Abstract
The control algorithm of the agent is an important research area of artificial intelligence, which reflects the way the agent completes tasks in a dynamic environment. Reinforcement learning is an advanced agent learning algorithm, which makes decision learning through the interaction between the agent and the dynamic environment, improves the strategy after repeated experiments, and finally gets the optimal action plan. In this paper, the PUMA560 robotic arm is selected as the agent body for simulation experiment research. In order to design a high-precision, low-error, fast-control robotic arm control method, we use Webots simulation software as a platform, use image-based visual servoing control algorithms, and use reinforcement learning algorithm Q-learning as the basis to improve and optimize control method.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call