Genetic and Evolutionary Algorithms for PSP - Digital Signal Processing in Power System Protection and Control - page 277

Parent Offspring

0 000 110 0 1110 0000 110 1 1110

Fig. 13.6 Mutation example

(a) (b)

parent
possible
offspring

parent

possible
offspring

variable 1

variable 1

Fig. 13.7 Recombination examples: a 2-D grid locations, b locations along a line

• Represent the problem domain as a chromosome,
• Define a fitness (goal) function to evaluate the chromosome performance,
• Construct the genetic operators (way of implementation for given encoding),
• Define probabilities for genetic operations (constant or changing),
• Generate initial population of individuals,
• Run the GA and tune its parameters.

13.2 Application Examples

The genetic and evolutionary algorithms have been widely used for many tech-
nical problems, including optimization in nonlinear and dynamical systems,
designing neural networks, strategy planning, function minima finding, etc. They
are also used for machine learning and for evolving simple programs.
The applications of GA and ES techniques for power-system problems are also
numerous, including:

• Power-system planning (optimal location of FACTS devices) [ 9 ],
• Economic load dispatch [ 15 ],
• Optimal power flow [ 17 ],
• Unit commitment problem solving [ 3 ].

and specifically for protection and control issues:

• Supply restoration and optimal load shedding [ 18 ],
• Power-system stabilizer design [ 1 ],