Exploring PID tuning strategies considering noise impact in the IPDT plant control

Abstract

This paper deals with one of the aspects influencing exploitation of the new opportunities provided by the development of the automated control technology base and resulting from the ability to work with relatively short sampling periods. It explores features of two integrated tuning procedures (ITP) for design of filtered series PID controllers of integral-plus-dead-time (IPDT) plants. Under ITP it means a joint design of both the PID controller and the nth order binomial low-pass filter parameters. Since the multiple real dominant pole method is able to yield closed loop performance scalable in a broad range, for an effective tuning of two filer parameters it is necessary to propose some rules helping in exploiting created options. For this purpose, shape related requirements put on the ideal input and output step responses are used. These performance measures are based on concept of piecewise monotonic signals. They represent additional constraints respected in designing as fast as possible transients. Deviations from the ideal shapes due to the noise and plant-model mismatch are quantified in terms of a modified total variation and used in different ways in the proposed ITPs. Possible tuning scenarios and superiority of the proposed approaches in comparison to the traditional PID controller design are demonstrated by simulation. Importance of the presented design for applications in different embedded controllers and programmable devices is stressed.