APPLICATION OF ARTIFICIAL INTELLIGENCE AND REINFORCEMENT MACHINE LEARNING ON ROBOTIC ARM









Abstract

This research work is based on deep Reinforcement Learning on a robotic arm by controlling it and performing task training on it. Touching the practicalities of the aspect of creating a robot in a virtual environment and training a TensorFlow NN model using python along with hyper-parameter optimization. All in All, in this research work we would be looking at the practical aspects of making an artificially intelligent custom robot with any kinematic parameters, like the length of linkages, types, and the number of joints in a virtual environment and training it in the same virtual environment, the trained model then being used in the robotic arm directly with the help of transfer learning. In the following research work, we would also revisit the basics of dynamic analysis of open chain linkages, second-order response systems, robot components, types of grippers, sensors, and some aspects of programming in the Unity game engine. The part dealing with machine vision, sensors and applications will not be discussed for the sake of brevity. Keywords: Robot, ML Agents, ROS, Unity, Neural Networks, Artificial Intelligence.


Modules


Algorithms


Software And Hardware

• Hardware: Processor: i3 ,i5 RAM: 4GB Hard disk: 16 GB • Software: operating System : Windws2000/XP/7/8/10 Anaconda,jupyter,spyder,flask Frontend :-python Backend:- MYSQL