Improvement of Transient Response by Means of Variable Setpoint Weighting

Performance improvement of PI controllers through dynamic set-point weighting

“…Set-point filtering and set-point weighting methods [27][28][29][30][31][32][33][34][35][36][37][38][39][40][41][42][43] are used for achieving an improved transient response due to setpoint change, but usually at the cost of responsiveness with no improvement in the load disturbance response. Instead of a fixed set-point weighting [27][28][29][30][33][34][35][36][37][38][39][40][41][42][43], variable set-point weighting techniques are proposed in [31,32].…”

“…Instead of a fixed set-point weighting [27][28][29][30][33][34][35][36][37][38][39][40][41][42][43], variable set-point weighting techniques are proposed in [31,32]. Leva and Colombo [33] proposed integral square error (ISE) based optimization technique to get the weighting factor.…”

“…These rule-based techniques involve a large number of tunable parameters associated with linguistic models of the weighting factor, and certainly demand more computational efforts. It is found that most of these above methods [11][12][13][28][29][30][31][32][33][34][35][36][37][38] are mainly focused on the improvement of transient response due to set-point change, but only a few are developed for improving the overall response [9,10,16,17,40].…”

“…In the present study, instead of a fixed value or three distinct values [31] of β, a continuously varying weighting factor (β d ) is proposed to get the dynamically weighted set-point (y dw ) for the proportional component of ZNPI. Here, the process operator's knowledge is incorporated in the set-point weighting through a simple parametric form.…”

“…The idea of introducing the dynamic set-point weighting comes from the basic knowledge of a process operator while running a plant manually: ''when the process is moving rapidly towards its set-point, proportional action should be reduced to prevent possible large overshoot and/or undershoot; on the other hand, when the process is moving away very fast from the set-point, proportional action should be increased to restrict such deviations for a quick recovery of the process to its desired value''. We try to implement the above strategy through the dynamic weighting factor β d , defined on the instantaneous change of error (∆e r ) of the controlled variable and normalized dead-time (θ ) of the process [31]. In this context, it is worth mentioning that the present study is motivated by the successful implementations of such operator's knowledge in real-time gain adjustment of ZNPI [9], ZNPID [10], and PI-type self-tuning fuzzy controllers [47,48] with improved set-point response as well as load regulation.…”