282,164 research outputs found
Analysis of gradient projection anti-windup scheme
No full textThe gradient projection anti-windup (GPAW) scheme was recently proposed as an anti-windup method for nonlinear multi-input-multi-output systems/controllers, the solution of which was recognized as a largely open problem in a recent survey paper. This paper analyzes the properties of the GPAW scheme applied to an input constrained first order linear time invariant (LTI) system driven by a first order LTI controller, where the objective is to regulate the system state about the origin. We show that the GPAW compensated system is in fact a projected dynamical system (PDS), and use results in the PDS literature to assert existence and uniqueness of its solutions. The main result is that the GPAW scheme can only maintain/enlarge the exact region of attraction of the uncompensated system.Singapore. DSO National LaboratoriesUnited States. Air Force Office of Scientific Research (grant FA9550-08-1-0086
Geometric properties of gradient projection anti-windup compensated systems
No full textThe gradient projection anti-windup (GPAW) scheme was recently proposed as an anti-windup method for nonlinear multi-input-multi-output systems/controllers, which was recognized as a largely open problem in a recent survey paper. Here, we show that for controllers whose output equation depends only on its state, the GPAW compensated controller achieves exact state-output consistency when appropriately initialized. In a related paper analyzing the GPAW scheme on a simple constrained system, this property was crucial in proving that the GPAW scheme can only maintain/enlarge the exact region of attraction of the uncompensated system. When the nominal controller does not have the required structure, an arbitrarily close approximating controller can be constructed. Further geometric properties of GPAW compensated systems are then presented, which illuminates the role of the GPAW tuning parameter.United States. Air Force Office of Scientific Research (grant FA9550-08-1-0086)Singapore. DSO National Laboratorie
Bayesian Nonparametric Inverse Reinforcement Learning
No full textInverse reinforcement learning (IRL) is the task of learning the reward function of a Markov Decision Process (MDP) given the transition function and a set of observed demonstrations in the form of state-action pairs. Current IRL algorithms attempt to find a single reward function which explains the entire observation set. In practice, this leads to a computationally-costly search over a large (typically infinite) space of complex reward functions. This paper proposes the notion that if the observations can be partitioned into smaller groups, a class of much simpler reward functions can be used to explain each group. The proposed method uses a Bayesian nonparametric mixture model to automatically partition the data and find a set of simple reward functions corresponding to each partition. The simple rewards are interpreted intuitively as subgoals, which can be used to predict actions or analyze which states are important to the demonstrator. Experimental results are given for simple examples showing comparable performance to other IRL algorithms in nominal situations. Moreover, the proposed method handles cyclic tasks (where the agent begins and ends in the same state) that would break existing algorithms without modification. Finally, the new algorithm has a fundamentally different structure than previous methods, making it more computationally efficient in a real-world learning scenario where the state space is large but the demonstration set is small
Carthy J. D. — Animal navigation. How animals find, their way about. London, Allen and Urwin, 1956
No full textP. J. Carthy J. D. — Animal navigation. How animals find, their way about. London, Allen and Urwin, 1956. In: La Terre et La Vie, Revue d'Histoire naturelle, tome 11, n°1, 1957. p. 88
A Human-Interactive Course of Action Planner for Aircraft Carrier Deck Operations
No full textAircraft carrier deck operations present a complex and uncertain environment in which time-critical scheduling and planning must be done, and to date all course of action planning is done solely by human operators who rely on experience and training to safely negotiate off -nominal situations. A computer decision support system could provide the operator with both a vital resource in emergency scenarios as well as suggestions to improve e fficiency during normal operations. Such a decision support system would generate a schedule of coordinated deck operations for all active aircraft (taxi, refuel, take o ff, queue in Marshal stack, land, etc.) that is optimized for effi ciency, amenable to the operator, and robust to the many types of uncertainty inherent in the aircraft carrier deck environment. This paper describes the design, implementation, and testing of a human-interactive aircraft carrier deck course of action planner. The planning problem is cast in the MDP framework such that a wide range of current literature can be used to fi nd an optimal policy. It is designed such that human operators can specify priority aircraft and suggest scheduling orders. Inverse reinforcement learning techniques are applied that allow the planner to learn from recorded expert demonstrations. Results are presented that compare various types of human and learned policies, and show qualitative and quantitative matching between expert demonstrations and learned policies.United States. Office of Naval Research (Science of Autonomy Program
L1 Adaptive Control for Indoor Autonomous Vehicles: Design Process and Flight Testing
No full textAdaptive control techniques have the potential to address many of the special
performance and robustness requirements of flight control for unmanned aerial vehicles. L[subscript 1] adaptive control offers potential benefits in terms of performance and
robustness. An L[subscript 1] adaptive output feedback control design process is presented
here in which control parameters are systematically determined based on intuitive
desired performance and robustness metrics set by the designer. Flight test results
verify the process for an indoor autonomous quadrotor helicopter, demonstrating that designer specifications correspond to the expected physical responses. In
flight tests comparing it with the baseline linear controller, the augmented adaptive system shows definite performance and robustness improvements conforming
the potential of L[subscript 1] adaptive control as a useful tool for autonomous aircraft.United States. Air Force Office of Scientific Research (Grant FA9550-08-1-0086
The safety and effectiveness of different methods of ear wax removal: a systematic review and economic evaluation
No full textEar wax (cerumen) is a natural secretion produced to protect the inner ear from dirt and other fragments by moving these particles towards the outer ear. If this process does not happen properly, wax may build up causing blockage in the ear canal and the possibility of impaction. People with a build up of ear wax may suffer from hearing loss, discomfort and, on occasions, infection. It may present problems in assessing hearing, blocking the view of the ear drum during medical examination and interfering with the fitting or function of hearing aids. Although it is thought to affect between 2% and 6% of the population in the England and Wales, some groups may be at a higher risk, such as those using hearing aids or with small ear canals and/or skin conditions. Recurrence is thought to be high among some of these groups. The consequences of the build up of ear wax in the ear canal are thought to be a common reason for consultation and cost in general practice with over 2 million consultations per year in the NHS.Methods of removal of ear wax include drops, flushing with water in general practice, and removal with suction or probes in specialist clinics. The relative safety and benefits of these different methods of removal remains uncertain. This research will systematically review published and unpublished evidence on the clinical and cost effectiveness of different methods for the removal of ear wax. Where appropriate, it will develop an economic model using data from this systematic review and other relevant sources to estimate the relative costs and benefits of different methods. In addition, the project will provide recommendations for future research to try to help answer any remaining areas of uncertainty
Dispelling the Myths Behind First-author Citation Counts
No full textWe conducted a full-scale evaluative citation analysis study of scholars in the XML research field to explore just how different from each other author rankings resulting from different citation counting methods actually are, and to demonstrate the capability of emerging data and tools on the Web in supporting more realistic citation counting methods. Our results contest some common arguments for the continued
use of first-author citation counts in the evaluation of scholars, such as high correlations between author rankings by first-author citation counts and other citation
counting methods, and high costs of using more realistic citation counting methods that are not well-supported by the ISI databases. It is argued that increasingly available digital full text research papers make it possible for citation analysis studies to go beyond what the ISI databases have directly supported and to employ more
sophisticated methods
Scalable reward learning from demonstration
No full textReward learning from demonstration is the task of inferring the intents or goals of an agent demonstrating a task. Inverse reinforcement learning methods utilize the Markov decision process (MDP) framework to learn rewards, but typically scale poorly since they rely on the calculation of optimal value functions. Several key modifications are made to a previously developed Bayesian nonparametric inverse reinforcement learning algorithm that avoid calculation of an optimal value function and no longer require discretization of the state or action spaces. Experimental results given demonstrate the ability of the resulting algorithm to scale to larger problems and learn in domains with continuous demonstrations.United States. Office of Naval Research (Autonomy Program Contract N000140910625
The subzero microbiome: Microbial activity in frozen and thawing soils
No full textMost of the Earth's biosphere is characterized by low temperatures (<5 °C) and cold-adapted microorganisms are widespread. These psychrophiles have evolved a complex range of adaptations of all cellular constituents to counteract the potentially deleterious effects of low kinetic energy environments and the freezing of water. Microbial life continues into the subzero temperature range, and this activity contributes to carbon and nitrogen flux in and out of ecosystems, ultimately affecting global processes. Microbial responses to climate warming and in particular, thawing of frozen soils are not yet well understood although the threat of microbial contribution to positive feedback of carbon flux is substantial. To date, several studies have examined microbial community dynamics in frozen soils and permafrost due to changing environmental conditions, and some have undertaken the complicated task of characterizing microbial functional groups and how their activity changes with changing conditions, either in situ or by isolating and characterizing macromolecules. With increasing temperature and wetter conditions microbial activity of key microbes and subsequent efflux of greenhouse gases also increase. In this review, we aim to provide an overview of microbial activity in seasonally frozen soils and permafrost. With a more detailed understanding of the microbiological activities in these vulnerable soil ecosystems, we can begin to predict and model future expectations for carbon release and climate change.Peer reviewe
- …
