motoko reloaded - msp430 neural net

Project

motoko reloaded - msp430 neural network line follower

Press

to mark as completed

Attachments

Introduction

This is an automatic import from our previous community platform. Some things can look imperfect.

If you are the original author, please access your User Control Panel and update it.

line follower using two linear neurons (controllers) PD for differential and I for speed; both with antiwindup, error filtering and simulated annealing tunning (take about 80 loops in short testing line loop); second controll stage is curve predictor - table of curves to prevent overshoot (still present)

motoko reloaded - msp430 neural network line follower
motoko reloaded - msp430 neural network line follower

motoko reloaded - msp430 neural network line follower

line follower

Actuators / output devices: DC Motors, pololu
Control method: modified pid with curve predictor, simulated annealing learning
CPU: msp430f2232
Operating system: none
Power source: 2S lipol 250mAh / 400mAh
Programming language: C, ruby
Sensors / input devices: accelerometer, gyro, 8x cny70, 3x ir
Target environment: indoor

Flag this post

motoko reloaded - msp430 neural network line follower

motoko reloaded - msp430 neural network line follower

Thanks for helping to keep our community civil!