1,721,050 research outputs found
Vision-based relative navigation system for autonomous proximity orbital operations
No full textThe paper presents the research project of the author, focused on developing, implementing and testing a vision based relative navigation system for spacecraft. A temporal organization of the project is presented, with tasks assigned to each year of the PhD programme, while at the same time two main technical stages, “final” and “far” rendezvous are introduced together with their scientific objectives. Being an experimental work, the envisioned implementation of the system on COTS computing platforms is introduced as well as the experiments planned to gather real imagery to validate the algorithms. Finally, possible fields of application of the project are discussed
A Fuzzy Guidance System for Rendezvous and Pursuit of Moving Targets
Full text linkThis article presents the development of a fuzzy guidance system (FGS) for unmanned aerial
vehicles capable of pursuing and performing rendezvous with static and mobile targets. The system is
designed to allow the vehicle to approach a maneuvering target from a desired direction of arrival and
to terminate the rendezvous at a constant distance from the target. In order to perform a rendezvous
with a maneuvering target, the desired direction of arrival is adjusted over time to always approach
the target from behind, so that the aircraft and target velocity vectors become aligned. The proposed
guidance system assumes the presence of an autopilot and uses a set of Takagi–Sugeno fuzzy controllers
to generate the orientation and speed references for the velocity and heading control loops, given the
relative position and velocity between the aircraft and the target. The FGS treats the target as a mobile
waypoint in a 4-D space (position in 2-dimensions, desired crossing heading and speed) and guides
the aircraft on suitable trajectories towards the target. Only when the vehicle is close enough to the
rendezvous point, the guidance law is complemented with an additional linear controller to manage
the terminal formation keeping phase. The capabilities of the proposed rendezvous-FGS are verified in
simulation on both maneuvering and non-maneuvering targets. Finally, experimental results using a
multi-rotor aerial system are presented for both fixed and accelerating targets
Human- and Machine Design: Resonant-Size Antennas
Full text linkRandom-based global optimization algorithms have been widely used for antenna shape design, primarily in situations where a human-knowledge based solution is not available. In this contribution we study the behavior of random-based global optimization in situations where the design can be addressed with a standard human-based design approach and human-driven parameter tweaking via simulations. The present case study is a resonant patch-type antenna with probe feeding
Real time Optimal Allocation for I-AUV with Interacting Thrusters
No full textEnergy saving is a relevant issue for battery powered Intervention Autonomous Underwater Vehicles which are designed for both short and long-range mission. The energy consumption of an I-AUV is affected by several effects like hydrodynamics, onboard electronics and thrusters cross-couplings. I-AUVs are usually over-actuated system, thus the actuator's interaction takes relevance and should take part within the energy saving process. In this paper the authors present a study on optimal control allocation that aims at considering the interactions between the propellers of an over actuated vehicle where usually two or more thrusters can interfere each other resulting in reduction of the allocation efficiency. The paper presents the mathematical formulation for interacting propeller considering the wake effect and proposes to dynamically adjust the control allocation matrix in order to obtain a cost effective control allocation without modifying the control layer. The existence of a minimum in the energy consumption during the cruising task in function of the parametric control allocation matrix is proved numerically. Thus a perturbation-based extremum seeking approach is used in order to dynamically adapt the parametric allocation matrix and seek the optimal allocation setpoint without explicit knowledge of the real coupling
An Energy-Aware Decision-Making Scheme for Mobile Robots on a Graph Map Based on Deep Reinforcement Learning
No full textAutonomous decision-making has always been one of the primary goals to pursue as concerns mobile robots. Researchers of this field have recently turned their attention to Deep Reinforcement Learning (DRL). This paper presents a Double Deep Q Network architecture for managing the high level decisions of a mobile robot involved in a site servicing task. We imagined a scenario where an autonomous service robot must react to alarms due to failures in its area of interest; the robot must have onboard the necessary servicing tool by resorting to a tool change station if needed, reach the area of the failure and fix it, while at the same time handling its battery status. One of the key properties, yet rarely examined, when it comes to robots' long-term independence is the energy-awareness, namely the ability of autonomously managing the charge state as a function of current and future needs. The proposed Deep Q Network training reward scheme is defined specifically to obtain an energy-aware high-level controller, by penalizing both extremely low levels of battery charge as well as unnecessary recharges. The model is numerically simulated on a graph scenario constituted of several failure and charging nodes. Results show that the trained agent always succeeds in reaching the destination without ever incurring in a complete discharge, as it promptly performs temporary stops at charging locations whenever needed
A diagnosis of progressive myoclonic ataxia guided by blood biomarkers:Expert commentary
Full text linkDecentralized control of a swarm of unmanned aerial vehicles
No full textIn this paper, we present a framework for non-linear control of a swarm of agents based on the artificial potential characterized by attractive and repulsive properties. In this context, the swarm is able to reach a configuration and to maintain it, while migrating as a group and avoiding collisions among agents. Therefore, the behaviors of the swarm system proposed in this study are group migration and configuration, including collision avoidance capabilities. Different potentials expressions are evaluated and one proposed, in order to determine how quickly the swarm converges to a desired direction and velocity, and how robust the swarm is against collisions among the agents. Furthermore we provide two metrics that estimate which potential is the best one in a certain scenario. One quantifies how quickly the swarm converges to the given velocity, and the second evaluates how robust the potential is against collisions. Numerical simulations are included for verification purposes
Low Cost and Low Weight Technologies for planetary Surface Exploration
No full textThis paper presents the results of recent efforts towards the design of low cost and low weight technologies for autonomous vehicles. The test bench for such technologies is an Autonomous Ground Vehicle (UGV) designed and built at University of Pisa. The main issues covered are related to guidance and navigation, by using an integrated system based on odometric and inertial sensors. The problem of obstacle recognition and avoidance is addressed by online vision-based algorithms, and compared with current methods (such as SLAM), presently used for this type of application. Cooperative localization and multi-vehicle coordination are also addressed using dynamic path allocation methods applied in the past to terrestrial unmanned air vehicles coordination and cooperative control. The proposed methodologies are validated through computer simulation, as well as via limited field experiments
- …
