub", "full_text_status": "public", "note": "Attribution 4.0 International (CC BY 4.0).\n\nThis research supported by the National Science Foundation award CNS-1545126. \n\nWe would like to acknowledge K. Mani Chandy who provided valuable input to the problem formulation and presentation of ideas in the manuscript and to Giovanna Amorim for her contributions to the simulation code. \n\nAUTHOR CONTRIBUTIONS. K.X.C., R.M.M., and T.P-M. jointly conceived the conceptual framework. K.X.C. and T.P-M. jointly developed the problem formulation and theoretical approach. K.X.C. worked out the main proofs with input from T.P-M. K.X.C. drafted the manuscript and figures with input from T.P-M. S-J.C. and R.M.M. provided guidance on the overall approach and provided feedback on the final manuscript.\n\nSubmitted - 2011.14148.pdf
", "abstract": "The ability to guarantee safety and progress for all vehicles is vital to the success of the autonomous vehicle industry. We present a framework for the distributed control of autonomous vehicles that is safe and guarantees progress for all agents. In this paper, we first introduce a new game paradigm which we term the quasi-simultaneous discrete-time game. We then define an Agent Protocol agents must use to make decisions in this quasi-simultaneous discrete-time game setting. According to the protocol, agents first select an intended action and then each agent determines whether it can take its intended action or not, given its proposed intention and the intentions of nearby agents. The protocol so defined will ensure safety under all traffic conditions and liveness for all agents under \"sparse\" traffic conditions. These guarantees, however, are predicated on the premise that all agents are operating with the aforementioned protocol. We provide proofs of correctness of the protocol and validate our results in simulation.", "date": "2021-01-21", "date_type": "published", "publisher": "arXiv", "id_number": "CaltechAUTHORS:20210120-165255737", "official_url": "https://resolver.caltech.edu/CaltechAUTHORS:20210120-165255737", "rights": "No commercial reproduction, distribution, display or performance rights in this work are provided.", "funders": { "items": [ { "agency": "NSF", "grant_number": "CNS-1545126" } ] }, "local_group": { "items": [ { "id": "Division-of-Biology-and-Biological-Engineering" } ] }, "doi": "10.48550/arXiv.2011.14148", "primary_object": { "basename": "2011.14148.pdf", "url": "https://authors.library.caltech.edu/records/rhcjt-68j20/files/2011.14148.pdf" }, "resource_type": "monograph", "pub_year": "2021", "author_list": "Cai, Karena X.; Phan-Minh, Tung; et el." }, { "id": "https://authors.library.caltech.edu/records/nnyhw-95c64", "eprint_id": 106583, "eprint_status": "archive", "datestamp": "2023-08-19 23:57:34", "lastmod": "2023-10-20 23:36:27", "type": "monograph", "metadata_visibility": "show", "creators": { "items": [ { "id": "Yu-Chenkai", "name": { "family": "Yu", "given": "Chenkai" } }, { "id": "Shi-Guanya", "name": { "family": "Shi", "given": "Guanya" } }, { "id": "Chung-Soon-Jo", "name": { "family": "Chung", "given": "Soon-Jo" }, "orcid": "0000-0002-6657-3907" }, { "id": "Yue-Yisong", "name": { "family": "Yue", "given": "Yisong" }, "orcid": "0000-0001-9127-1989" }, { "id": "Wierman-A", "name": { "family": "Wierman", "given": "Adam" } } ] }, "title": "Competitive Control with Delayed Imperfect Information", "ispublished": "unpub", "full_text_status": "public", "note": "Submitted - 2010.11637.pdf
", "abstract": "This paper studies the impact of imperfect information in online control with adversarial disturbances. In particular, we consider both delayed state feedback and inexact predictions of future disturbances. We introduce a greedy, myopic policy that yields a constant competitive ratio against the offline optimal policy with delayed feedback and inexact predictions. A special case of our result is a constant competitive policy for the case of exact predictions and no delay, a previously open problem. We also analyze the fundamental limits of online control with limited information by showing that our competitive ratio bounds for the greedy, myopic policy in the adversarial setting match (up to lower-order terms) lower bounds in the stochastic setting.", "date": "2020-11-10", "date_type": "published", "publisher": "arXiv", "id_number": "CaltechAUTHORS:20201110-082106076", "official_url": "https://resolver.caltech.edu/CaltechAUTHORS:20201110-082106076", "rights": "No commercial reproduction, distribution, display or performance rights in this work are provided.", "local_group": { "items": [ { "id": "GALCIT" } ] }, "doi": "10.48550/arXiv.2010.11637", "primary_object": { "basename": "2010.11637.pdf", "url": "https://authors.library.caltech.edu/records/nnyhw-95c64/files/2010.11637.pdf" }, "resource_type": "monograph", "pub_year": "2020", "author_list": "Yu, Chenkai; Shi, Guanya; et el." }, { "id": "https://authors.library.caltech.edu/records/3xzsq-ybr76", "eprint_id": 106578, "eprint_status": "archive", "datestamp": "2023-08-19 23:31:59", "lastmod": "2023-10-20 23:36:13", "type": "monograph", "metadata_visibility": "show", "creators": { "items": [ { "id": "Shi-Xichen", "name": { "family": "Shi", "given": "Xichen" } }, { "id": "O'Connell-Michael", "name": { "family": "O'Connell", "given": "Michael" } }, { "id": "Chung-Soon-Jo", "name": { "family": "Chung", "given": "Soon-Jo" }, "orcid": "0000-0002-6657-3907" } ] }, "title": "Numerical Predictive Control for Delay Compensation", "ispublished": "unpub", "full_text_status": "public", "note": "Submitted - 2009.14450.pdf
", "abstract": "We present a delay-compensating control method that transforms exponentially stabilizing controllers for an undelayed system into a sample-based predictive controller with numerical integration. Our method handles both first-order and transport delays in actuators and trades-off numerical accuracy with computation delay to guaranteed stability under hardware limitations. Through hybrid stability analysis and numerical simulation, we demonstrate the efficacy of our method from both theoretical and simulation perspectives.", "date": "2020-11-10", "date_type": "published", "publisher": "arXiv", "id_number": "CaltechAUTHORS:20201110-074747495", "official_url": "https://resolver.caltech.edu/CaltechAUTHORS:20201110-074747495", "rights": "No commercial reproduction, distribution, display or performance rights in this work are provided.", "local_group": { "items": [ { "id": "GALCIT" } ] }, "doi": "10.48550/arXiv.2009.14450", "primary_object": { "basename": "2009.14450.pdf", "url": "https://authors.library.caltech.edu/records/3xzsq-ybr76/files/2009.14450.pdf" }, "resource_type": "monograph", "pub_year": "2020", "author_list": "Shi, Xichen; O'Connell, Michael; et el." }, { "id": "https://authors.library.caltech.edu/records/8fk4n-b1212", "eprint_id": 106599, "eprint_status": "archive", "datestamp": "2023-08-19 23:55:06", "lastmod": "2023-10-20 23:37:07", "type": "monograph", "metadata_visibility": "show", "creators": { "items": [ { "id": "Paranjape-A-A", "name": { "family": "Paranjape", "given": "Aditya A." }, "orcid": "0000-0002-3164-3215" }, { "id": "Chung-Soon-Jo", "name": { "family": "Chung", "given": "Soon-Jo" }, "orcid": "0000-0002-6657-3907" } ] }, "title": "Sub-Optimality of a Dyadic Adaptive Control Architecture", "ispublished": "unpub", "full_text_status": "public", "note": "Submitted - 2010.10329.pdf
", "abstract": "The dyadic adaptive control architecture evolved as a solution to the problem of designing control laws for nonlinear systems with unmatched nonlinearities, disturbances and uncertainties. A salient feature of this framework is its ability to work with infinite as well as finite dimensional systems, and with a wide range of control and adaptive laws. In this paper, we consider the case where a control law based on the linear quadratic regulator theory is employed for designing the control law. We benchmark the closed-loop system against standard linear quadratic control laws as well as those based on the state-dependent Riccati equation. We pose the problem of designing a part of the control law as a Nehari problem. We obtain analytical expressions for the bounds on the sub-optimality of the control law.", "date": "2020-11-10", "date_type": "published", "publisher": "arXiv", "id_number": "CaltechAUTHORS:20201110-154646316", "official_url": "https://resolver.caltech.edu/CaltechAUTHORS:20201110-154646316", "rights": "No commercial reproduction, distribution, display or performance rights in this work are provided.", "local_group": { "items": [ { "id": "GALCIT" } ] }, "doi": "10.48550/arXiv.2010.10329", "primary_object": { "basename": "2010.10329.pdf", "url": "https://authors.library.caltech.edu/records/8fk4n-b1212/files/2010.10329.pdf" }, "resource_type": "monograph", "pub_year": "2020", "author_list": "Paranjape, Aditya A. and Chung, Soon-Jo" }, { "id": "https://authors.library.caltech.edu/records/vk7vz-zp479", "eprint_id": 104236, "eprint_status": "archive", "datestamp": "2023-08-19 21:51:40", "lastmod": "2023-10-20 19:12:00", "type": "monograph", "metadata_visibility": "show", "creators": { "items": [ { "id": "Yu-Chenkai", "name": { "family": "Yu", "given": "Chenkai" }, "orcid": "0000-0001-8683-7773" }, { "id": "Shi-Guanya", "name": { "family": "Shi", "given": "Guanya" }, "orcid": "0000-0002-9075-3705" }, { "id": "Chung-Soon-Jo", "name": { "family": "Chung", "given": "Soon-Jo" }, "orcid": "0000-0002-6657-3907" }, { "id": "Yue-Yisong", "name": { "family": "Yue", "given": "Yisong" }, "orcid": "0000-0001-9127-1989" }, { "id": "Wierman-A", "name": { "family": "Wierman", "given": "Adam" }, "orcid": "0000-0002-5923-0199" } ] }, "title": "The Power of Predictions in Online Control", "ispublished": "unpub", "full_text_status": "public", "note": "This project was supported in part by funding from Raytheon, DARPA PAI, AitF-1637598 and CNS-1518941, with additional support for Guanya Shi provided by the Simoudis Discovery Prize. \n\nWe see no ethical concerns related to the results in this paper.\n\nAccepted Version - 2006.07569.pdf
", "abstract": "We study the impact of predictions in online Linear Quadratic Regulator control with both stochastic and adversarial disturbances in the dynamics. In both settings, we characterize the optimal policy and derive tight bounds on the minimum cost and dynamic regret. Perhaps surprisingly, our analysis shows that the conventional greedy MPC approach is a near-optimal policy in both stochastic and adversarial settings. Specifically, for length-T problems, MPC requires only O(logT) predictions to reach O(1) dynamic regret, which matches (up to lower-order terms) our lower bound on the required prediction horizon for constant regret.", "date": "2020-07-07", "date_type": "published", "publisher": "arXiv", "id_number": "CaltechAUTHORS:20200707-094715120", "official_url": "https://resolver.caltech.edu/CaltechAUTHORS:20200707-094715120", "rights": "No commercial reproduction, distribution, display or performance rights in this work are provided.", "funders": { "items": [ { "agency": "Raytheon Company" }, { "agency": "Defense Advanced Research Projects Agency (DARPA)" }, { "agency": "NSF", "grant_number": "CCF-1637598" }, { "agency": "NSF", "grant_number": "CNS-1518941" }, { "agency": "Simoudis Discovery Prize" } ] }, "local_group": { "items": [ { "id": "GALCIT" } ] }, "doi": "10.48550/arXiv.2006.07569", "primary_object": { "basename": "2006.07569.pdf", "url": "https://authors.library.caltech.edu/records/vk7vz-zp479/files/2006.07569.pdf" }, "resource_type": "monograph", "pub_year": "2020", "author_list": "Yu, Chenkai; Shi, Guanya; et el." }, { "id": "https://authors.library.caltech.edu/records/xrywz-v0w33", "eprint_id": 101303, "eprint_status": "archive", "datestamp": "2023-08-19 19:58:08", "lastmod": "2023-10-19 22:36:37", "type": "monograph", "metadata_visibility": "show", "creators": { "items": [ { "id": "Shi-Guanya", "name": { "family": "Shi", "given": "Guanya" }, "orcid": "0000-0002-9075-3705" }, { "id": "Lin-Yiheng", "name": { "family": "Lin", "given": "Yiheng" }, "orcid": "0000-0001-6524-2877" }, { "id": "Chung-Soon-Jo", "name": { "family": "Chung", "given": "Soon-Jo" }, "orcid": "0000-0002-6657-3907" }, { "id": "Yue-Yisong", "name": { "family": "Yue", "given": "Yisong" }, "orcid": "0000-0001-9127-1989" }, { "id": "Wierman-A", "name": { "family": "Wierman", "given": "Adam" }, "orcid": "0000-0002-5923-0199" } ] }, "title": "Online Optimization with Memory and Competitive Control", "ispublished": "unpub", "full_text_status": "public", "note": "This project was supported in part by funding from Raytheon, DARPA PAI, AitF-1637598 and CNS-1518941, with additional support for Guanya Shi provided by the Simoudis Discovery Prize. \n\nWe see no ethical concerns related to the results in this paper.\n\nAccepted Version - 2002.05318.pdf
", "abstract": "This paper presents competitive algorithms for a novel class of online optimization problems with memory. We consider a setting where the learner seeks to minimize the sum of a hitting cost and a switching cost that depends on the previous p decisions. This setting generalizes Smoothed Online Convex Optimization. The proposed approach, Optimistic Regularized Online Balanced Descent, achieves a constant, dimension-free competitive ratio. Further, we show a connection between online optimization with memory and online control with adversarial disturbances. This connection, in turn, leads to a new constant-competitive policy for a rich class of online control problems.", "date": "2020-02-14", "date_type": "published", "publisher": "arXiv", "id_number": "CaltechAUTHORS:20200214-105606928", "official_url": "https://resolver.caltech.edu/CaltechAUTHORS:20200214-105606928", "rights": "No commercial reproduction, distribution, display or performance rights in this work are provided.", "funders": { "items": [ { "agency": "Raytheon Company" }, { "agency": "Defense Advanced Research Projects Agency (DARPA)" }, { "agency": "NSF", "grant_number": "CCF-1637598" }, { "agency": "NSF", "grant_number": "CNS-1518941" }, { "agency": "Simoudis Discovery Prize" } ] }, "local_group": { "items": [ { "id": "GALCIT" } ] }, "doi": "10.48550/arXiv.2002.05318", "primary_object": { "basename": "2002.05318.pdf", "url": "https://authors.library.caltech.edu/records/xrywz-v0w33/files/2002.05318.pdf" }, "resource_type": "monograph", "pub_year": "2020", "author_list": "Shi, Guanya; Lin, Yiheng; et el." }, { "id": "https://authors.library.caltech.edu/records/wjr0t-sh732", "eprint_id": 99547, "eprint_status": "archive", "datestamp": "2023-08-19 17:40:18", "lastmod": "2023-10-18 18:33:01", "type": "monograph", "metadata_visibility": "show", "creators": { "items": [ { "id": "O'Connell-Michael", "name": { "family": "O'Connell", "given": "Michael" } }, { "id": "Shi-Guanya", "name": { "family": "Shi", "given": "Guanya" }, "orcid": "0000-0002-9075-3705" }, { "id": "Shi-Xichen", "name": { "family": "Shi", "given": "Xichen" } }, { "id": "Chung-Soon-Jo", "name": { "family": "Chung", "given": "Soon-Jo" }, "orcid": "0000-0002-6657-3907" } ] }, "title": "Meta-Learning-Based Robust Adaptive Flight Control Under Uncertain Wind Conditions", "ispublished": "unpub", "full_text_status": "public", "note": "We thank Yisong Yue, Animashree Anandkumar, Kamyar Azizzadenesheli, Joel Burdick, Mory Gharib, Daniel Pastor Moreno, and Anqi Liu for helpful discussions. The work is funded in part by Caltech's Center for Autonomous Systems and Technologies and Raytheon Company.\n\nSubmitted - 2103.01932v1.pdf
", "abstract": "Realtime model learning proves challenging for complex dynamical systems, such as drones flying in variable wind conditions. Machine learning technique such as deep neural networks have high representation power but is often too slow to update onboard. On the other hand, adaptive control relies on simple linear parameter models can update as fast as the feedback control loop. We propose an online composite adaptation method that treats outputs from a deep neural network as a set of basis functions capable of representing different wind conditions. To help with training, meta-learning techniques are used to optimize the network output useful for adaptation. We validate our approach by flying a drone in an open air wind tunnel under varying wind conditions and along challenging trajectories. We compare the result with other adaptive controller with different basis function sets and show improvement over tracking and prediction errors.", "date": "2019-10-29", "date_type": "published", "id_number": "CaltechAUTHORS:20191029-154625952", "official_url": "https://resolver.caltech.edu/CaltechAUTHORS:20191029-154625952", "rights": "No commercial reproduction, distribution, display or performance rights in this work are provided.", "funders": { "items": [ { "agency": "Center for Autonomous Systems and Technologies" }, { "agency": "Raytheon Company" } ] }, "local_group": { "items": [ { "id": "GALCIT" }, { "id": "Center-for-Autonomous-Systems-and-Technologies-(CAST)" } ] }, "doi": "10.48550/arXiv.2103.01932", "primary_object": { "basename": "2103.01932v1.pdf", "url": "https://authors.library.caltech.edu/records/wjr0t-sh732/files/2103.01932v1.pdf" }, "resource_type": "monograph", "pub_year": "2019", "author_list": "O'Connell, Michael; Shi, Guanya; et el." }, { "id": "https://authors.library.caltech.edu/records/607xd-w9259", "eprint_id": 98458, "eprint_status": "archive", "datestamp": "2023-08-19 16:18:06", "lastmod": "2023-10-18 17:23:00", "type": "monograph", "metadata_visibility": "show", "creators": { "items": [ { "id": "Liu-Anqi", "name": { "family": "Liu", "given": "Anqi" } }, { "id": "Shi-Guanya", "name": { "family": "Shi", "given": "Guanya" } }, { "id": "Chung-Soon-Jo", "name": { "family": "Chung", "given": "Soon-Jo" }, "orcid": "0000-0002-6657-3907" }, { "id": "Anandkumar-A", "name": { "family": "Anandkumar", "given": "Anima" } }, { "id": "Yue-Yisong", "name": { "family": "Yue", "given": "Yisong" }, "orcid": "0000-0001-9127-1989" } ] }, "title": "Robust Regression for Safe Exploration in Control", "ispublished": "unpub", "full_text_status": "public", "note": "Submitted - 1906.05819.pdf
", "abstract": "We study the problem of safe learning and exploration in sequential control problems. The goal is to safely collect data samples from an operating environment to learn an optimal controller. A central challenge in this setting is how to quantify uncertainty in order to choose provably-safe actions that allow us to collect useful data and reduce uncertainty, thereby achieving both improved safety and optimality. To address this challenge, we present a deep robust regression model that is trained to directly predict the uncertainty bounds for safe exploration. We then show how to integrate our robust regression approach with model-based control methods by learning a dynamic model with robustness bounds. We derive generalization bounds under domain shifts for learning and connect them with safety and stability bounds in control. We demonstrate empirically that our robust regression approach can outperform conventional Gaussian process (GP) based safe exploration in settings where it is difficult to specify a good GP prior.", "date": "2019-09-05", "date_type": "published", "publisher": "arXiv", "id_number": "CaltechAUTHORS:20190905-154307157", "official_url": "https://resolver.caltech.edu/CaltechAUTHORS:20190905-154307157", "rights": "No commercial reproduction, distribution, display or performance rights in this work are provided.", "doi": "10.48550/arXiv.1906.05819", "primary_object": { "basename": "1906.05819.pdf", "url": "https://authors.library.caltech.edu/records/607xd-w9259/files/1906.05819.pdf" }, "resource_type": "monograph", "pub_year": "2019", "author_list": "Liu, Anqi; Shi, Guanya; et el." } ]