Novel Tuning Rules of PD Controller for Industrial Processes

Abstract

The paper presents a simple and an efficient design method of Proportional-Derivative (PD) controller. This type of controllers has been extensively used in industry e.g. to eliminate position error at robotic manipulators and other mechatronic systems. Analytical tuning formulas of PD controller for typical representatives of industrial processes in terms of a single tuning parameter ${\lambda}$ have been derived. The desired performance and robustness indices of the closed-loop system are achieved by suitable selection of the adjustable parameter ${\lambda}$. The PD controller and feed-forward filter are primarily designed to achieve step reference response with negligible overshoot and good response speed, while maximum of the sensitivity function, ${M}_{\mathrm{s}}$, and maximum of the complementary sensitivity functions, ${M}_{\mathrm{p}}$, are used as robustness indices, simultaneously considering maximum, ${M}_{\mathrm{n}}$, of the sensitivity to measurement noise. The effectiveness of the proposed design procedure is verified via numerical simulations on twelve representatives of typical industrial processes: stable, integral and unstable with and without transport delay.