/Fuzzy-Algorithm-HW

Primary LanguagePython

Fuzzy Algorithm apply on self-driving car

Introduction

This program is the homework of Computational Intelligence in NCU. This program applies the fuzzy algorithm on a simple self-driving car system.

The program is written in python3.6. The require package is shown below.

import sys, pygame
import os
import time
import math
  • The map is set in map.txt
  • Fuzzy type is sellected in Method.txt. There are three type of fuzzy I wrote, which is
    1. Functional base fuzzy
    2. Center Gravity fuzzy
    3. Mean Max fuzzy

To edit the content in Methdo.txt as Fuzzy, FuzzyGravity or FuzzyMean can change the program fuzzy type.

The main program is autoCar.py. Run the game2exe.py can convert the python script to the executable program.

An output Log and the screenshot will save at working direction after the program is finished.

Fuzzy design

  • The input distance function Front distance Left and Right distance

  • The output degree function

Result

For the comparison of the three methods I wrote.

The Functional base fuzzy is as below. It shows that froms the first corner, the car had an extra turn before it back to the right path

And in Center Gravity or Mean Max method, both of them are very stable tour to the end. But the Center Gravity is more stable than Mean Max due to less shaking when the car moved. (Center Gravity) (Mean Max)