Genetic-Algorithm-N-Queens

Solving N-queens problem with Genetic Algorithm

Overview

N-queens puzzle:

The N-queens puzzle is the problem of placing N chess queens on an N*N chessboard so that no two queens threaten each other; thus, a solution requires that no two queens share the same row, column, or diagonal.

Genetic algorithm:

Genetic algorithms are commonly used to generate high-quality solutions to optimization and search problems by relying on biologically inspired operators such as mutation, crossover, and selection. (for more information, you can use this link).

Goal

In this project, I have used the Genetic algorithm to solve the N-queens puzzle. I have implemented two selection methods, two different types of crossover and, two mutation methods. In the table below, you can see how to run each model.

How to Run

1. gcc -Wall -o output main.c init.c random.c eval.c stats.c gen.c select.c xover.c rep.c
2. ./output [enter number 1 or 2 or 3]

Mode Number	Selection	Crossover	Mutation
./output 1	Tournament Selection	Two-Point	Uniform
./output 2	Roulette Wheel	One-Point	Uniform
./output 3	Tournament Selection	Two-Point	Swap Mutation

THE BEST ANSWER 🎉

+ THE BEST ANSWER is for mode number 1, It can solve the 88-Queens Problem

88-Queens is a version of the N-Queens puzzle on an 88x88 chessboard with 88 queens. The object is to calculate positions for all 88 queens such that no queen is attacking any other queen.

---

Mode number 1️⃣:

Population Size	MaxGen	pCross	Start pMut
1000	1000	0.8	0.001

If maxFitness > 0.97 then pMut = 0.01;

Output:

N-Queens Problem	Result
n = 8	Global Optimal in 14th generation
n = 9	Global Optimal in 36th generation
n = 11	Global Optimal in 104th generation
n = 20	Global Optimal in 71th generation
n = 25	Local Optimal with 0.9966 fitness
n = 30	Global Optimal in 942th generation
n = 40	Local Optimal with 0.9974 fitness
n = 45	Global Optimal in 159th generation
n = 48	Global Optimal in 240th generation
n = 50	Local Optimal with 0.9991 fitness
n = 75	Local Optimal with 0.9996 fitness
n = 80	Global Optimal in 870th generation
n = 88 🎉	Global Optimal in 745th generation
n = 100	Local Optimal with 0.9993 fitness

---

Mode number 2️⃣:

Population Size	MaxGen	pCross	Start pMut
1000	1000	0.8	0.001

If maxFitness > 0.97 then pMut = 0.01;

Output:

N-Queens Problem	Result
n = 8	Global Optimal in 67th generation
n = 9	Global Optimal in 46th generation
n = 10	Global Optimal in 173th generation
n = 11	Global Optimal in 467th generation
n = 12	Global Optimal in 252th generation
n = 13	Local Optimal with 0.96 fitness
n = 14	Global Optimal in 739th generation
n = 15	Local Optimal with 0.97 fitness

---

Mode number 3️⃣:

Population Size	MaxGen	pCross	Start pMut
1000	1000	0.8	0.001

If maxFitness > 0.97 then pMut = 0.01;

Output:

N-Queens Problem	Result
n = 8	Global Optimal in 9th generation
n = 9	Global Optimal in 35th generation
n = 10	Global Optimal in 45th generation
n = 25	Local Optimal with 0.9966 fitness
n = 50	Local Optimal with 0.9934 fitness

---

Support

Reach out to me at riasiarman@yahoo.com