An Application of Genetic Algorithms to Active Vibration Control

Abstract

A genetic algorithm is developed for the calculation of the time optimal control force for flexible systems. The algorithm is a robust random search method which uses a simple and compact coding of the control signal as a binary string of maximum pos itive and negative forces. A random population of admissible strings is allowed to evolve through the three processes of selection, reproduction and mutation to produce strings with increasing fitness and optimality. The introduction of stop characters allows the time optimal control problem to be solved in one procedure. A simulation is presented to illus trate the behaviour of the algorithm.