Using a Homogeneous Transformation Matrix to Combine Rotation and Translation

  Рет қаралды 11,226

Christopher Lum

Christopher Lum

Күн бұрын

In this video we discuss how to properly deal with coordinate frames that are both rotated and translated from one another. We develop a homogeneous transformation matrix which combines a standard rotation matrix with a translation. This matrix can be used to relate the pose of a system in one frame to the pose in the other frame.
Topics and Timestamps:
0:00 - Introduction
1:24 - Development of homogeneous transformation in 2D
27:08 - Example
41:15 - Composite/successive transformations
59:35 - Extension to 3D
Lecture notes and code can be downloaded from github.com/clum/KZfaq/tree/...
All Flight Mechanics videos in a single playlist ( • Flight Mechanics )
#FlightMechanics
You can support this channel via Patreon at / christopherwlum or by clicking on the ‘Thanks’ button underneath the video. Thank you for your help!

Пікірлер: 64
@ChristopherLum
@ChristopherLum 4 жыл бұрын
In case it is helpful, here are all the Flight Mechanics videos in a single playlist kzfaq.info/sun/PLxdnSsBqCrrEx3A6W94sQGClk6Q4YCg-h. Please let me know what you think in the comments. Thanks for watching!
@Gholdoian
@Gholdoian 2 ай бұрын
AE512: Interesting to see the breakdown of translation and rotation in a single matrix, I've usually performed the operations separately but nice to see you can easily perform successive transformations.
@aimeepak717
@aimeepak717 2 ай бұрын
AE512: Great lecture. I was able to do the homework following this step by step.
@milesrobertroane955
@milesrobertroane955 4 ай бұрын
AA516: Good video! Augmenting the matrix makes total sense after seeing the composite transformations
@edwardmau5877
@edwardmau5877 2 ай бұрын
AE 512 Very impressive the measurements on the board matched the mathematical calculations, nice steady drawing skills.
@SayedTorak
@SayedTorak 2 ай бұрын
AE 512: great video prof. Lum
@zaneyosif
@zaneyosif 2 ай бұрын
AE512: Great lecture, especially the 3D example. Really helped put things into perspective with different objects with different reference frames (translationally/rotationally) and how to relate them all together via transformation matrices.
@reesetaylor3506
@reesetaylor3506 2 ай бұрын
AE 512: Nice lecture, the physical examples with measurments definitely help with my understanding, thanks!
@user-hu5zs4ur7i
@user-hu5zs4ur7i 5 ай бұрын
AA516: Allie S, this is really neat! I haven't thought about this type of transformation being possible and I am excited to learn even more uses for it.
@anthonysu1712
@anthonysu1712 2 жыл бұрын
- no questions; just wanted to say I'm so thankful you derived the 2D case and extended it to 3D instead of directly deriving the 3D case :)
@rowellcastro2683
@rowellcastro2683 5 ай бұрын
AA516: I was going to ask this in Office Hours but I believe I figured it out. For anyone that might be wondering in 33:10 why the rotation matrix is different from what we learned in the past rotation lecture (Flight Mech 02), it is because the rotation from frame b to frame 1 is left handed! That is why it is transposed/inversed.
@rowellcastro2683
@rowellcastro2683 5 ай бұрын
That also might be why Dr. Lum wrote it incorrectly in the beginning.
@Kumky605
@Kumky605 4 ай бұрын
AA516: Forgot to comment on this when I watched it the first time but great video! These types of transformations were super interesting and I was excited to go more indepth
@steahrcase8379
@steahrcase8379 2 жыл бұрын
AE512: I absolutely love your real world examples and enthusiasm
@ChristopherLum
@ChristopherLum 2 жыл бұрын
Great, I'm glad it is entertaining, hopefully it applies to HW as well :)
@manitaregmi6932
@manitaregmi6932 2 жыл бұрын
AA 516 - You made it real easy to understand by Developing the homogeneous transformation in 2D and then doing the Extension to 3D. Great explanation on material as always!
@yaffetbedru6612
@yaffetbedru6612 5 ай бұрын
AA516: Even though the reference frames started adding up, it still was easy to follow and see how powerful the homogeneous transformation matrix is.
@charlesharmon9065
@charlesharmon9065 2 жыл бұрын
AE 512 - Great video. Sketches of frames/coordinate systems and inclusion of numerical examples were extremely helpful in making the material 'stick.'
@boeing797screamliner
@boeing797screamliner 3 жыл бұрын
AA516 - I really like how you incorporate real-world examples into these lectures. It really helps me grasp the material
@ChristopherLum
@ChristopherLum 3 жыл бұрын
Gary, thanks for the feedback. And thanks for actively participating in the discussion, you are rocking the participation points 😀!
@koshiroyamaguchi9613
@koshiroyamaguchi9613 2 жыл бұрын
AA516 - So far I really appreciate Prof. Lum's lecture styles with various real world/computaitonal examples. Looking forward to learning more lectures!
@PatrickGalvin519
@PatrickGalvin519 5 ай бұрын
AA 516: Great video, very informative lecture.
@sianhoan833
@sianhoan833 Жыл бұрын
AA 516: Another great lecture! Glad to see that our aircraft can start flying around instead of only rotating in one place and excited for what comes next.
@ChristopherLum
@ChristopherLum Жыл бұрын
Yes, and it only gets better, we'll start flying in earnest soon!
@jackpascho2561
@jackpascho2561 Жыл бұрын
AA 516: One of my favorite lectures so far. Doing a multi-stage chain transformation like this has always been something I thought about particularly in regard to air combat with all of the sensor and radar data that gets distributed between aircraft, and has been something I wish I could do on a variety of projects. Very excited that I understand the fundamentals behind this process!
@ChristopherLum
@ChristopherLum Жыл бұрын
Jack, great, I'm glad it was simulating. This comes into play a lot when working with robotics.
@shavykashyap
@shavykashyap 5 ай бұрын
AA 516 - Great Lecture
@bsgove
@bsgove 28 күн бұрын
AE512: It's amazing how someone can keep all of these building blocks in mind while trying to create a dynamic simulation where this is only one small part.
@NikolajKuntner
@NikolajKuntner 4 жыл бұрын
Good topic, I'm also very much into spatial transformations atm.
@ChristopherLum
@ChristopherLum 4 жыл бұрын
I'm glad it was helpful. There are other similar videos on the channel, please feel free to check them out and let me know what you think. Thanks for watching!
@davidtelgen8114
@davidtelgen8114 2 ай бұрын
AE512: now I won't be confused when I hear a GNC engineer talk about 'pose'!
@petermay6090
@petermay6090 5 ай бұрын
AA516: Very useful!
@nathanlipshutz3140
@nathanlipshutz3140 2 ай бұрын
AE512: During a rigid body dynamics calculation, I have previously done something where I used a combination of a vector addition and rotation matrices to accomplish something similar. I think this could have expedited my workflow.
@thomasstiglich3484
@thomasstiglich3484 4 жыл бұрын
Really enjoying your videos, hope to get through each one. I did find one thing confusing though, as may others. In a previous video, "Expressing Vectors in Different Frames Using Rotation Matrices", when a vector is known in Fr and desired in F1, the angle Theta was calculated in the RHS from Fr to F1, resulting in [cos(theta) sin(theta); -sin(theta) cos(theta)]. In this video, the rotation from Fb to F1 was [cos(alpha) -sin(alpha); sin(alpha) cos(alpha)] as the result of the angle being defined as the RHS angle from F1 to Fb. If the angle was from Fb to F1 as in the prior video, the prior rotation matrix could be used. Wolfram Mathworld uses [c(psi) s(psi); -s(psi) c(psi)] where psi is the RHS angle from the reference frame to the rotated frame, as in your first video. Do you have a preference as to how the angle is defined?
@AliG.G
@AliG.G 3 жыл бұрын
As far as I understand is both r same. Here is a short video demonstration LHS and RHS referencing kzfaq.info/get/bejne/gctjm5p_t7LNYoE.html. If I remember right in the other video Z was facing down i.e. NED frame of reference and in this video, Z comes out of the screen towards you in ENU frame.
@divinadileep9464
@divinadileep9464 3 жыл бұрын
tnku so much sir
@princekeoki4603
@princekeoki4603 5 ай бұрын
AA516 : Cool how successive transformations stack on top of each other.
@sethwhittington28
@sethwhittington28 2 жыл бұрын
AE512: these transformations keep building on eachother and getting more complicated! I think you’re doing a great job breaking it down piecewise though.
@ChristopherLum
@ChristopherLum 2 жыл бұрын
These come in handy, especially in robotics applications.
@sethwhittington28
@sethwhittington28 2 жыл бұрын
@@ChristopherLum @28:10 you measure a 40 degree angle to define the rotation matrix that takes you from the b frame to the 1 frame C1/b(psi). In the diagram on the right side of the board, it looks like this 40 degree angle is actually the rotation that would take you from the 1 frame to the b frame, but in your homogenous matrix at the bottom of the screen you define C1/b which i think is the rotation that takes you from the b frame to the 1 frame. Should you have used NEGATIVE 40 degrees here to go from b to 1...?
@WalkingDeaDJ
@WalkingDeaDJ 2 ай бұрын
Jason-AE512: This is helpful to understand the homework question 1
@steahrcase8379
@steahrcase8379 2 жыл бұрын
AE512: at 46:30, why do you measure theta from x2 to xr all the way around? In the previous case, x1 to xr, it was the smaller angle. Why is it different here (why do we not measure from xr to x2)?
@tonymcdaniel6878
@tonymcdaniel6878 5 ай бұрын
AA 516: On the last statement made in the video, when you went to the 3D rotation and translation matrix, what if the target reference frame was at two different rotation ie a different psi and theta
@prathammaheshwari4489
@prathammaheshwari4489 3 жыл бұрын
sir can you upload dynamics also like forces and moments
@BryanLiberman
@BryanLiberman 2 ай бұрын
AE512: Thanks Dr. Lum! Off the top of your head, do you have any resources for how to concatenate matrices in mathematica? I think it might be easier to build up the zeta matrix one piece at a time, and put it all together!
@ChristopherLum
@ChristopherLum 2 ай бұрын
Yes, try the Join function. I have a video on the channel showing matrices in Mathematica if you would like more details
@habesharockets7460
@habesharockets7460 2 жыл бұрын
AA516: Hi professor! The video is great as always. At 19:00 min, we used r(t/b) frame 1 instead of frame r (reference frame). Is that because r(t/b) origin aligns with frame 1? (Daniel Teshome) Thank you!
@ChristopherLum
@ChristopherLum 2 жыл бұрын
Hi Daniel, yes, sometimes we need to translate (but not rotate) the frames to make the downstream rotations work.
@kaikim8402
@kaikim8402 3 жыл бұрын
Question : @16:14 Please correct me if i am wrong. The rotational matrix C1/b(psi) from Fb to F1, shouldn't the psi in the -ve direction since the Z^r is coming out of the plane?
@ChristopherLum
@ChristopherLum 3 жыл бұрын
Hi Kai, Great to hear from you again. I've started answering questions directly through my Patreon page at www.patreon.com/christopherwlum. Please free to submit your question there and we can discuss. Thanks for watching and for supporting the channel!
@ojasvikamboj6083
@ojasvikamboj6083 Жыл бұрын
A A 516: Ojasvi Kamboj
@ToanLuu158
@ToanLuu158 2 ай бұрын
AE512: I wondered what is the maximum frames that someone had been developing in the real world using this transformation matrix method.
@timproby7624
@timproby7624 2 ай бұрын
[AE 512] 42:40 You are drawing the vector from car to origin because you have to keep the vector addition "tip to tail". Basically, however many vectors you add, you have to make sure they are tip to tail to keep the equations valid?
@ChristopherLum
@ChristopherLum 2 ай бұрын
Yep, that is a good way to think of it. Maybe add this to your notation cheat-sheet :)
@kennethwilsey6260
@kennethwilsey6260 2 жыл бұрын
AA516 are there other applications of this transformation where we could expand to dimensions greater than 3D?
@ChristopherLum
@ChristopherLum 2 жыл бұрын
For most of our applications 3D is probably the largest dimension we can deal with.
@ryancusack5909
@ryancusack5909 6 ай бұрын
Are you sure you wrote that rotation matrix wrong at 16:38? It looks like the same type of rotation in your video Expressing Vectors in Different Frames Using Rotation Matrices at 19:16 in that video. It is a right hand angle, rotating from the reference frame to a 1 frame. If the angle is positive, it looks like what you wrote is correct, or matches the other video.
@chayweaver.2995
@chayweaver.2995 2 ай бұрын
[AE 512]: @16:41 you say that C_1/b is a 2x2 rotation matrix with the angle being psi, where psi is defined as FROM the 1 axes TO the b axes. However, we want to rotate FROM the b axes TO the 1 axes, so isn't the rotation matrix as you've defined it here incorrect? Shouldn't C_1/b actually be the transpose of what's on the board?
@user-hw8di2xk7f
@user-hw8di2xk7f 4 ай бұрын
AA 516: Celeste Yuan
@Colin_Baxter_UW
@Colin_Baxter_UW 5 ай бұрын
AA 516, no questions
@RNeilen216
@RNeilen216 2 жыл бұрын
???? why are there so many letters, and what does this have to do with physical posistions
@edengebretsadik7793
@edengebretsadik7793 2 ай бұрын
AE512
@user-ng5eu1pt5w
@user-ng5eu1pt5w 5 ай бұрын
AA516
@h3ngy1ng68
@h3ngy1ng68 7 ай бұрын
"I dont wanna get the s i g n of s i n wrong" hahaha (thr goes my 1st ever comment in yt) 33:59
Trimming a Model of a Dynamic System Using Numerical Optimization
1:00:44
Euler Angles and the Euler Rotation Sequence
1:10:52
Christopher Lum
Рет қаралды 47 М.
Жайдарман | Туған күн 2024 | Алматы
2:22:55
Jaidarman OFFICIAL / JCI
Рет қаралды 1,4 МЛН
ROCK PAPER SCISSOR! (55 MLN SUBS!) feat @PANDAGIRLOFFICIAL #shorts
00:31
Just try to use a cool gadget 😍
00:33
123 GO! SHORTS
Рет қаралды 85 МЛН
Manipulating Aerodynamic Coefficients
25:33
Christopher Lum
Рет қаралды 7 М.
Homogenous Transformations in Robotics | Robotics 101
7:00
Engineering Simplified
Рет қаралды 4,9 М.
Matrix Transformations and the Homogeneous Coordinate System
26:02
Muffin Express Games
Рет қаралды 67 М.
6 Levels of Thinking Every Student MUST Master
17:12
Justin Sung
Рет қаралды 981 М.
Geodetic Coordinates: Computing Latitude and Longitude
31:29
Christopher Lum
Рет қаралды 18 М.
Жайдарман | Туған күн 2024 | Алматы
2:22:55
Jaidarman OFFICIAL / JCI
Рет қаралды 1,4 МЛН