Stabilization subspaces based DPGA for optimizing PID controllers

A distributed population genetic algorithm (DPGA) based on stabilization subspaces of PID controller parameters obtained by generalized Hermite-Biehler theorem for unstable first order plus dead time (FOPDT) process is developed. According to generalized Hermite-Biehler theorem, the stabilization regions of PID controller parameters of stable controlled plant can be obtained, then the initial subspaces of DPGA can be derived. As to an unstable FOPDT process, it is first stabilized by the inner loop of P or PI controller, and the outer loop of PID controller is optimized by DPGA. Simulation results of several unstable FOPDT plants and bioreactor show the efficiency of the proposed method.