Ma · Mathematics

How to Find the Impulse and Step Responses of a First-Order ODE

Published By the 9to5Science editorial team

At a glance

  1. 1. Write the model

    Put the equation in the form y' + a y = b x(t), with constant coefficients.

  2. 2. Assume rest

    Use zero initial conditions to find the system's zero-state response.

  3. 3. Apply an impulse

    Set x(t) = δ(t); the output is the impulse response.

  4. 4. Apply a step

    Set x(t) = u(t); the output is the step response.

  5. 5. Check initial conditions

    Add the natural response separately if the system does not start at rest.

A first-order differential equation can describe how an output changes in response to an input. The impulse response is the output caused by a unit impulse, while the step response is the output caused by a unit step. For a constant-coefficient equation, both follow directly from the Laplace transform.

Use the standard causal, zero-initial-state model

y′(t)+ay(t)=bx(t),y'(t)+a y(t)=b x(t),

where x(t)x(t) is the input and y(t)y(t) is the output. For the familiar stable case, assume a>0a>0. If your equation has different coefficients, first divide through by the coefficient of y′y'. The formulas below also show what changes when the initial state is not zero.

Find the transfer function

Take the Laplace transform with zero initial conditions. Since L{y′}=sY(s)\mathcal{L}\{y'\}=sY(s) under that assumption,

sY(s)+aY(s)=bX(s).sY(s)+aY(s)=bX(s).

The transfer function is the ratio of output to input in the Laplace domain:

H(s)=Y(s)X(s)=bs+a.H(s)=\frac{Y(s)}{X(s)}=\frac{b}{s+a}.

The impulse and step responses are found by multiplying this transfer function by the Laplace transform of the corresponding input.

Impulse response

A unit impulse has input x(t)=δ(t)x(t)=\delta(t), whose Laplace transform is 11. Therefore,

Y(s)=H(s)=bs+a.Y(s)=H(s)=\frac{b}{s+a}.

Taking the inverse Laplace transform gives

h(t)=be−atu(t),h(t)=b e^{-at}u(t),

where u(t)u(t) is the unit step. The step factor expresses causality: the response is zero before the impulse is applied.

An impulse at time zero produces an instantaneous change in the output. Integrating the differential equation across the impulse gives the jump condition

y(0+)−y(0−)=b.y(0^+)-y(0^-)=b.

This is consistent with the impulse response: for a system initially at rest, the output jumps from zero to bb and then decays exponentially. If the input impulse occurs at time t0t_0 instead, the response is be−a(t−t0)u(t−t0)b e^{-a(t-t_0)}u(t-t_0).

Unit-step response

For a unit-step input, x(t)=u(t)x(t)=u(t) and X(s)=1/sX(s)=1/s. Thus

Y(s)=H(s)1s=bs(s+a).Y(s)=H(s)\frac{1}{s}=\frac{b}{s(s+a)}.

Using partial fractions,

bs(s+a)=ba(1s−1s+a).\frac{b}{s(s+a)}=\frac{b}{a}\left(\frac{1}{s}-\frac{1}{s+a}\right).

The inverse transform is

s(t)=ba(1−e−at)u(t).s(t)=\frac{b}{a}(1-e^{-at})u(t).

For a>0a>0, the output starts at zero and approaches the steady value b/ab/a. Its transient part decays with time constant 1/a1/a.

Worked example

Find the impulse and unit-step responses of

y′+4y=12x(t)y'+4y=12x(t)

with zero initial state. Here a=4a=4 and b=12b=12, so the transfer function is

H(s)=12s+4.H(s)=\frac{12}{s+4}.

For an impulse input, X(s)=1X(s)=1, giving

Y(s)=12s+4⟹h(t)=12e−4tu(t).Y(s)=\frac{12}{s+4} \quad\Longrightarrow\quad h(t)=12e^{-4t}u(t).

The output jumps by 1212 at the impulse, then decays toward zero. For a unit-step input, X(s)=1/sX(s)=1/s:

Y(s)=12s(s+4)=3(1s−1s+4),Y(s)=\frac{12}{s(s+4)} =3\left(\frac{1}{s}-\frac{1}{s+4}\right),

so

y(t)=3(1−e−4t)u(t).y(t)=3(1-e^{-4t})u(t).

The step response begins at zero and tends to 33, which also follows by setting the derivative to zero in the steady-state equation 4y=124y=12.

Nonzero initial conditions

The impulse and step responses above are zero-state responses: they exclude effects caused by an initial stored value. If y(0)=y0y(0)=y_0 and the input is x(t)x(t) for t≥0t\ge 0, the solution is

y(t)=y0e−at+b∫0te−a(t−τ)x(τ) dτ.y(t)=y_0e^{-at}+b\int_0^t e^{-a(t-\tau)}x(\tau)\,d\tau.

The first term is the natural response from the initial condition; the integral is the zero-state response. For a unit step and y(0)=y0y(0)=y_0, this becomes

y(t)=y0e−at+ba(1−e−at),t≥0.y(t)=y_0e^{-at}+\frac{b}{a}(1-e^{-at}),\qquad t\ge 0.

Common mistakes

  • Forgetting the input gain: In y′+ay=bxy'+ay=bx, the impulse response includes bb; it is not just e−ate^{-at}.
  • Confusing the impulse and step inputs: Their Laplace transforms are 11 and 1/s1/s, respectively.
  • Ignoring initial conditions: The transfer-function method gives the zero-state response. A nonzero initial value contributes a separate natural response.
  • Assuming every case is stable: The formulas apply algebraically when a≠0a\ne0, but decay and a finite step steady state require a>0a>0. If a=0a=0, then y′=bxy'=bx gives impulse response bu(t)b u(t) and step response btu(t)bt u(t).

References

  1. Elementary Differential Equations and Boundary Value Problems, William E. Boyce, Richard C. DiPrima, and Douglas B. Meade
  2. Signals and Systems, Alan V. Oppenheim, Alan S. Willsky, and S. Hamid Nawab

Spotted an error? Send a correction. Read how we write and check articles in our editorial policy.