Scribd
Upload
110 views6 pages

Convolution Integral Tutorial and Examples

The convolution integral is used to compute the forced response of an LTI system. It involves integrating the product of one function shifted and flipped over the other. To evaluate the convolution integral, the key is determining the correct integration limits by sketching the graphs of the functions and seeing their overlap at different values of t. Examples are worked through step-by-step to demonstrate finding the integration limits and evaluating the convolution for both continuous and discontinuous functions.

Uploaded by

MaMa Baloch
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
110 views6 pages

Convolution Integral Tutorial and Examples

The convolution integral is used to compute the forced response of an LTI system. It involves integrating the product of one function shifted and flipped over the other. To evaluate the convolution integral, the key is determining the correct integration limits by sketching the graphs of the functions and seeing their overlap at different values of t. Examples are worked through step-by-step to demonstrate finding the integration limits and evaluating the convolution for both continuous and discontinuous functions.

Uploaded by

MaMa Baloch
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

Convolution Integral

Reading and Watching:


Convolution animation short video:
[Link]
[Link]
The convolution integral is a well-known mathematical operation. In this course, we use it to
compute the forced response of an LTI system, but in this lecture note we will talk about the
convolution in general.
When we say y(t) is the convolution of x(t) and z(t) we write it with an asterisk notation as
follow:

First, we should note that:

𝑦 𝑡 𝑥 𝜆 𝑧 𝑡 𝜆 𝑑𝜆 𝑧 𝜆 𝑥 𝑡 𝜆 𝑑𝜆

If x(t) is the input and h(t) is the system impulse response, then the forced response is:

Note that the integration limits go from ∞ 𝑡𝑜 ∞. Of course the limits change based on the
two functions being convolved. For instant, if the system is “causal”, its impulse response, h(t)
is zero for negative time and we frequently see the limits of the convolution starts at zero (not
negative infinity).
When evaluating the convolution integral, we frequently find that the main difficulty is in
finding the correct integration limits (rather than the integration itself).
Example (1)
consider the following convolution example:

The two waveforms are shown below:


Let us write y(t) as follow:

It’s important to note that, while performing the above integration, “t” is constant, and we are
integrating over . Also note that u(t) is either “1” or “0”, so the integration is trivial, however,
just looking at the integration expression, one needs to know the limits of the integration, and
you should also note that the integration limits will change based on the value of “t”.
To obtain the integration limits correctly, you are highly advised to sketch the function and use
the graph to determine the limits for different ranges of “t” as shown below.

In this problem there are three regions:


- For 𝑡 0 there is no overlap between x() and z(t-), then the output y(t)=0
- For 0 𝑡 2 The overlap interval is between =0 and =t. Then:
- For 𝑡 2 The overlap interval is between =0 and =2. Then:

Therefore, the convolution result y(t) can be written as:

While solving this problem, we went through the following steps:


- Choose one of the two function, x(t) or z(t) (we chose x(t)), express it in terms of  
x() and fix it.
- Take the second function (z(t)), express it in terms of   z() then: flip it around the y-
axis  z(-) and slide it by “t”  z(t-)
- Multiply x() and z(t-)
- Integrate the product.
- Repeat for different range of “t”
The number of time intervals depends on the two convolving functions. Finding the integration
limits is easy when the two functions are continuous, but discontinued functions must be handled
with patience.
Example (2):
Convolve 𝑥 𝑡 𝛿 𝑡 and ℎ 𝑡 10 𝑒 u(t)
This problem is simple; we do not need to graph it.
𝑦 𝑡 𝑥 𝑡 ∗ℎ 𝑡 ℎ 𝜆 𝑥 𝑡 𝜆 𝑑𝜆

𝑦 𝑡 10 𝑒 𝛿 𝑡 𝜆 𝑑𝜆

𝑦 𝑡 10 𝑒 𝑢 𝑡 ℎ 𝑡
As expected. One of the two signals is the delta function, then the output is the second function
(recall that the forced output of an LTI system is the input convolved with the system impulse
response. If the input is the delta function the output is the impulse response).
Example 3:
Convolve the following two functions: 𝑥 𝑡 𝑢 𝑡 and ℎ 𝑡 10 𝑒 u(t)

Example 4:
Convolve x(t) and h(t) defined as follows:
Let us write the convolution as follow:

We chose to keep h fixed while flipping and sliding x


There are 5 regions: [ 𝑡 2, 2 𝑡 4, 4 𝑡 5, 5 𝑡 7, 𝑡 7]
Carry out the integration in each region. The convolution result is y(t) as follow

More convolution examples and background explanation can be found in: E. Kamen and B.
Heck, Fundamentals of Signals and Systems using the Web and MATLAB, 3rd Ed., Pearson
Prentice Hall, 2007. The relevant section is copied and is included in a separate file as a reading
material. Please note that in SYSC3600 we only focus on continuous time signals so you may
ignore the discrete examples in Kamen.

Common questions

Powered by AI

Handling continuous functions in a convolution integral is generally more straightforward because their overlap regions are clear and consistent. However, for discontinuous functions, patience is required as one must carefully account for abrupt changes that affect how the functions overlap. These changes can introduce multiple integration limits within the same interval of t, requiring detailed analysis to ensure accuracy in the convolution result .

The integration limits in a convolution integral change based on the overlap between the functions x(λ) and z(t-λ) as time t varies. This typically occurs when the functions have different support or length. If t changes the manner in which the functions overlap (e.g., if one function is causal and steps in amplitude or duration), the integration limits need to reflect these overlaps at each incremental step of t. Accurate integration limits are essential to calculate the convolution correctly .

Graphing is unnecessary when a delta function is involved in convolution because the delta function simplifies the convolution process. Its nature as an identity in convolution means it 'selects' the corresponding point from the other function, making the graphical overlap analysis redundant. Thus, the result is simply the value of the other function at the point of the delta's spike .

When the input function is a delta function, the convolution process simplifies significantly because the result of the convolution is simply the system's impulse response. The convolution integral reduces to evaluating the impulse response at the point of impulse, as the delta function acts as an 'identity' under convolution .

Solving a convolution integral for continuous functions involves several key steps: choosing one of the functions (e.g., x(t)), expressing it in terms of a variable (e.g., λ), flipping and sliding the other function (e.g., z(t) to z(t-λ)), multiplying the two functions, and integrating the product. Finding the correct integration limits is crucial because the integration is performed over the region where the functions overlap, which can vary with t. Incorrect limits result in an inaccurate calculation of the convolution .

In causal systems, where the impulse response is zero for negative time, the convolution integral’s limits are often adjusted to start at zero instead of negative infinity. This change impacts the calculation by considering only the portions of the input that coincide with the system’s active response, thus respecting the causal nature by not allowing future inputs to affect the output prior to their input occurrence .

Flipping and sliding one of the functions in the convolution process is critical to align and analyze the overlap of the input signal with the system's response. By flipping the function across the y-axis and sliding it incrementally over the other function, this method facilitates the precise determination of overlap intervals as a function of time, directly affecting the setup of the correct integration limits needed to evaluate the convolution .

The convolution operation relates the system's impulse response to the system's input and output signals by providing a systematic way to calculate the output signal as a function of time. In an LTI system, the output y(t) is defined as the convolution of the input signal x(t) with the impulse response h(t), reflecting how each component of the input affects the system at different times .

Graphical methods aid in determining the integration limits by providing a visual means to identify where the two functions overlap as they are being convolved. By sketching the functions and their shifts, one can easily see how the overlap regions change with respect to time, allowing precise determination of integration limits necessary for evaluating the convolution integral accurately .

The convolution integral is critical for determining the forced response of an LTI system because it mathematically describes how an input signal is transformed by the system's impulse response. By convolving the input x(t) with the system's impulse response h(t), one can determine the system's output y(t), which reflects how the system reacts over time to the input signal .

You might also like