Modeling a continuous-time feedback system
For the system illustrated in Figure 4.12,
(a) Find its zero-input response, the response with x(t)=0, if the initial value of {y}(t) is \mathbf{y}(0)=1, the initial rate of change of y(t) is y^{\prime}(t)|_{t=0}=0, a = 1, b = 0 and c = 4.
(b) Let b = 5 and find the zero-input response for the same initial conditions as in part (a).
(c) Let the system be initially at rest and let the input signal {x}(t) be a unit step. Find the zero-state response for a = 1, c = 4 and b = -1,1,5.
(a) From the diagram we can write the differential equation for this system by realizing that the output signal from the summing junction is y^{\prime\prime}(t) and it must equal the sum of its input signals
\mathbf{y}^{\prime\prime}(t)=\mathbf{x}(t)-[b\mathbf{y}^{\prime}(t)+c\mathbf{y}(t)]
With b = 0 and c = 4, the response is described by the differential equation \mathbf{y}^{\prime\prime}(t)+4\mathbf{y}(t)=\mathbf{x}(t). The eigenfunction is the complex exponential e^{st} and the eigenvalues are the solutions of the characteristic equation s^{2}+4=0\Rightarrow s_{1,2}=\pm j2. The homogeneous solution is \mathbf{y}(t)=K_{h1}e^{j2t}+K_{h2}e^{-j2t}. Since there is no excitation, this is also the total solution. Applying the initial conditions \mathbf{y}(0)=K_{h\mathbf{1}}+K_{h\mathbf{2}}=1 and y^{\prime}(t)\vert_{t=0}=j2K_{h1}-j2K_{h2}=0 and solving, K_{h1}=K_{h2}=0.5. The total solution is y(t)=0.5\left(e^{j2t}+e^{-j2t}\right)=\cos(2t){\mathrm{~,~}}t\geq0. So, with b = 0, the zero-input response is a sinusoid .
(b) Now b = 5. The differential equation is \ y^{\prime\prime}(t)+5y^{\prime}(t)+4y(t)=x(t), the eigenvalues are s_{1,2}=-1,-4 and the solution is y(t) = K_{h1}e^{-t} + K_{h2} e^{-4t} Applying initial conditions, y(0) = K_{h1} + K_{h2} =1 and {y^\prime}(t)|_{t=0}=-K_{h1}-4K_{h2}=0. Solving for the constants , K_{h1}=4/3,K_{h2}=-1/3 and y({t}){=}(4/3)e^{-t}-(1/3)e^{-4t},t\geq0. This zero-input response approaches zero for times, t > 0.
(c) In this case x(t) is not zero and the total solution of the differential equation includes the particular solution. After t = 0 the input signal is a constant, so the particular solution
is also a constant K_{p}. The differential equation is y^{\prime \prime}(t)+b y^{\prime}(t)+4y(t)=x(t). Solving for K_{p} we get K_{p} = 0 25 . and the total solution is \mathrm{y}(t)=K_{h1}e^{s_{1}t}+K_{h2}e^{s_{2}t}+0.25 where s_{1,2}=\left(-b\pm{\sqrt{b^{2}-16}}\right)/2. The response and its first derivative are both zero at t = 0. Applying initial conditions and solving for the remaining two constants,
(table (a))
The solutions are (table(b))
These zero-state responses are graphed in Figure 4.13.
Obviously, when b = −1 the zero-state response grows without bound and this feedback
system is unstable. System dynamics are strongly infl uenced by feedback.
table (a) | ||||
b | {{s _1}} | {{s _2}} | K_{h{1}} | K_{h{2}} |
-1 | 0.5+j1.9365 | 0.5-j1.9365 | -0.125-j0.0323 | -0.125+j0.0323 |
1 | -0.5+j1.9365 | -0.5-j1.9365 | -0.125+j0.0323 | -0.125-j0.0323 |
5 | -4 | -1 | 0.0833 | -0.3333 |
table (b) | |
b | \mathbf{y}(t) |
-1 | 0.25-e^{0.5t}[0.25\mathrm{cos}(1.9365t)-0.0646\mathrm{sin}(1.9365t)] |
1 | 0.25-e^{-0.5t}[0.25\mathrm{cos}(1.9365t)+0.0646\mathrm{sin}(1.9365t)] |
5 | 0.08333e^{-4t}-0.3333e^{-t}+0.25 |