CASCADE CONTROL



Description of the simulated systems

In this simulator two control systems are simulated in parallel (i.e. simultaneously), see the front panel of the lab:

  • A cascade control system consisting of two control loops - the primary loop and the secondary loop, each containing a PID-controller.
  • An ordinary, single-loop control system containing one PID-controller.

Note that both control systems are excited by the same setpoint, r1, and the same disturbance, d.

In both control systems the process to be controlled consists of two sub-processes in series:

  • Sub-process 2 ("quick"): y2 = P2*u + d, where u is the manipulated variable, and d is the disturbance. P2 is dynamically a third order system with gain 1 and response-time 1 sec.
  • Sub-process 2 ("slow"): y1 = P1*y2, where y2 is the output of sub-process 2. y1 = y is the output of the whole process (consisting of the two sub-processes in series). P1 is dynamically a second order system with gain 1 and response-time approx. 5 sec.

The combined blocks of sensors and subsequent scaling M, M1 and M2 are here simply gains of 1.

The controllers available are PID and on/off. The PID-controllers shall be used under normal operation of the control systems, but the on/off-controllers can be used for controller parameter tuning (in the Åstrøm-Hägglund's autotuning method). The controller functions (PID-controller and on/off-controller) are described here.

The performance of the controllers is quantified through the individual IAE indices. The IAE is the time-integral of the absolute value of the control error. The less IAE, the better control. However, the IAE does not measure the usage of the control variable.


Aims

The aims of the tasks given below are

  • to give experience in tuning the controller parameters
  • to demonstrate the benefits of using cascade control compared to ordinary signle-loop control, with particular focus on disturbance compensation.

Motivation

Cascade control is a common control structure in both process control systems and in servomechanisms. In process control cascade control is used for temperatur control, level control, pressure control, and quality control. These are the primary control loops, while the secondary loop typically performs flow control or pressure control.

In servomechanisms the primary loops perform positional control or speed control, while the secondary loop typically performs speed control (in a primary positional loop), or current control (in electrical servomechanisms).

Cascade control can give a much better compensation for disturbances than single loop control can do.


Tasks

In the tasks below the starting point is that the process is in it's nominal operating point which is characterized as follows: The set-point is 40 (unit %), the disturbance is  -10, and the process outputs y1 and y2 are 40%. The nominal value of the manipulated variable, which keeps the process at the operating point, is 50% for all the controllers.

  1. Controller tuning: Find proper parameters for the primary controller C1 and for the secondary controller C2 in the cascade control system, and for the (single) controller C in the single-loop control system. (You may use the Ziegler-Nichols' closed-loop method, or Åstrøm-Hägglund's autotuner.) Note: C1 will be used as a PID-controller, while C2 will be used as a PI-controller. C will be used as a PID-controller.
    (Tip: Remember to tune the secondary controller before the primary controller, and the latter must then be in manual mode, with a proper nominal value of the manipulated variable. Excite the control system with a small step in the set-point. The controller parameters should not be very far from the following: C1: Kp=7.2, Ti=5, Td=1.25. C2: Kp=1.8, Ti=2, Td=0. C: Kp=4.2, Ti=6, Td=1.5.)

     

  2. Cascade control versus single-loop control: 
    1. Disturbance compensation: Observe how the two control systems manage to compensate for a step in the disturbance. Let the step go from -10 to -30, but you may well try other step heights, too. Which  of the two control systems gives best compensation (compare the IAE values)?

      Observe also how the control variable works in the two control systems. In which of them is the control variable working most actively?

    2. Setpoint tracking: Observe how the two control systems are able to make the process output track or follow a step in the setpoint. Which of them are best (compare the IAE values)?

[KYBSIM] [TechTeach]


Updated April 26, 2006. Developed by Finn Haugen. E-mail: finn@techteach.no.