Neural network with genetic algorithm
These self-learning cars learn to drive with neural networks and genetic algorithms.
Generation #1
Lap #1
Show only one
All cars can detect the distance to the wall in three directions. Every drawn direction in the image is an input to the network.
All the directions where the car can see.
The angels are written in radians.
There is the current neural network of the blue car in the image. Black dots are neurons. Bigger neurons have a bigger bias value. Blue lines are negative connections and red lines positive.
There is the current neural network of the blue car in the image. Black dots are neurons. Bigger neurons have a bigger bias value. Blue lines are negative connections and red lines positive.