In this work artificial intelligence learning methods are analyzed for robot control. Most often in practice robots solve problems by using analytic algorithms, however, for these algorithms to work correctly robot has to be manufactured with very high precision, that requires expensive parts, not only that to make these algorithms high amount of high quality specialist assistance is required, but even that not always helps to solve all the problems to evaluate all possible errors. For this reason it can be useful to use artificial intelligence methods to control the robot. This thesis's main focus is a robot that can adapt to its surrounding ever changing environment, for this reason methods that use large amounts of data (pretrained models) do not suit this problem for this reason reinforcement learning methods are used. The environment that allows low effort design, configure a robot, control and train it.


    Zugriff

    Download


    Exportieren, teilen und zitieren



    Titel :

    Dirbtinio intelekto metodų panaudojimas autonominiam besimokančiam robotui ; Autonomous self-learning robot


    Beteiligte:

    Erscheinungsdatum :

    2022-06-07


    Medientyp :

    Hochschulschrift


    Format :

    Elektronische Ressource


    Sprache :

    Lithuanian , Englisch


    Klassifikation :

    DDC:    629