Expert autotuner for multiloop SISO controllers

A novel expert autotuner for multivariable control applications is proposed. The main idea of the autotuning control methodology is to apply the SISO (single-input-single-output) PID controllers in the MIMO (multi-input-multi-output) environment with appropriate tuning based on the pattern recognition technique. An expert tuning processor is designed to supervise the proportional, integral, derivative (PID) controller to provide good tuning parameters from the observed MIMO response. A customized SISO process identification technique based on pattern recognition is developed for closed loop application. Multivariable tuning is then activated using a detuning factor on the SISO settings. A new technique to identify the relative disturbance gain online is also included to assist the decision on tuning the controller upon natural load disturbances. The time response of the autotuning controller is compared to some tuning methods reported in the literature