Doubts in genetic algorithm (Using toolbox)

Hello everyone i have some doubts regarding mixed integer optimization using genetic algorithm in toolbox, please help:

1. what does it mean by population type double?

2. For integer constraint why population should be of type double vector only not bit string or any other ?

3. How population type double vector is different from population type bit string?

4. How a particular string of variables is being represented in a population?

Most important of all :

5. Optimization toolbox uses real coded genetic algorithm or binary coded genetic algorithm??

Thanking you in advance.

Answer 1

[Alan Weiss]

1. Use normal default MATLAB variables, not any other type.
2. The mixed-integer solver, as its name suggests, is MIXED-integer, meaning some variables can be continuous. To keep it simple, one data type is used for all variables.
3. bitstring variables have many limitations, documented here, such as no constraints are allowed.
4. Variables are standard MATLAB variables. The solution is a vector.
5. For mixed-integer optimization, variables are standard doubles. If you want a binary variable, give integer type. a lower bound of 0, and an upper bound of 1. See the documentation, which I tried to make clear.

Alan Weiss

MATLAB mathematical toolbox documentation

[Parikshit Sharma]

Hey alan thanks for your prompt reply ,

You understood my last (5th) question wrongly, i am not asking about binary variable, i wanted to ask that for vector type double, variables are being represented by binary string or real values are used directly for crossover and other operations.

[Walter Roberson]

Resl values are used directly.