I haven't done anything new on the robot for a few months but I hope to resume the project anytime soon and update this thread.
Here is a video of how it's performing at the moment:
The angle is controlled by a common PD controller and the speed is set in correlation with the angle error. I haven't yet tried any other controller structures, but have a few ideas laying on the stack so to say.
Many parts are overdimensioned and contributes to unnecessary rotational inertia and the sensor array that senses the line should have a lot higher resolution in order to be highly performing i believe.
Next step is to build a PC interface with a GUI to make troubleshooting and optimization easier. Then I would try to get the best possible perfomance out of the current hardware before building a next concept.
Your robot's steering response is very smooth. It travels about the same speed as the line follower I once had, from a store-bought kit called 'Movit'.
The control circuit was surprisingly simple. Two photosensors and a comparator, alternately switching on two drive motors, very rapidly.
hi
If you want to get a higher speed, I suggest you lengthening the appropriate distance between the sensor and the car,Using more sensors that can help you better access path.
Thanks