Aug-25-2020, 10:11 AM
I am very new to Python and I am trying to understand the implementation of genetic algorithms (GA) for transport applications. I have done some research and found various GA codes and packages, although am hoping that someone might be able to point me in a better direction.
My goal is to be able to implement transport data from a .csv file to create a population of road networks with different origin/destination matrices. These can then be tested against an optimisation fitness function to produce an evolved population, and extract the output into a text file. The fitness function will be input into the GA (for example, determining network structures with the best travel times) to select the networks used in the population.
Input data - origin/destination matrix with 1's indicating a link between an origin and destination node, and 0's indicating no link.
Output data - adjusted origin/destination matrix (with an evolved allocation of 1's and 0's).
Most of what I have tried are from examples online, although I have not had any luck in working out how to apply them to the specific problem - that is, inputting an OD matrix (such as a 10 x 10) with 1's and 0's and evolving it.
I understand there are numerous approaches and examples of codes online, and I have tried to utilise and understand these, although what I have found appears to focus on either using random values or finding a specific output (such as pygad, ga and NetworkX). I find that these are quite complex and often incorporate randomised data throughout, whereas I would like to utilise a specific population of matrices. I understand that it is ideal for me to post my own code and work off that, although I have not yet been able to really achieve any successful deviation from these examples.
One example of code (it is quite long) used can be found in the following link: https://github.com/gmichaelson/GA_in_pyt...mple.ipynb
I would greatly appreciate it if someone is able to suggest an approach that is relatively simple but will allow me to incorporate this heuristic method.
Thanks in advance!!
My goal is to be able to implement transport data from a .csv file to create a population of road networks with different origin/destination matrices. These can then be tested against an optimisation fitness function to produce an evolved population, and extract the output into a text file. The fitness function will be input into the GA (for example, determining network structures with the best travel times) to select the networks used in the population.
Input data - origin/destination matrix with 1's indicating a link between an origin and destination node, and 0's indicating no link.
Output data - adjusted origin/destination matrix (with an evolved allocation of 1's and 0's).
Most of what I have tried are from examples online, although I have not had any luck in working out how to apply them to the specific problem - that is, inputting an OD matrix (such as a 10 x 10) with 1's and 0's and evolving it.
I understand there are numerous approaches and examples of codes online, and I have tried to utilise and understand these, although what I have found appears to focus on either using random values or finding a specific output (such as pygad, ga and NetworkX). I find that these are quite complex and often incorporate randomised data throughout, whereas I would like to utilise a specific population of matrices. I understand that it is ideal for me to post my own code and work off that, although I have not yet been able to really achieve any successful deviation from these examples.
One example of code (it is quite long) used can be found in the following link: https://github.com/gmichaelson/GA_in_pyt...mple.ipynb
I would greatly appreciate it if someone is able to suggest an approach that is relatively simple but will allow me to incorporate this heuristic method.
Thanks in advance!!
