Scribd Logo
Upload

Welcome to Scribd!

  • 32 views

    Pid Control For Boiler

    Uploaded by

    Engr Faisal Saleem Janjua
    The document discusses the backward difference method for digital implementation of a compensator transfer function. It shows the derivation of the backward difference equation from the Laplace domain transfer function. It then implements this difference equation in Simulink and compares the step responses at different sample times, showing that reducing the sample time provides a better approximation of the continuous system. It also contains questions on open and closed loop systems, with derivations of characteristics equations, root locus plots, and step responses.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd

    Pid Control For Boiler

    Uploaded by

    Engr Faisal Saleem Janjua
    32 views13 pages
    The document discusses the backward difference method for digital implementation of a compensator transfer function. It shows the derivation of the backward difference equation from the Laplace domain transfer function. It then implements this difference equation in Simulink and compares the step responses at different sample times, showing that reducing the sample time provides a better approximation of the continuous system. It also contains questions on open and closed loop systems, with derivations of characteristics equations, root locus plots, and step responses.

    Original Description:

    this document helps tuning pid controller

    Original Title

    pid control for boiler

    Copyright

    © © All Rights Reserved

    Available Formats

    DOCX, PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document discusses the backward difference method for digital implementation of a compensator transfer function. It shows the derivation of the backward difference equation from the Laplace domain transfer function. It then implements this difference equation in Simulink and compares the step responses at different sample times, showing that reducing the sample time provides a better approximation of the continuous system. It also contains questions on open and closed loop systems, with derivations of characteristics equations, root locus plots, and step responses.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    32 views13 pages

    Pid Control For Boiler

    Uploaded by

    Engr Faisal Saleem Janjua
    The document discusses the backward difference method for digital implementation of a compensator transfer function. It shows the derivation of the backward difference equation from the Laplace domain transfer function. It then implements this difference equation in Simulink and compares the step responses at different sample times, showing that reducing the sample time provides a better approximation of the continuous system. It also contains questions on open and closed loop systems, with derivations of characteristics equations, root locus plots, and step responses.

    Copyright:

    © All Rights Reserved
    Download as DOCX, PDF, TXT or read online from Scribd
    Download as docx, pdf, or txt
    You are on page 1of 13

    Question no.

    01
    Backward differencing method:
    𝑈(𝑠) 𝑠 + 𝑎
    𝐶(𝑠) = =
    𝐸(𝑠) 𝑠 + 𝑏

    Simplifying the above equation, we get:


    𝑠𝑈(𝑠) + 𝑏𝑈(𝑠) = 𝑠𝐸(𝑠) + 𝑎𝐸(𝑠)
    By taking the inverse Laplace:
    𝑑𝑢(𝑡) 𝑑𝑒(𝑡)
    + 𝑏𝑢(𝑡) = + 𝑎𝑒(𝑡)
    𝑑𝑡 𝑑𝑡
    Applying the Euler’s backward differentiation, we get:
    𝑢(𝑘) − 𝑢(𝑘 − 1) 𝑒(𝑘) − 𝑒(𝑘 − 1)
    + 𝑏𝑢(𝑘) = + 𝑎𝑒(𝑘)
    𝑇 𝑇
    𝑢(𝑘) − 𝑢(𝑘 − 1 + 𝑇𝑏𝑢(𝑘)) 𝑒(𝑘) − 𝑒(𝑘 − 1) + 𝑇𝑎𝑒(𝑘)
    =
    𝑇 𝑇
    After simplification of the above equation:
    (1 + 𝑏𝑇)𝑢(𝑘) = 𝑢(𝑘 − 1) − 𝑒(𝑘 − 1) + 𝑒(𝑘)(1 + 𝑎𝑇)
    1 1 1 + 𝑎𝑇
    𝑢(𝑘) = 𝑢(𝑘 − 1) − 𝑒(𝑘 − 1) + 𝑒(𝑘)
    (1 + 𝑏𝑇) (1 + 𝑏𝑇) (1 + 𝑏𝑇)
    By putting 𝑘 = 𝑘 + 1 (next sample of digital signal):
    1 1 1 + 𝑎𝑇
    𝑢(𝑘 + 1) = 𝑢(𝑘) − 𝑒(𝑘) + 𝑒(𝑘 + 1)
    (1 + 𝑏𝑇) (1 + 𝑏𝑇) (1 + 𝑏𝑇)

    Hence it is proved that, the backward difference method results in digital implementation of
    compensator.
    Question no. 02
    Continuous time representation of system using
    Part (a): Forward differencing implementation
    The forward differencing implementation of equation
    1 1 1 + 𝑎𝑇
    𝑢(𝑘 + 1) = 𝑢(𝑘) − 𝑒(𝑘) + 𝑒(𝑘 + 1)
    (1 + 𝑏𝑇) (1 + 𝑏𝑇) (1 + 𝑏𝑇)
    is done in MATLAB Simulink and the is shown in figure 1:

    Figure 1: Forward difference implementation in Simulink

    The step response of figure 1 with sample time of 0.001 𝑠𝑒𝑐, 0.01 sec and 157 𝑚𝑠𝑒𝑐 is shown in figure
    2 to figure 4 respectively.
    Figure 2: Step response of figure 1 with Ts=0.001 s:

    Figure 3: Step response of figure 1 with Ts=0.01 s


    Figure 4: Step response of figure 1 with Ts=157 ms

    Part (a): Forward differencing implementation


    The backward differencing implementation of equation
    1 1 1 + 𝑎𝑇
    𝑢(𝑘) = 𝑢(𝑘 − 1) − 𝑒(𝑘 − 1) + 𝑒(𝑘)
    (1 + 𝑏𝑇) (1 + 𝑏𝑇) (1 + 𝑏𝑇)
    is done in MATLAB Simulink and the is shown in figure 5:

    Figure 5: Backward difference implementation in Simulink


    The step response of figure 5 with sample time of 0.001 𝑠𝑒𝑐, 0.01 sec and 157 𝑚𝑠𝑒𝑐 is shown in figure
    6 to figure 8 respectively.

    Figure 6: Step response of figure 5 with Ts=0.001 s


    Figure 7: Step response of figure 5 with Ts=0.01 s

    Figure 8: Step response of figure 5 with Ts=157 m s


    It can be observed from the step responses of figure 1 and figure 5 at different sample times that,
    by reducing the sample time, discrete system becomes a good approximation of continuous system.
    By increasing the sample time, the approximation becomes very poor and ultimately the false
    approximation of continuous system.

    Question 03
    Problem A-6-19
    Part (1): Open loop transfer function
    10(𝑠 + 𝑎)
    𝐺(𝑠) =
    𝑠(𝑠 + 1)(𝑠 + 8)
    Part (2):
    Closed loop characteristics equation:
    𝑠 3 + 9𝑠 2 + 19𝑠 + 10𝑎 = 0
    Part (3):
    Simulink diagram:

    Figure 9: Simulink diagram


    Part (4):
    Root locus:
    %% problem 6-19
    K1=tf(1,[1 8]);
    G1=tf(10,[1 1 0]);

    CL1=feedback(G1*K1,1);
    figure;
    rlocus(CL1)
    grid on

    Figure 10: Root locus of figure 9


    Part (5)-(7)
    Value of a:

    𝑎=2
    figure;
    plot(t,st,'linewidth',1.5)
    hold on
    plot(t,sr6_19,'linewidth',1.5)
    grid on
    title('Step response')
    legend('Reference','Output')

    Figure 11: Step response of figure 9


    Question 04
    Problem A-6-20
    Part (1): Open loop transfer function
    10
    𝐺(𝑠) =
    𝑠(𝑠 + 1 + 10𝑘)

    Part (2):
    Closed loop transfer function:
    10
    𝐺𝑐 (𝑠) =
    𝑠(𝑠 + 1 + 10𝑘)
    10
    1+
    𝑠(𝑠 + 1 + 10𝑘)
    After simplification:
    1
    𝐺𝑐(𝑠) =
    𝑠2 + 𝑠(1 + 10𝑘) + 10

    Part (3):
    Closed loop characteristics equation:
    𝑠 2 + 𝑠 + 10𝑘𝑠 + 10 = 0
    Now, rearranging above equation:
    10𝑘𝑠
    1+ =0
    𝑠2 + 𝑠 + 10
    Taking 𝐾 = 10𝑘, we get:
    𝐾𝑠
    1+ =0
    𝑠2 + 𝑠 + 10
    Part (4)-(6):
    Simulink diagram:

    Figure 12: Simulink diagram

    Part (5)-(6):
    %% problem 6-20
    k=1;
    sys=tf(10,[1 1]);

    CL20a=feedback(sys,k);
    CL20=feedback(CL20a*tf(1,[1 0]),1);
    figure;
    rlocus(CL20)
    grid on
    Figure 13: Root locus of figure 12

    Step Response:
    figure;
    plot(t,st20,'linewidth',1.5)
    hold on
    plot(t,sr6_20,'linewidth',1.5)
    grid on
    title('Step response')
    legend('Reference','Output')
    ylim([0 1.2])
    Figure 14: Step response of figure 12

    You might also like