Reinforcement learning algorithm for industrial robot programming by demonstration

Programming by demonstration represent a significant subject in the field of robotics and it is developing more and more in the direction of robots for services and humanoid robots. Programming by demonstration is much less researched, when we talk about industrial robots. One of the reasons is that...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:2012 13th International Conference on Optimization of Electrical and Electronic Equipment s. 1517 - 1524
Hlavní autoři: Stoica, M., Sisak, F., Morosan, A. D.
Médium: Konferenční příspěvek
Jazyk:angličtina
Vydáno: IEEE 01.05.2012
Témata:
Classification algorithms
Learning
Programming
Robot kinematics
Service robots
Software
ISBN:9781467316507, 1467316504
ISSN:1842-0133
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Programming by demonstration represent a significant subject in the field of robotics and it is developing more and more in the direction of robots for services and humanoid robots. Programming by demonstration is much less researched, when we talk about industrial robots. One of the reasons is that an industrial robot has to act in a precise and certain manner. However, extending research regarding programming by demonstration in industrial robots area, could lead to development of intelligent systems, where the industrial robot could be programmed in an easier way. In this paper we proposed an algorithm based on reinforcement learning and we developing, implementing and testing this algorithm which can offer flexibility in intelligent systems. Initially, we have focused our research on the creation of a reasoning algorithm based on artificial neural networks, but the results of this algorithm weren't satisfying, so we have switched our focus towards proposed algorithm. The results of this algorithm is that the robot will be capable to learn from its mistakes and he will know how to act in unknown situation; this will be possible because the robot will get marks for each possible action and he will updates its behavior.
ISBN:9781467316507
1467316504
ISSN:1842-0133
DOI:10.1109/OPTIM.2012.6231926