12 research outputs found
An online algorithm for constrained POMDPs
No full textThis work seeks to address the problem of planning in the presence of uncertainty and constraints. Such problems arise in many situations, including the basis of this work, which involves planning for a team of first responders (both humans and robots) operating in an urban environment. The problem is framed as a Partially-Observable Markov Decision Process (POMDP) with constraints, and it is shown that even in a relatively simple planning problem, modeling constraints as large penalties does not lead to good solutions. The main contribution of the work is a new online algorithm that explicitly ensures constraint feasibility while remaining computationally tractable. Its performance is demonstrated on an example problem and it is demonstrated that our online algorithm generates policies comparable to an offline constrained POMDP algorithm.United States. Office of Naval Research (grant N00014-07-1-0749
Optimal trajectories for maneuvering reentry vehicles
No full textThesis (S.M.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 2007.Includes bibliographical references (p. 187-189).Many demanding aerospace missions today require maneuverable re-entry vehicles that can fly trajectories that have stringent path and terminal constraints, including those that cannot be written as drag or energy constraints. This work presents a method based on trajectory optimization techniques to assess the capabilities of the re-entry vehicle by computing the landing and re-entry footprints while meeting these conditions. The models used also account for important non-linear effects seen during hypersonic flight. Several different vehicles are studied, and the effects of parameters such the maximum G-loading, stagnation point heat rate, and the maximum L/D are analyzed.by Aditya Undurti.S.M
Planning under uncertainty and constraints for teams of autonomous agents
No full textThesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Aeronautics and Astronautics, 2011.Cataloged from PDF version of thesis.Includes bibliographical references (p. 155-164).One of the main advantages of unmanned, autonomous vehicles is their potential use in dangerous situations, such as victim search and rescue in the aftermath of an urban disaster. Unmanned vehicles can complement human first responders by performing tasks that do not require human expertise (e.g., communication) and supplement them by providing capabilities a human first responder would not have immediately available (e.g., aerial surveillance). However, for unmanned vehicles to work seamlessly and unintrusively with human responders, a high degree of autonomy and planning is necessary. In particular, the unmanned vehicles should be able to account for the dynamic nature of their operating environment, the uncertain nature of their tasks and outcomes, and the risks that are inherent in working in such a situation. This thesis therefore addresses the problem of planning under uncertainty in the presence of risk. This work formulates the planning problem as a Markov Decision Process with constraints, and offers a formal definition for the notion of "risk". Then, a fast and computationally efficient solution is proposed. Next, the complications that arise when planning for large teams of unmanned vehicles are considered, and a decentralized approach is investigated and shown to be efficient under some assumptions. However some of these assumptions place restrictions - specifically on the amount of risk each agent can take. These restrictions hamper individual agents' ability to adapt to a changing environment. Hence a consensus-based approach that allows agents to take more risk is introduced and shown to be effective in achieving high reward. Finally, some experimental results are presented that validate the performance of the solution techniques proposed.by Aditya Undurti.Ph.D
An intelligent cooperative control architecture
No full textThis paper presents an extension of existing cooperative control algorithms that have been developed for multi-UAV applications to utilize real-time observations and/or performance metric(s) in conjunction with learning methods to generate a more intelligent planner response. We approach this issue from a cooperative control perspective and embed elements of feedback control and active learning, resulting in an new intelligent Cooperative Control Architecture (iCCA). We describe this architecture, discuss some of the issues that must be addressed, and present illustrative examples of cooperative control problems where iCCA can be applied effectively.United States. Air Force Office of Scientific Research (grant FA9550-08-1-0086)Boeing Scientific Research Laboratorie
A decentralized approach to multi-agent planning in the presence of constraints and uncertainty
No full text
Innovations in Agricultural Forecasting: A Multivariate Regression Study on Global Crop Yield Prediction
No full textThe prediction of crop yields internationally is a crucial objective in
agricultural research. Thus, this study implements 6 regression models (Linear,
Tree, Gradient Descent, Gradient Boosting, K Nearest Neighbors, and Random
Forest) to predict crop yields in 37 developing countries over 27 years. Given
4 key training parameters, insecticides (tonnes), rainfall (mm), temperature
(Celsius), and yield (hg/ha), it was found that our Random Forest Regression
model achieved a determination coefficient (r2) of 0.94, with a margin of error
(ME) of .03. The models were trained and tested using the Food and Agricultural
Organization of the United Nations data, along with the World Bank Climate
Change Data Catalog. Furthermore, each parameter was analyzed to understand how
varying factors could impact overall yield. We used unconventional models,
contrary to generally used Deep Learning (DL) and Machine Learning (ML) models,
combined with recently collected data to implement a unique approach in our
research. Existing scholarship would benefit from understanding the most
optimal model for agricultural research, specifically using the United Nations
data.Comment: 12 pages, 8 figures, 1 table, Guided by Dr. Aditya Undurt
Pro- and anti-inflammatory bioactive lipids imbalance contributes to the pathobiology of autoimmune diseases
No full textAutoimmune diseases are driven by TH17 cells that secrete pro-inflammatory cytokines, especially IL-17. Under normal physiological conditions, autoreactive T cells are suppressed by TGF-β and IL-10 secreted by microglia and dendritic cells. When this balance is upset due to injury, infection and other causes, leukocyte recruitment and macrophage activation occurs resulting in secretion of pro-inflammatory IL-6, TNF-α, IL-17 and PGE2, LTs (leukotrienes) accompanied by a deficiency of anti-inflammatory LXA4, resolvins, protecting, and maresins. PGE2 facilitates TH1 cell differentiation and promotes immune-mediated inflammation through TH17 expansion. There is evidence to suggest that autoimmune diseases can be suppressed by anti-inflammatory bioactive lipids LXA4, resolvins, protecting, and maresins. These results imply that systemic and/or local application of LXA4, resolvins, protecting, and maresins and administration of their precursors AA/EPA/DHA could form a potential therapeutic approach in the prevention and treatment of autoimmune diseases. © 2022, The Author(s), under exclusive licence to Springer Nature Limited
An Intelligent Cooperative Control Architecture
No full textThis paper presents an extension of existing cooperative control algorithms that have been developed for multi-UAV applications to utilize real-time observations and/or performance metric(s) in conjunction with learning methods to generate a more intelligent planner response. We approach this issue from a decentralized
cooperative control perspective and embed elements of feedback control
and active learning, resulting in an new intelligent Cooperative Control Architecture (iCCA). We describe this architecture, discuss some of the issues that must be addressed, and present illustrative examples of cooperative control problems where iCCA can be applied effectively
