Genetic Algorithms (GA) are algorithms that can deal with problems that cannot be solved analytically. What is involved in this situation are search processes that are inspired by biological evolution and combine the principle of the "survival of the fittest" with stochastic components. A set (population) of solutions (individuals) is randomly produced and then those which best correspond to a defined criterion are selected (selection). Their properties (parameter values) are then slightly altered (mutation) and combined with one another (recombination) in order to produce a new population of solutions (the new generation). Through repeated selection, recombination and elimination more and more individuals are produced with the aim of increasing the fitness runs of the generations.
Genetic algorithms are the simplest evolutionary optimisation processes, as a result of which they can be rapidly implemented and matched to new problems.
>Search for solutions in multidimensional spaces with the aid of genetic algorithms (GA)
Particle Swarm Optimization
Particle Swarm Optimization (PSO) is a population-based optimisation process that imitates the social behaviour of individuals living in groups: the movement of the particle depends on its own experiences and also on the behaviour of the neighbouring particle. In this way, one can gain an insight into the social behaviour of the population or with regard to technical applications. PSO was applied for the first time in 1995 by J. Kennedy and R. C. Eberhart.
>Search for solutions in multidimensional spaces with the aid of particle swarm optimisation (PSO)
On the basis of this work, many different extensions to particle swarm optimisation have been put forward in the last few years that have aimed, above all, to improve the robustness, speed of convergence and running time of the process and to vali-date it using test functions.
