An important field of study that has been addressed over the last few years is autonomous mobile robotics. One of the main tasks that an autonomous mobile robot is required to perform is exploration, that amounts to incrementally discover an initially unknown environment in order to build its map. During an exploration, an autonomous robot perceives a portion of the environment with its sensors, integrates its sensor readings within its current map, identifies the boundaries between known and unknown portions of the environment, evaluates such boundaries according to an utility function in order to select one of them according to an exploration strategy and, finally, reaches it. In this thesis, we propose a multi-criteria exploration strategy that uses the current map that has been built during the exploration (which represents the explored part of the environment) to predict the layout of the unknown part of the environment and that exploits this knowledge to enhance the exploration process. We evaluate the performance of our system in di erent simulated indoor environments, obtaining a consistent speed up at the end of the exploration process.
La robotica mobile autonoma è un ambito di ricerca che ha ricevuto molta attenzione negli ultimi anni. Uno dei compiti principali che un robot mobile autonomo deve essere capace di compiere è l'esplorazione, definito come la scoperta incrementale di un ambiente inizialmente sconosciuto. Un robot mobile autonomo durante l'esplorazione usa i sensori per percepire l'ambiente, integra i dati raccolti in una mappa, identifica le frontiere tra le porzioni di mappa esplorata e le porzioni di mappa inesplorata, valuta le frontiere utilizzando una funzione di utilità e ne seleziona una utilizzando una strategia di esplorazione, e infine la raggiunge. In questa tesi proponiamo una strategia di esplorazione multicriterio che usa la mappa costruita durante l'esplorazione (che rappresenta la parte esplorata dell'ambiente) per predirre il layout dalla parte sconosciuta dell'ambiente e che sfrutta questa conoscenza per migliorare il processo esplorativo. Valutiamo le prestazioni del nostro sistema in molteplici ambienti interni simulati, mostrando che migliora il processo esplorativo.
Use of predicted layouts of indoor environments to improve exploration strategies for autonomous mobile robots
FOCHETTA, LUCA
2017/2018
Abstract
An important field of study that has been addressed over the last few years is autonomous mobile robotics. One of the main tasks that an autonomous mobile robot is required to perform is exploration, that amounts to incrementally discover an initially unknown environment in order to build its map. During an exploration, an autonomous robot perceives a portion of the environment with its sensors, integrates its sensor readings within its current map, identifies the boundaries between known and unknown portions of the environment, evaluates such boundaries according to an utility function in order to select one of them according to an exploration strategy and, finally, reaches it. In this thesis, we propose a multi-criteria exploration strategy that uses the current map that has been built during the exploration (which represents the explored part of the environment) to predict the layout of the unknown part of the environment and that exploits this knowledge to enhance the exploration process. We evaluate the performance of our system in di erent simulated indoor environments, obtaining a consistent speed up at the end of the exploration process.File | Dimensione | Formato | |
---|---|---|---|
2019_04_Fochetta.pdf
accessibile in internet per tutti
Descrizione: Testo della tesi
Dimensione
6.07 MB
Formato
Adobe PDF
|
6.07 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/147956