genetic_algorithm_based_pid_controller

Genetic algorithm is a optimization method involving selection, crossover and mutation steps to evolve the individuals to adapt in given environment.

Project Description

genetic algorithm-based PID controller is adapted to optimize the response of a simple mass-spring-damper system.It is based upon the principle of natural evolution aiming at minimizing the integral of time-weighted absolute error(ITAE) as the fitness function to improve the system response. Fitness function needs to be reduces over a time to get optimal result.

Requirements:

• matlab or octave

Installation

Run this command on command prompt to clone the repository

git clone https://github.com/love481/genetic_algorithm_based_pid_controller.git

Running code

To simulate the response, run on command line

geneticMain.m

Code Structure

• geneticMain.m --> initialize the total population, gene size, ellite_size and iteratation involving evolution process
• GeneticAlgorithm.m --> Class implementation of algorithm contains fitness function, binary to float conversion and pid controller
• getBestChromo.m --> select chromosome having smallest fitness value
• selctTournament.m --> Involves game tournament between two or more chromosomes to find the fittest individuals
• crossover.m --> Crossing between two chromosomes.
• mutation.m --> mutation process seen on the gene of the chromosomes to ensure diversity
• singleIteration --> single iteration of evolution(selection, crossover, mutation)

Results

step response

fitness function

Contact Information

Please feel free to contact me if any help needed

Email: 075bei016.love@pcampus.edu.np