University of Twente Student Theses


Heterogeneous cooperative target tracking using Nonlinear Model Predictive Control

Kivits, Max P.W. (2021) Heterogeneous cooperative target tracking using Nonlinear Model Predictive Control.

[img] PDF
Abstract:The autonomy of robots is limited by the accuracy of the information of their environment. The availability of accurate state estimates of the environment is crucial. This work researches how to obtain these state estimates and proposes a method to cooperatively track moving targets using a team of heterogeneous agents. A state estimation scheme fuses the noisy intermittent target measurements and its covariance is used to quantify the estimation quality. To this end, a C++ implementation of an Intermittent Kalman filter in the Genom framework is constructed. A NMPC method from previous work is adapted to include a the Intermittent Kalman filter covariance trace as perception objective. The proposed NMPC is able to compute the actuator inputs for a team of heterogeneous generic multi-rotors in order to minimize their collective state estimation covariance, whilst honoring actuator constraints. This removes the need for a cascaded control scheme. The proposed method is validated using simulation experiments. The NMPC shown to drive two quadrotor sensing agent into a configuration that minimizes their state estimate covariance beyond what it can achieve using a single quadrotor. The method is able to tracking moving targets using both single and dual quadrotors, with a control frequency of over 600Hz and 300Hz respectively.
Item Type:Essay (Master)
LAAS, Toulouse, France
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:53 electrotechnology, 54 computer science
Programme:Electrical Engineering MSc (60353)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page