Asynchronous Distributed Averaging, Cooperative Tilt Estimation, and Modular Robotic Self-Assembly
Open access
Author
Date
2017Type
- Doctoral Thesis
ETH Bibliography
yes
Altmetrics
Abstract
This dissertation provides an advanced platform for research on distributed estimation and control, and investigates three different topics that pertain to these fields of research: (1) a generic asynchronous distributed averaging algorithm, (2) cooperative tilt and altitude estimation from multiple distance sensors, and (3) modular robotic self-assembly through rendezvous with bearing-only information.
This work employs the Distributed Flight Array, which is a flying platform consisting of multiple autonomous single propeller modules that are able to drive on the ground, self-assemble with their peers, and when assembled fly in a coordinated fashion. As part of this work, the Distributed Flight Array has been completely revised enhancing its usability for research by honing its hardware and devising a software ecosystem that incorporates the embedded and distributed nature of the platform, and can be simulated to a large amount. The three different research topics of this work have been implemented and evaluated in experiments on this platform.
The research on distributed averaging considers the context of a distributed embedded system with multiple agents connected through a communication network, where adversities such as switching network topologies, agents joining or leaving the network, and unreliable communication links may arise. To address these adversities, we propose an asynchronous implementation of a distributed averaging algorithm that has the following properties: (1) unbiased average, (2) homogeneous implementation, (3) mitigation of above network adversities, and (4) well-defined parameters.
In addition, this work proposes a generalized algorithm for computing the altitude and tilt of a rigid body with respect to an inertial frame using a set of distance measurements obtained from a sensor network. In the case where all sensors are centrally measurable, a linear-optimal estimate is obtained. To account for communication bandwidth constraints, a scalable, distributed scheme is borrowed from literature, where local information is shared only with immediate neighbors. In the limit of sharing information, each agent asymptotically computes the linear-optimal tilt and altitude estimate.
Last, this work extends an algorithm for rendezvous, which is found in literature, and enables mobile agents to meet in a bounded region based only on bearing information of other agents within their vicinity. Each agent repeatedly employs a stop-and-go strategy consisting of the following three actions: (1) estimate the bearing of agents in its vicinity, (2) compute a target point based on the estimates, and (3) move to that target point. We leverage this algorithm for self-assembly of the Distributed Flight Array such that several modules that are distributed on the ground collide with each other by employing the proposed rendezvous strategy. Upon collision the magnets on their chassis align and assemble the modules into clusters, which coordinate their actions and continue to rendezvous with other modules and clusters until they assemble into one large structure. Show more
Permanent link
https://doi.org/10.3929/ethz-b-000258117Publication status
publishedExternal links
Search print copy at ETH Library
Publisher
ETH ZurichOrganisational unit
03758 - D'Andrea, Raffaello / D'Andrea, Raffaello
Funding
146717 - Distributed Estimation and Control of Mechatronic Systems (SNF)
More
Show all metadata
ETH Bibliography
yes
Altmetrics