Process Exam 1-1

pdf

School

University of Houston *

*We aren’t endorsed by this school

Course

3132

Subject

Aerospace Engineering

Date

Apr 3, 2024

Type

pdf

Pages

10

Uploaded by CountCapybaraMaster28

Report
True/ False: 1. A PID controller with a derivative on measurement feature will experience a spike due to a set point change 2. In general, the tuning parameter for set point tracking is more conservative due to a smaller closed-loop time constant. 3. A feedforward controller is realizable if the disturbance dead time is shorter than the process dead time. 4. In a cascade controller, the outer loop is tuned for the set point tracking. 5. One of the reasons to adopt a combined FF/FB controller is to take care of unmeasured disturbance. 6. In a bumpless transfer from the manual mode to auto mode, you set the correct process variable value a the immediate set point (before ramping) and the current valve output as the bias. 7. An override controller is one type of PID controller that is used to compete with another PID controller used for normal operations. NOT False PID controller with derivative on measurement feature is design to have sudden changes leg : setpoint change) : prevent spike in control signal - False tuning parameter is more aggressive because smaller loop allows the system to respond faster to set point changes - > quicker convergence and tighter tracking . - LONGER ! ' False False outer loop : provide fast disturbance rejection inner loop : tunned for setpoint tracking True FF : directly compensate disturbance FB : Correct any remaining error to ensure accurate control - False current controller al output NOT Falze Wrong ! override controller is used to temporarily take control of a system when necessary (eg : emergency)
8. A reverse acting controller has a negative KC. 9. An interacting PID controller is a PI controller in series with a PD controller 10. In gerneral, the tuning parameter set for set point tracking is more conservative due to a smaller closed-loop time constant. 11. A regression has an explained variance of 60 out of a total variance corrected for the mean of 100, then the R2 is 0.4 12. A PID controller with a derivative (on measurement) feature will experience a spike due to a set point change 13. The performance of a PID controller with filter can make the controller output more sluggish and smoother compared to a regular PID controller. 14. In a cascade controller, the inner loop is tuned for disturbance rejection 15. In the primary loop process modeling of cascade control of the jacketed reactor (WS11). The manipulated variable is the set point of the secondary controller (jacket temperature control) set point. False TITIVE True more aggressive False aiming for faster response and tighter tracking False R = = 60/100 = 0 6 False NOT True adding a filter will cause rapid changes in the control signal making the output response slower but smoother False outer loop : provide fast disturbance rejection inner loop : tunned for setpoint tracking True
16. In ITAE, the errors occur at the later part of the response period will contribute more to the penalty criteria 17. One of the reasons to adopt a combined feedforward/feedback controller is to take care of unmeasured disturbances. 18. A feed-forward controller is realizable if the disturbance dead time is shorter than the process dead time 19. A PI controller is often used to avoid excessive valve movements due to noises and to eliminate offset. 20. A PI controller does not have a fixed bias. 21. In a bumpless transfer from the manual mode to the auto mode, you set the current process variable value as the immediate set point (before ramping) and the current valve output as the bias. 22. A bias in a P-only control is the controller output (for valve opening) that will control the process variable as the immediate set point (before ramping) and the current valve output as the bias. - > Integral Time Absolute Error are contranted more d True heavily in the penalty criteria True FF : directly compensate disturbance FB : Correct any remaining error to ensure accurate control men False LONGER ! True PI controller help smooth out control action by integrating past errors , reducing impact of noise and eliminating steady-state offset which can prevent excessive value moment True Pl controller adjust based on the accumulated error over time -- False current - controller output True
Your preview ends here
Eager to read complete document? Join bartleby learn and gain access to the full version
  • Access to all documents
  • Unlimited textbook solutions
  • 24/7 expert homework help
23. A reverse acting controller has a negative K Control Station: 1. What is the bias of a P-Only controller? 2. What is the bias in a PI controller? 3. What is an open loop? 4. What is “proportional band”? 5. What are the major concerns of applying the derivative actions in the presence of large measurement noises? 6. What is the drawback of a P-only controller? What is the drawback of the derivative action? 7. What are the causes of offset? 8. What is the impact of derivative action on response? 9. What is the impact of noise on derivative action? Positive False the controller output (CO) at the designed set point and expected disturbance the controller output at the end of last control period and the beginning of a new control period in manual mode . i . e , without feedback or feed forward control PB = 100/ KC This is the input change that cause 100 % change in the output loss of symmetry in control action Offset is the drawback for 4-only controller . Derivative action amplifies noisy measurements into excessive control value movement and is no good for noisy processes such as flow the cause of offset in a control system include imperfect turning of controller parameters , dead time in system response , and disturbances that the controller cannot fully compensate for Du = Kc (SP-PV) > Du : derived from offset and is needed to reach new SP or counterbalance new disturbance Anti-change , anti-slope , and anti-oscillation Noise is amplified in the controller output . So derivative action is usually avoided in flow and other noisy control loops
10. Write down the expression of derivative on measurement. 11. What is ITAE? IAE? ISE? 12. In PID controller, which action accounts for past history, which for current trajectory, and which for current value? 13. What is PID disturbance rejection? What is set point tracking? 14. What is R-squared in a dynamic model fit? 15. What is a bumpless transfer? 16. When doing a plant test to fit the process model or the disturbance model, what condition is preferred, open-loop or closed-loop? 17. What is a reverse acting controller? 18. What is the impact of too much filtering? - Ec . Ta . dt ITAE : Integral time weighted absolute error/integral of time multiplied by absolute error IAE : Integral of absolute error ISE : Integral of square of error Integral : past history derivative : current trajectory proportional : Current value changes ability to maintain the process variable close to Disturbance rejection : veter to a control system's I in the setpoint despite external disturbance operating condition setpoint tracking : ability of a control system to adjust the process variable to match desired match point R : the goodness fit or the explained variation < z(y + 5) = (y + y) + z(yi y * R2 = SS due to regression = z(yi - j) : SS about the SS due to SS about Total SS corrected for the mean ( 2 (yi - y 2 mean regression regression refer to a seamless transition from manual control mode to automatic control mode , eversa in control system . (it is smoothly transitioned to ensure stability in control system) Ubias (bias controller output) : set to current value of manipulated variable (MV) setpoint (SP) : Set to current value of the process variable (CV) , ensuring smooth transition without sudden changes open-loop . However , almost all plant data are closed loop . Open-loop data are from special tests output : manipulated variable (MU) , controller output (k) , reverse acting controller is characterized by : denoted as "4" Ec > 0 and kp > O Input : Process Variable (PV) /controlled variable (CV) , denoted as "y" increased sluggish in controlled variable (CV) responses lead to larger overshoot . This effect resembles the behavior of P1-controller when a = 1 if a >1 , the responses become even more sluggish than a P1 controller
19. Why does the performance of a PID controller without a derivative filter degrade when the noise causes the controller output to hit an upper or lower limit? 20. Why does Tank 2 remain in manual mode in workshop 7? 21. What are the two things that hinder regulatory control? 22. What is the effect of reset windup on response? 23. What can you say about the balance between the robustness and performance of a controller? 24. If you do the closed-loop plant testing, what change do you have to make to see the dynamics? 25. In general, which tuning parameter set is more conservative, SP tracking or disturbance rejection? 26. What is a direct acting controller? Without a derivative filter , uneven (un-symmetric) manipulated actions can result when the controller output hit an upper or lower limit . Because the "equality or symmetry in randomless" is lost to avoid interaction , we only tune Tank 1 (upper & lower) Deadtime and noise right control is increasingly difficult when Qp > 0 7 Ip it causes delayed return to the new SP because it has to generate enough error to "unwind" the saturated manipulated variable . leg) if during saturation , it continues to integrate to generate excessive "I error that has to be balanced with an equal amount of "-" error by not approaching the new Sp Take jet fighter control scheme as an example . In the attack mode , we stress the performance , which has fast responses and short closed loop Ip . It may , however , cause overshoot and instability . In the landing mode , the stability (safety) requirements demand robustness . No overshoot is tolerated or needed setpoint changes or measured disturbance changes . We can take the controller output data and the response data to figure out open-loop dynamics Disturbance rejection is more conservative Direct acting means that when the input (measurement) is up , the output (controller out to see signal) is also up . Usually it indicate a negative Kc .
Your preview ends here
Eager to read complete document? Join bartleby learn and gain access to the full version
  • Access to all documents
  • Unlimited textbook solutions
  • 24/7 expert homework help
27. What is the benefit and drawback of no overshoot? 28. In PID control, how do you detect the nonlinear behavior in a closed-loop setting? 29. In the jacketed reactor temperature feedforward control workshop (control station), which disturbance variable would you change, and which process variable would you observe to obtain the disturbance model? 30. In the jacketed reactor temperature cascade control workshop (control station), which manipulated variable would you change, and which process variable would you observe to get the open-loop, first order plus dead time model for the primary loop? 31. What are the units of the input (DV) and the output (CV) in the disturbance model in the feed forward plus feedback control of jacketed CSTR (Workshop II)? What is the unit of KD? 32. In the same Workshop 11, what is the unit of the input (MV) and what is the unit of output (CV) in the process model? What is the unit of'Kp? (6%) What is the unit of Kc and why? (4%) Drawback : sluggish response Benefit : minimal variance and less chance of upsetting the unit 1 . Change setpoint 2 . Observe difference in overshoot/oscillation/settling time I Disturbance variable : temperature inlet to cooling jacket Process variable : reactor outlet temperature To Manipulate variable : setpoint of secondary controller (jacket temp control Process variable : reactor outlet temperature [DV) = 02 Ep = % % C or dimensionles [CV] = ° C [MV] = % [Kc = Tip = [CV] = °C [Rp7 : %4 %
33. What is the meaning of P=RA used to characterize a PIDF controller in Control Station case studies plots? Is Kc, positive or negative in this case (4%) 34. What are the meanings ofF=off, D= ideal (meas), and I=ARW? (6%) 35. What is the bias for a 4-mode PIDF controller (4%) 36. What are the reasons to adopt a combined feedforward/feedback controller? 37. What are the units of Kp and Kc of the inner loop in WS10 cascade control of a jacketed reactor 38. where the manipulated variable is the cooling water flow rate (in %) and the controlled variable is 39. the reactor outlet temperature (in °C)? 40. What is the difference between “fluctuations” and “oscillations”? Dynamics: 1. What are the major causes of dead time? Reverse action or Ke > O RA : reverse action Kc > O No filter , -KTpA , Anti Rest Windu the CO is in the beginning of a control cycle Bias is the controller output at the last control period /the beginning of a new control period to take care of unmeasured disturbance for model mismatch · for metering * [Kp] = 0C/ % [Rc] = % / & * Fluctuation : control value Movement(s) controller output Oscillation : peak overshoot ratio higher-order response (like , heat and mass transfer delay) , transportation lag (like fluid How delay) , and simple or instrument lag (like sensor delay such as GC analysis
2. When Kp_ is negative, what does it mean? 3. What is pulse testing? Doublet Testing? 4. What is a typical sampling time? 5. What is a noise band? What is the normally adopted signal to noise ratio? 6. If you do the closed loop plant testing, what change do you have to make to see the dynamics? 7. What is overdamped? 8. How can you tell a process is nonlinear from a y vs. u (steady state) plot? How can you tell a process is nonlinear from )y/)u? 9. What are characteristics that define a process being linear? 10. Write down the first order plus dead time model in terms of a differential equation (time domain expression). Du : the controller output (CO) moves in the opposite direction Dy : The process variable (PV) moves in the opposite direction of the setpoint (SP) Pulse testing : a step change followed by another step change returning to the baseline . Doublet testing : a step change followed by another step change to the opposite side of the baseline , then a final step change returning to the baseline . is is approximately 0 1 times the process time constant Ip noise band : + 32 or + 30 ! signal to noise ratio : 10 : I setpoint changes or measured disturbance changes . We can take the controller output data and the response data to figure out open-loop dynamics the damping coefficient is greater than 1 in a second-order system you can tell a process is nonlinear from a steady-state y vs a plot if the curve is not linear . similarly a process is nonlinear from the slope by if it's not constant a the gain (A) is constant or a steady state y vs . U plot is a straight t line throughout the operating range , kp , Ip and Op are constant y(t) : process variable (PV) at time + Ip : process time constant [p (t) + y(t) = kp m u(t D) kp : process gain m : Input multiplier ult) : input signal D : dead time
Your preview ends here
Eager to read complete document? Join bartleby learn and gain access to the full version
  • Access to all documents
  • Unlimited textbook solutions
  • 24/7 expert homework help
11. In PID control, how do you detect the nonlinear behavior in the presence of large measurement noises? 12. What are the reasons to adopt a combined feedforward/feedback controller? 13. What are the units of Kp and Kc of the inner loop in WS10 cascade control of a jacketed reactor 14. where the manipulated variable is the cooling water flow rate (in %) and the controlled variable is the reactor outlet temperature (in °C)? 15. What is the diference between “fluctuations” and “oscillations”? change setpoint (SP) observe differences in overshoot/oscillation/settling time to take care of unmeasured disturbance for model mismatch · for metering * [Kp] = 0C/ % [Rc] = % / Fluctuation : randon or irregular variations in signal or system oscillation : repetitive and periodic variations around central value 16 . What is the sense of a controller ? sense is an indication of positive or negative (the sign of Kel 17 . What is the difference between PV & CV ? PV : Process variable , which is resulting and recorded (observed) output CV : Controlled Variable , which is the designed output