search

reiniscimurs / DRL-robot-navigation

Deep Reinforcement Learning for mobile robot navigation in ROS Gazebo simulator. Using Twin Delayed Deep Deterministic Policy Gradient (TD3) neural network, a robot learns to navigate to a random goal point in a simulated environment while avoiding obstacles.
MIT License
486 stars 97 forks source link

Changing the lidar angel from 180 to 360 degrees will make the model to perform better or not? #131

Closed am-amani closed 2 months ago

am-amani commented 2 months ago

Dear @reiniscimurs ,

One question. Have you tried or heard someone try to successfully train the mobile robot with lidar at a 360-degree angle instead of 180 degrees? In your code, you are using 180 degrees. I was wondering if changing it to 360 degrees will increase the trained agent's performance. I understand it will take longer to train the model, but I wondered if changing the environment dimension from 20 to 40 will help the robot perform better.

Thank you for your attention.

reiniscimurs commented 2 months ago

Hi,

You can go through some still open and closed issues. Some people have tried this approach with varying results. There are papers out there that use 360 degree FOV but it is hard for me to say if that improves the performance.

am-amani commented 2 months ago

Thank you for your reply!