Høgskolen i Agder: MAS107 Reguleringsteknikk

Prosjektoppgave:

Simulator for nivåreguleringssystem for flistank

Hva oppgaven handler om

  • Utvikling av simulator for flistank med nivåreguleringssystem i LabVIEW Simulation Module
  • Innstilling av nivåregulator
  • Stasjonær settpunktsfølging og forstyrrelseskompensering
  • Målestøyens utslag på pådraget og betydningen av målestøyfiltrering
  • Reguleringssystemets robusthet overfor endringer i prosessparametre
  • Sammenlikning av regulatorers ytelse vha. IAE-indeksen

Utstyr

  • PC med LabVIEW inkl. Simulation Module

Praktiske forhold

Oppgaven utføres fortrinnsvis på datasal 259 (men egentlig kan hvilken som helst av salene benyttes).

Oppgaver

Where you need additional assumptions for your solution, make these assumptions yourself.

  1. Implement a simulator of the wood chip tank level control system quite similar to this KYBSIM simulator. (However, you can design your front panel as you like it.) The mathematical model (prosessmodell) is available via the link "Her er en prosessmodell". The parameter values are available from the fron panel of the simulator. Include the following in the simulator:

    • Random measurement noise (uniformly distributed random noise, also denoted white noise) in the simulator. This noise should be added to the level measurement signal (assumed to have unit of percent). The maximum amplitude of the random noise can be set to 1% (of the measurement range). Use the Uniform White Noise PtByPt (Point-By-Point) function for the noise.

    • A measurement lowpass filter in the form of a first order transfer function with time constant 6s.

    • A PID controller in the form of the PID Advanced function which is described here.
      Note: In the simulator do not open the PID block to use the block (for most function blocks you can open the block by double-clicking it). In stead, wire signals to the inputs and the output of the block.

    • Parameters of the controller, the process and the measurement filter should be adjustable via the front panel.

    • Use the Runge-Kutta 2 method with a step size of 1s. Remember to use seconds or minutes concistently (hours is not convenient here) as time unit in all parts of the simulator where time is involved. I suggest you use minutes here. Probably you want to speed up the simulator (because the chip tank is a very slow process). This is done using the Period parameter of the Input Node of the Simulation Loop.

  2. Check by some proper manual (hand-) calculation that the process part of the simulator (controller and measurement filter not included) behaves correctly. Why is is important that to do such a check?

  3. Tune the PID controller using the Ziegler-Nichols' Closed-loop Method. Is the stability of the control system satisfactory for both controllers? In the subsequent tasks it is assumed that these PID parameter settings are used, unless other instructions are given.

  4. Is the setpoint tracking and disturbance compensation perfect in steady-state with the PID controller?

  5. Demonstrate that the lowpass filter reduces the variance (statistical) of the control signal which is the signal out from the controller. To calculate the variance, us the Variance PtByPt function with the sample length parameter equal to 50. In general, what may be the drawback of large variations of the control signal? Note: The value of the variance is meaningful only under steady-state conditions (thus, the setpoint and the process disturbance should be constant).

  6. The simulator can be used to find out (much about) the robustness of the control system. Assume that, for some reason, the conveyor belt speed is reduced so that the transportation time is increased. Find out from the simulator what is the maximum value of the transportation time that is allowed before the control system becomes unstable. (After this increase of the transportation time, set it back to the nominal value.)

  7. Compare the disturbance compensation performance of the control system with the following controllers: P-controller, PI- controller, and PID controller. Assume that the setpoint is constant. Use the IAE index to express the performance. (You must implement the IAE index calculation in the simulator.)

[Til undervisningsplanen]

Oppdatert 23.4.07 av Finn Haugen, faglærer (e-postadresse: finn@techteach.no).