Optimal FOLPD approximation by GAs and an assessment method

In this paper, an optimal first order lag plus delay (FOLPD) approximation of linear systems is presented, based on minimizing ITAE criterion, using modified genetic algorithms (GAs) with gray coding method and real coding method respectively. An Embedded MATLAB Function is created in order that the population could be manipulated in parallel mode in SIMULINK. One high order process is employed for numerical example. The illustration in time domain demonstrates that the step responses of the proposed methods are more approximative to that of the original plant than some other methods. A new method for the assessment of model reduction is also proposed, which is based on the comparison between the PID robust regions of the original process and of these reduced FOLPD processes. The result reveals that only the proposed method could hold the PID robust regions for the used process.