At this stage, the genetic algorithm is hybridized with the quadratic programming routine.
The routine solves the problem for each chromosome and delivers the corresponding tracking error to the fitness function of the genetic algorithm.
This approach combines the search procedure of the genetic algorithm in the solution space with the local convergence properties of the quadratic programming
solver.
In this paper we use deterministic tournament selection as selection operator. Tournament selection runs a tournament among a few individuals and selects the winner (the one with the best fitness).
Tournament selection has several benefits:it is efficient to code, works on parallel architectures, and allows the selection pressure to be easily adjusted [27].