In recent times, the control of multiple mobile robots is becoming increasingly successful. Multi-robot structures can perform tasks that cannot be done by a single robot. Therefore, the study of the simultaneous motion of multiple robots is becoming more and more common. Different kinds of robots that can be used for this purpose; in this thesis, two of them are studied, a classical non-holonomous type and another omnidirectional or holonomous one, that is capable of moving freely without necessarily change orientation. The project is inspired by an existing system capable of moving omni- directionally and supporting high loads. The robot is developed by the company KUKA and is called OmniMove. It is capable of transporting objects such as train wagons, aircraft parts, etc. The study attempts to recreate a similar structure consisting of two holonomous robots connected with a platform. From the literature, the type of control used in structures already developed in academic institutions were studied. Once the type of control to be implemented in our structure had been decided, we moved on to write a programme that could be implemented in ROS (Robot Operating System), a software capable of interact to real or virtual robots via our commands. The programme was written in Python and involves acquiring and sending data to control and monitor the robots. These will have to move in a synchronised manner, for whatever path is given to them. The programme was then implemented with a closed-loop feedforeward controller that allows us to correct the robot’s trajectory according to errors that can be measured during movement. Another method for measuring errors was studied and implemented, observing its advantages and drawbacks. In addition to programming the movement of the two robots, a first prototype for connecting the robots was designed and realised. A platform that rests on three load-bearing spheres, one in the centre of each robot and one in the centre of the platform that supports it from the ground. These spheres allow the robots to deviate slightly from the trajectory described by the synchronism. If the platform’s anchoring points were rigid, the robots’ components could be ruined after only few applications. In order to support the load-bearing spheres and not move the platform under normal working conditions, soft material supports were created to allow the robots to move without repercussions and to support the platform. Several simulations were carried out both in the design area, to create and realise the soft components and to design the platform, and to control the navigation of the robots. From the prototype, it was possible to observe the actual behaviour of the platform and deduce conclusions and future improvements applicable to the platform itself as well as to the entire system.
Negli ultimi tempi il controllo di molteplici robot mobili sta riscontrando sempre più successo. Con strutture multi-robot si possano compiere compiti che non possono essere eseguiti da un singolo robot. Pertanto, lo studio della movimentazione simultanea di molteplici robot è sempre più frequente. Diverse sono le tipologie di robot possono essere usati per a tale scopo. In questa tesi di laurea ne vengono studiati due, uno di tipo classico non olonomo e un altro omnirezionale o olonomo, capace cioè, di muoversi liberamente senza dover cambiare forzatamente orientamento. Il progetto si inspira ad un sistema già esistente capace di muoversi in maniera omnidirezionale e di supportare elevati carichi. Il robot in questione è sviluppato dall’azienda KUKA e si chiama OmniMove. È capace di trasportare oggetti come vagoni di treni, parti di aerei ecc. Lo studio cerca di ricreare una struttura simile composta da due robot olonomi collegati con una piattaforma. Dalla letteratura sono state studiate le strutture già sviluppate in ambito accademico, concentrandosi sulla tipologia di controllo usata. Una volta decisa la tipologia di controllo da implementare nella nostra struttura si è passati alla scrittura di un programma implementabile nel software ROS (Robot Operating System) in grado di far interagire i nostri comandi, scritti nei codici, con dei robot reali o virtuali. Il programma è stato scritto con Python e prevede l’acquisizione e l’invio di dati per controllare e monitorare i robot. Questi si dovranno muovere in maniera sincronizzata, per qualsiasi percorso fornitogli. Si è poi implementato il programma con un controllore a ciclo chiuso in retroazione che ci permette di correggere la traiettoria del robot in base a degli errori misurabili nel corso della movimentazione. Un altro metodo di misurazione degli errori è stato studiato e implementato osservandone i pregi e i difetti. Oltre alla programmazione della movimentazione dei due robot è stato progettato e realizzato un primo prototipo di connessione tra gli stessi. Una piattaforma che appoggia su tre sfere portanti, una al centro di ciascun robot e una al centro della piattaforma che tramite un supporto la sostiene dal terreno. Queste sfere consentono ai robot di potersi discostare leggermente dalla traiettoria descritta dal sincronismo. Se i punti di ancoraggio della piattaforma fossero stati rigidi la componentistica dei robot si sarebbe potuta rovinare dopo poche applicazioni. Per sostenere le sfere portanti e non far muovere la piattaforma nelle condizioni normali di lavoro, sono state create dei supporti di materiale morbido che consentono di far muovere i robot senza ripercussioni e, contemporaneamente, di supportare la piattaforma. Sono state compiute svariate simulazioni sia nell’ambito progettuale, per realizzare i supporti morbidi e per progettare la piattaforma che per controllare la navigazione dei robot. Dal prototipo è stato possibile osservare il comportamento reale della piattaforma, dedurne delle conclusioni e ipotizzare dei futuri miglioramenti, applicabili sia alla piattaforma stessa che all’intero sistema.
Trasporto collaborativo di oggetti con due robot mobili olonomi
FAVA, MARCO
2021/2022
Abstract
In recent times, the control of multiple mobile robots is becoming increasingly successful. Multi-robot structures can perform tasks that cannot be done by a single robot. Therefore, the study of the simultaneous motion of multiple robots is becoming more and more common. Different kinds of robots that can be used for this purpose; in this thesis, two of them are studied, a classical non-holonomous type and another omnidirectional or holonomous one, that is capable of moving freely without necessarily change orientation. The project is inspired by an existing system capable of moving omni- directionally and supporting high loads. The robot is developed by the company KUKA and is called OmniMove. It is capable of transporting objects such as train wagons, aircraft parts, etc. The study attempts to recreate a similar structure consisting of two holonomous robots connected with a platform. From the literature, the type of control used in structures already developed in academic institutions were studied. Once the type of control to be implemented in our structure had been decided, we moved on to write a programme that could be implemented in ROS (Robot Operating System), a software capable of interact to real or virtual robots via our commands. The programme was written in Python and involves acquiring and sending data to control and monitor the robots. These will have to move in a synchronised manner, for whatever path is given to them. The programme was then implemented with a closed-loop feedforeward controller that allows us to correct the robot’s trajectory according to errors that can be measured during movement. Another method for measuring errors was studied and implemented, observing its advantages and drawbacks. In addition to programming the movement of the two robots, a first prototype for connecting the robots was designed and realised. A platform that rests on three load-bearing spheres, one in the centre of each robot and one in the centre of the platform that supports it from the ground. These spheres allow the robots to deviate slightly from the trajectory described by the synchronism. If the platform’s anchoring points were rigid, the robots’ components could be ruined after only few applications. In order to support the load-bearing spheres and not move the platform under normal working conditions, soft material supports were created to allow the robots to move without repercussions and to support the platform. Several simulations were carried out both in the design area, to create and realise the soft components and to design the platform, and to control the navigation of the robots. From the prototype, it was possible to observe the actual behaviour of the platform and deduce conclusions and future improvements applicable to the platform itself as well as to the entire system.File | Dimensione | Formato | |
---|---|---|---|
MASTER_THESIS_Fava_Marco.pdf
accesso aperto
Dimensione
8.45 MB
Formato
Adobe PDF
|
8.45 MB | Adobe PDF | Visualizza/Apri |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.12075/10716