1 / 9

Introduction to PID control

Introduction to PID control. Sigurd Skogestad. e. y m. Block diagram of negative feedback control. PID controller. P-part: Input change ( Δ u) proportional to error I-part: Add contribution proportional to integrated error. Will eventually make e=0 (no steady-state offset!)

jspurlock
Download Presentation

Introduction to PID control

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Introduction to PID control Sigurd Skogestad

  2. e ym Block diagram of negative feedback control

  3. PID controller • P-part: Input change (Δu) proportional to error • I-part: Add contribution proportional to integrated error. • Will eventually make e=0 (no steady-state offset!) • Possible D-part: Add contribution proportional to change in (derivative of) error TUNING OF PID-CONTROLLER. Want the system to be (TRADE-OFF!) • Fast intitially (Kc large, D large) • Fast approach to steady state (I small) • Robust / stable (OPPOSITE: Kc small, I large) • Smooth use of inputs (OPPOSITE: Kc small, D small) NOTE: Always check the manual for your controller! Many variants are in use, for example: • I = Kc/I • Proportional band = 100/Kc • Reset rate = 1/ I

  4. Tuning of your PID controllerI. “Trial & error” approach (online) • P-part: Increase controller gain (Kc) until the process starts oscillating or the input saturates • Decrease the gain (~ factor 2) • I-part: Reduce the integral time (I) until the process starts oscillating • Increase a bit (~ factor 2) • Possible D-part: Increase D and see if there is any improvement

  5. II. Model-based tuning (SIMC rule) • From step response • k = Δy(∞)/ Δu – process gain •  - process time constant (63%) •  - process time delay • Proposed controller tunings

  6. Example SIMC rule) • From step response • k = Δy(∞)/ Δu = 10C / 1 kW = 10 •  = 0.4 min (time constant) •  = 0.3 min (delay) • Proposed controller tunings

  7. Simulation PID control • Setpoint change at t=0 and disturbance at t=5 min • Well tuned (SIMC): Kc=0.07, taui=0.4min • Too long integral time (Kc=0.07, taui=1 min) : settles slowly • Too large gain (Kc=0.15, taui=0.4 min) – oscillates • Too small integral time (Kc=0.07, taui=0.2 min) – oscillates • Even more aggressive (Kc=0.1, taui=0.2 min) – unstable (not shown on figure) 3 4 Output (y) 2 1 setpoint 1 2 3 4 Time [min]

  8. Comments tuning • Delay (θ) is feedback control’s worst enemy! • Common mistake: Wrong sign of controller! • Controller gain (Kc) should be such that controller counteracts changes in output • With negative feedback: Sign of Controller gain (Kc) is same as sign of process gain (k) • Alternatively, always use Kc positive and select between • ”Reverse acting” when k is positive (most common case) • because MV (u) should go down when CV (y) goes up • ”Direct acting” when k is negative • WARNING: Some reverse these definitions (e.g. wikipedia)

More Related