Optimal tuning of plant-friendly PID controllers

A plant-friendly proportional-integral-derivative (PID) controller optimization framework is proposed to make tradeoffs among set-point response, controller output variations and robustness. The objective function is chosen as the weighted sum of the integral of squared time-weighted error and the integral of squared time-weighted derivative of the control variable with respect to set-point response, while the robustness of the system is guaranteed by constraints on gain and phase margins. Due to the complex structure of the constraints, the problem is solved by genetic algorithms. Simulation analysis show the proposed method could efficiently reduce the controller output variations while maintaining a short settling time. Based on the simulation results, iterative tuning rules for the weighting factor in the objective function are obtained, which allows efficient simple proportional-integral(PI) tuning formulae to be derived.