Proširenje MultiAGV paketa za upravljanje autonomnim vozilima funkcionalnošću zaobilaženja nepoznatih prepreka

master's thesis

• Title (english): Implementation of Obstacle Avoidance in MultiAGV Package for Autonomous Vehicle Control
• Author: Vatroslav Čelan
• Mentor: Stjepan Bogdan (mentor)
• Committee member: Stjepan Bogdan (predsjednik povjerenstva)
• Committee member: Zdenko Kovačić (član povjerenstva)
• Committee member: Matko Orsag (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2017-07-18, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Electrical Engineering

Abstract

MultiAGV paket Robotskog operacijskog sustava (ROS) za upravljanje autonomnim vozilima proširen je analizom i klasifikacijom prepreka. Definirane su dvije vrste prepreka: statičke i dinamičke. Klasifikacija se vrši na temelju različtih faktora i kao rezultat daje procjenu tipa prepreke i ocjenu pouzdanosti u klasifikaciju. Dinamičkim preprekama omogućena je analiza gibanja i brzina.

Nadalje, funkcionalnost sustava proširena je na izbjegavanje statičkih prepreka. Uzevši u obzir isplaniranu putanju, dimenzije robota i laserska očitanja vrši se provjera dolazi li do sudara sa preprekom. U slučaju predviđenog sudara traži se planiranje nove putanje gdje se u obzir uzima originalna karta prostora i trenutna laserska očitanja, što rezultira novoisplaniranom putanjom koja zaobilazi statičku prepreku.

Simulacija sustava izvođena je u alatu Gazebo, a eksperimentalna evaluacija izvršena je na robotu Pioneer P3-DX u zgradi Fakulteta Elektrotehnike i Računarstva. Testovima je utvrđena točna klasifikacija prepreka i ispravnost sustava za izbjegavanje statičkih prepreka.

Abstract (english)

Robot operating system (ROS) package MultiAGV for autonomous vehicle control is upgraded with analysis and obstacle classification. There are two obstacle types defined: static and dynamic. Classification is handled based on several factors and provides obstacle type estimation and its reliability level. Position and speed analysis can be applied for dynamic obstacles.

Furthermore, the system is upgraded with static obstacle avoidance functionality. Collision detection is executed by having path plan, robot dimensions and laser sensor readings taken into account. In case the collision is predicted, path replanning is requested in which the original map and current laser readings are used. The result is a new path where the obstacle is avoided.

Simulations are executed in Gazebo, whereas experimental evaluation is carried out on Pioneer P3-DX robot inside the Faculty of Electrical Engineering and Computing builiding. The experiments have confirmed the obstacle classification accuracy and static obstacle avoidance performance.

• Keywords: autonomno vozilo; mobilna robotika; statičke prepreke; dinamičke prepreke; navigacija
• Keywords (english): autonomous vehicle; mobile robotics; obstacle avoidance; static obstacle; dynamic obstacle; navigation
• Language: croatian
• URN:NBN: urn:nbn:hr:168:083903
• Study programme: Title: Electrical Engineering and Information Technology; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka elektrotehnike i informacijske tehnologije (magistar/magistra inženjer/inženjerka elektrotehnike i informacijske tehnologije)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2019-03-13 19:43:14