A tuning method of two degrees of freedom PID controller

This paper proposes a new tuning method when using a two degrees of freedom proportional-integral-derivative (PID) controller. Its control performance for a first-order lag plus deadtime system is shown as an example of the commonly approximated controlled plants in the heating, ventilating, and air-conditioning (HVAC) field. Reference and disturbance input changes to a conventional PID controller do not necessarily give a satisfactory response. To overcome this problem, several two degrees of freedom PID (2 DOF PID) algorithms have been developed to replace the conventional PID controllers. However, when these techniques are applied to real plants, it is not usually easy to obtain a set of optimum tuning parameters. To evaluate the control performance, a comparison is carried out between the two tuning methods--the optimization technique and the partial model matching method--using simulation. Graphs by which the controller parameters can be related to dynamics of a popular plant are developed. The 2 DOF PID control is studied taking into account modeling error due to a change in plant characteristics. It is found that the 2 DOF PID controller designed based on the partial model matching method is robust and useful in simulations where a traditional PID controller would be used.