| Abstract |
The paper presents application of Genetic Algorithms
to a multicriterial optimisation of an electromechanical system
with some nonlinear elements taken into consideration. The
designed control has been tested using a separately excited dc
motor with constant magnetic flux, converter, control system,
AD/DA converters, etc. The control objective is to accomplish a
given distance on a sub-optimal trajectory. Results of GA
adaptation are simulations of a separately excited dc motor
operation with the controller designed by GA, which minimises
selected indices ( ƒ°e2, H ‡ and ( ƒ°e2 + H ‡)) as well as simulation
of the system operation with controllers designed using classical
methods (Kessler, H ‡) presented in this paper. |
