The present invention discloses an extension to a prior-art genetic algorithm, with which the routing of elevators based on the calls given in an elevator system is formed. A new type of gene, a so-called run type gene, is connected to the chromosome of the genetic algorithm according to the invention, with which gene the desired speed profile for the elevator trip can be set. In this way e.g. an upper limit can be set for the acceleration or for the maximum travel speed of the elevator. By means of the run type gene a kinetic energy term is included in the optimization. The energy consumed by the system can thus be minimized more effectively by means of the algorithm, because the varying travel speeds of the elevators create more freedom of choice for the chromosomes of the algorithm.