Razvoj programskega orodja za napovedovanje obnašanja mobilnega robota

magistrsko delo magistrskega študijskega programa II. stopnje Strojništvo

Gregor Bolka (Author), Rok Vrabič (Mentor)

Abstract

Avtonomni mobilni roboti v sodobnem industrijskem okolju so obkroženi s številnimi premikajočimi objekti, ki jih robot lahko spremlja s pomočjo svojih zaznaval. V tej nalogi smo za namen napovedovanja trajektorij preučili metode za analizo časovnih vrst s poudarkom na uporabi umetnih nevronskih mrež. Ugotovili smo, da se enkoder/dekoder LSTM mreža lahko uspešno nauči periodičnih vzorcev gibanja robota. Z nadgradnjo te arhitekture smo uspeli napovedovati tudi kratkoročne trajektorije, kar smo v praksi realizirali v obliki ROS vozlišča za napovedovanje trajektorij.

Keywords

magistrske naloge;mobilna robotika;časovne vrste;napovedovanje trajektorij;umetne nevronske mreže;LSTM mreže;GRU mreže;robotski sistem ROS;

Data

Language:	Slovenian
Year of publishing:	2021
Typology:	2.09 - Master's Thesis
Organization:	UL FS - Faculty of Mechanical Engineering
Publisher:	[G. Bolka]
UDC:	007.52:004.85:004.032.26(043.2)
COBISS:	51334147
Views:	430
Downloads:	100
Average score:	0 (0 votes)
Metadata:

Other data

Secondary language:	English
Secondary title:	Development of a software package for predicting mobile robot behaviour
Secondary abstract:	Autonomous mobile robots in the modern industrial environment are surrounded by numerous moving objects, which the robot is able to track using its sensors. Often the future position of such objects is needed, therefore we examined the usage of time series methods for trajectory prediction with an emphasis on neural network models. We showed that encoder-decoder LSTM model can successfully learn periodic patterns in the movement of a robot. Enhanced version of this architecture was used to predict short-term trajectories, which we implemented in practice as a ROS node for trajectory prediction.
Secondary keywords:	master thesis;mobile robotics;time series;trajectory prediction;artificial neural networks;LSTM networks;GRU networks;robotics middleware ROS;
Type (COBISS):	Master's thesis/paper
Study programme:	0
Embargo end date (OpenAIRE):	1970-01-01
Thesis comment:	Univ. Ljubljana, Fak. za strojništvo
Pages:	XXII, 68 str.
ID:	12515752