LoboVault Home
 

Distributed, adaptive deployment for nonholonomic mobile sensor networks : theory and experiments

LoboVault

Please use this identifier to cite or link to this item: http://hdl.handle.net/1928/10294

Distributed, adaptive deployment for nonholonomic mobile sensor networks : theory and experiments

Show full item record

Title: Distributed, adaptive deployment for nonholonomic mobile sensor networks : theory and experiments
Author: Luna-Castaneda, Jose Marcio
Advisor(s): Fierro, Rafael
Committee Member(s): Abdallah, Chaouki
Wood, John
Department: University of New Mexico. Dept. of Electrical and Computer Engineering
Subject(s): nonholonomic
consensus problems
mobile robots
coverage control
multivehicle control
sensor networks
LC Subject(s): Wireless sensor networks.
Self-tuning controllers.
Lyapunov stability.
Nonholonomic dynamical systems.
Degree Level: Masters
Abstract: In this work we show the Lyapunov stability and convergence of an adaptive and decentralized coverage control for a team of mobile sensors. This new approach assumes nonholonomic sensors rather than the usual holonomic sensors found in the literature. The kinematics of the unicycle model and a nonlinear control law in polar coordinates are used in order to prove the stability of the controller applied over a team of mobile sensors. This controller is adaptive, which means that the mobile sensors are able to estimate and map a density function in the sampling space without a previous knowledge of the environment. The controller is decentralized, which means that each mobile sensor has its own estimate and computes its own control input based on local information. In order to guarantee the estimate convergence, the mobile sensors implement a consensus protocol in continuous time assuming a fixed network topology and zero communication delays. The convergence and feasibility of the coverage control algorithm are verified through simulations in Matlab and Stage. The Matlab simulations consider only the kinematics of the mobile sensors and the Stage simulations consider the dynamics and the kinematics of the sensors. The Matlab simulations show successful results since the sensor network carries out the coverage task and distributes itself over the estimated density function. The adaptive law which is defined by a differential equation must be approximated by a difference equation to be implementable in Stage. The Stage simulations show positive results, however, the system is not able to achieve an accurate estimation of the density function. In spite of that, the sensors carry out the coverage task distributing themselves over the sampling space. Furthermore, some experiments are carried out using a team of four Pioneer 3-AT robots sensing a piecewise constant light distribution function. The experimental results are satisfactory since the robots carry out the coverage task. However, the accuracy of the estimation is affected by the approximation of the adaptation law by difference equations, the number of robots and sensor sensitivity. Based on the results of this research, the decentralized adaptive coverage control for nonholonomic vehicles has been analyzed from a theoretical approach and validated through simulation and experimentation with positive results. As a future work we will investigate: (i) new techniques to improve the implementation of the adaptive law in real time,(ii) the consideration of the dynamics of the mobile sensors, and (iii) the stability and convergence of the adaptive law for continuous-time variant density function.
Graduation Date: December 2009
URI: http://hdl.handle.net/1928/10294

Files in this item

Files Size Format View
J_M_Luna_Thesis1.pdf 1.103Mb PDF View/Open

This item appears in the following Collection(s)

Show full item record

UNM Libraries

Search LoboVault


Advanced Search

Browse

My Account