Genetic design of a gain-scheduled controller for a nonlinear concentration control system

The technique of genetic algorithms is proposed as a means of designing gain-scheduled PID control schemes for a class of nonlinear plants where the nonlinearity is a function of the plant output. It is shown that the use of genetic algorithms for this purpose results in highly effective gain-scheduled controller for a highly nonlinear discrete-time single-input single-output model of a concentration control system. The results also indicate that genetic algorithms can be used to design controllers for nonlinear plants where hitherto no design techniques existed.