Source author record

Andrea Garulli

Andrea Garulli appears in the imported research catalog. Authorship, coauthor and topic links are available while profile ownership is still unclaimed.

ResearcherUnclaimed source record

Systems and Control eess.SY math.OC physics.soc-ph Robotics Social and Information Networks

Catalog footprint

What is connected

4works

6topics

4close collaborators

Actions

Connect this record

Open graph Browse works

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.

preprint2020arXiv

Cooperative versus decentralized strategies in three-pursuer single-evader games

The value of cooperation in pursuit-evasion games is investigated. The considered setting is that of three pursuers chasing one evader in a planar environment. The optimal evader trajectory for a well-known decentralized pursuer strategy is characterized. This result is instrumental to derive upper and lower bounds to the game length, in the case in which the pursuers cooperate in the chasing strategy. It is shown that the cooperation cannot reduce the capture time by more than one half with respect to the decentralized case, and that such bound is tight.

preprint2020arXiv

Satellite Relative Motion Modeling and Estimation via Nodal Elements

In this paper, a new parametrization of the relative motion between two satellites orbiting a central body is presented. The parametrization is based on the nodal elements: a set of angles describing the orbit geometry with respect to the relative line of nodes. These are combined with classical orbital elements to yield a nonsingular relative motion description. The exact nonlinear, perturbed dynamic model resulting from the new parametrization is established. The proposed parameter set captures the fundamental Keplerian invariants, while retaining a simple relationship with local orbital coordinates. An angles-only relative navigation filter and a collision avoidance scheme are devised by exploiting these features. The navigation solution is validated on a case study of an asteroid flyby mission. It is shown that a collision can be detected early on in the estimation process, which allows one to issue a timely evasive maneuver.

preprint2019arXiv

An Adaptive Groundtrack Maintenance Scheme for Spacecraft with Electric Propulsion

In this paper, the repeat-groundtrack orbit maintenance problem is addressed for spacecraft driven by electric propulsion. An adaptive solution is proposed, which combines an hysteresis controller and a recursive least squares filter. The controller provides a pulse-width modulated command to the thruster, in compliance with the peculiarities of the electric propulsion technology. The filter takes care of estimating a set of environmental disturbance parameters, from inertial position and velocity measurements. The resulting control scheme is able to compensate for the groundtrack drift due to atmospheric drag, in a fully autonomous manner. A numerical study of a low Earth orbit mission confirms the effectiveness of the proposed method.

preprint2016arXiv

Asymptotic analysis of threshold models for social networks

A class of dynamic threshold models is proposed, for describing the upset of collective actions in social networks. The agents of the network have to decide whether to undertake a certain action or not. They make their decision by comparing the activity level of their neighbors with a time-varying threshold, evolving according to a time-invariant opinion dynamic model. Key features of the model are a parameter representing the degree of self-confidence of the agents, and the mechanism adopted by the agents to evaluate the activity level of their neighbors. The case in which a radical agent, initially eager to undertake the action, interacts with a group of ordinary agents, is considered. The main contribution of the paper is the complete analytic characterization of the asymptotic behaviors of the network, for three different graph topologies. The asymptotic activity patterns are determined as a function of the self-confidence parameter and of the initial threshold of the ordinary agents.