ikkurthis1998 / Genetic-Algorithm-Python

This python code is developed by Sreemannarayana Ikkurthi, as a part of course notes for the course 15AES477: Multidisciplinary Design Optimization (MDO). In support of Dr. Rajesh Senthil Kumar T., Assistant Professor, Department of Aerospace Engneering, Amrita Vishwa Vidyapeetham.

Genetic-Algorithm-Python

This python code is developed by Sreemannarayana Ikkurthi, as a part of course notes for the course 15AES477: Multidisciplinary Design Optimization (MDO), in the year 2019-20

In support of Dr. Rajesh Senthil Kumar T., Assistant Professor,

Department of Aerospace Engneering, Amrita Vishwa Vidyapeetham.

About Genetic Algorithm

Genetic Algorithm is an optimization method, mimicing natural selection process.

A detailed explanation about the method can be found in the text book:

Deb Kalyanmoy, Optimization for Engineering Design, Algorithms and Examples. (2012)

About Code

This code can be used to optimize an objective function of n variables and produce a contour plots of adjacent variables of all generations.

The code contains 8 important functions and 16 helping functions.

The 16 helping functions are made to help the 5 important functions as dividing tasks into smaller functions improves the efficiency of the program.

The 8 important functions and thier usage is as follows:

pop = ini_pop(no_bit, no_var, npop):

    Inputs:
          no_bit = No. of bits per variable
          no_var = No. of variables
          npop = Population size
    Outputs:
          pop = Binary population
    Usage: 
          This function is used to produce a set of binary population with given requirements.

nor_pop = norm_pop(co_pop, lim):

    Inputs:
          co_pop = Binary population
          lim = Limits of variables of form [a, b]; a: lower limit, b: upper limit
    Outputs:
          nor_pop = Normalized population
    Usage: 
          This function is used to convert binary population into normalized population, normalized to given limits.

mat_pool = rep_pop(co_pop, lim):

    Inputs:
          co_pop = Binary population
          lim = Limits of variables of form [a, b]; a: lower limit, b: upper limit
    Outputs:
          mat_pool = Mating pool
    Usage: 
          This function is used to get mating pool from the given population using roulette wheel selection.

int_pool = crov_pop(mat_pool, cp):

    Inputs:
          mat_pool = Mating pool
          cp = Cross-over probability
    Outputs:
          int_pool = Intermediate pool
    Usage: 
          This function is used to get intermediate pool from mating pool.

mut_pool = mut_pop(int_pop, mp):

    Inputs:
          int_pop = Intermediate pool
          mp = Mutation probability
    Outputs:
          mut_pool = Mutated pool
    Usage: 
          This function is used to get mutated pool from intermediate pool.

cont_plot(x, y, a, b):

    Inputs:
          x = x coordinate points
          y = y coordinate points
          a = index of x variable
          b = index of y variable
    Usage: 
          For plotting contour of x and y.

f = fun(x):

    Inputs:
          x = Variables
    Outputs:
          f = Function value
    Usage: 
          To get the value of function.

min_point, fun_val, gen = Genetic_A(fun, no_var, var_lim, no_bit, npop, n_gen):

    Inputs:
          fun = Objective function
          no_bit = No. of bits per variable
          no_var = No. of variables
          npop = Population size
          var_lim = Limits of variables
          n_gen = No. of generations
    Outputs:
          min_point = Minima point
          fun_val = Minimum function value
          gen = Variables of all generations
    Usage: 
          To optimize objective function.

Give appropriate objective function and input variables in INPUT ARENA to get the minima and plot contours.