Skip to content

A simple but realistic drone simulator, easily implementable in RNN's, offering 6 different parameters to tune while in the simulation and keyboard controls for you to test.

License

Notifications You must be signed in to change notification settings

Kaito0733/2D-Drone-Simulator-for-Reinforcement-Learning

Repository files navigation

2D-Drone-Simulator-for-Reinforcement-Learning

A simple but realistic drone simulator made in pygame, easily implementable in RNN's, offering 6 different parameters to tune while in the simulation and keyboard controls. You can enable radomly spawning target, to train the drone to fly to specific coordinates.

Use main_manual_control_inputs.py to play the simulation with left and right keys by yourself. Use example_with_random_inputs.py to see how the drone flies with randomly generated arrays of 0's and 1's, without manual steering from the keyboard. Use RNN.py to implement your RNN or another type of AI-model to let it control the drone with generated control inputs. Please read the comments in the scripts to understand how to implement AI-generated sequences and how to use shortcuts.

class_for_RNN.py and fancy_drone.png are resources, if you want, you can modify them.

About

A simple but realistic drone simulator, easily implementable in RNN's, offering 6 different parameters to tune while in the simulation and keyboard controls for you to test.

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages