Special Topics - The Kalman Filter (7 of 55) The Multi-Dimension Model 1

[@cjk9988]

Sir you are a legend...it is rare to find lecturers like this especially at university level...thank you so much... u r a life saver!

[@MichelvanBiezen]

Thank you. Glad you found our videos. 🙂

[@ma888u]

Dear Sir, your work is amazing! Please continue like that because you help a lot of people to understand very complex stuff in a way that it seems very simple and logical!!! You have a great gift to teach!!!

[@briankasmara9937]

This video makes me appreciate the step-by-step progression of explanations from previous videos. From simple 1-dimensional examples to more complicated multi-dimension and general Kalman Filter problems. I wish all lecturers taught this way. Thank you Michel!

[@galloden]

This video series is incredible in clarifying The Kalman Filter process. Thank you very much.

[@Aeon_Scipher]

Dear Michel, I have learned more from you, than pretty much all of my professors combined. Thank you for all that you do. I swear if I strike it rich someday, I will be donating a large sum to ilectureonline. Thank you sir!

[@MichelvanBiezen]

We are glad that you found our videos and you find them useful. Thank you for your comment. 🙂

[@Randomsimus]

9 Years and this is still the best series explaining what's kalman filter in terms that you can understand

[@sarastanway8861]

I can't thank you enough for this wonderful series! I learn much better by self-teaching than sitting through lectures and this content and the way you break down your examples is excellent!

[@WA-dq3me]

You're a godsend. You break the Kalman filter down and explain it in almost fundamental terms!

[@khaledwaleed9128]

Incredible amazing skill of explanation of the subject, reallllyyyyyyyyyy helpful and more clear than very known channels, hope this reaches more for what it deserves

[@AlexisPaques]

I am finally capable of understanding the Kalman Filter (1D) and soon able to implement it on my process ! Thanks a lot !

[@mkelly66]

This and your previous videos are the best explanation of the Kalman filter I've ever seen (and I've seen quite a few)! Well done!

[@adityasrivastava6288]

Sir, You are a gift to humanity. Thanks a lot for breaking down the complex stuff into easy explanation.

[@MichelvanBiezen]

Thank you for your kind words.

[@leec8977]

One of the best explanation I've ever seen about this topic. Thanks!!

[@MichelvanBiezen]

Thank you. Glad you found our videos.

[@yubrshen]

Thank you the most detailed, and the most approachable lectures on Kalman filter. Thank you very much!

[@zedraken369]

Dear Sir, I started to view that video serie on the Kalman filter to refresh my student souvenir (a quite long time ago indeed), and your explanations are very clear! Things are smoothly introduced to increase the complexity but you are able to show them as very simple elements with your very educative approach. My full congratulations for that course (I just finished video #7 and I am really excited to move on the next ones). Thanks a lot!

[@MichelvanBiezen]

We really appreciate your comment.

[@mahimagandhi]

You are an amazing teacher. I wish all teachers were like you. You have made this topic so simple. Thank you so much :)

[@simdenv]

Fantastic videos. Lots of explanation, first principles and details, perfect for breaking down some rather daunting closed loop control theory.

[@amksprofa3110]

I'm impressed, how could you make it so easy to understand. very helpful

[@aymakam3789]

I really do not know how to thank you for this astonishing series.

[@pawewojda2776]

Thank you sir. This is the most valuable channel on YT I know.

[@kylebroflovski5333]

You are wonderful! you just managed to explain a concept ive been struggling with for my final year masters modules. Thank you!

[@MichelvanBiezen]

Happy to help!

[@WangZhen-r2d]

Dear sir, you published this series of videos years ago, but it still helps me understanding Kalman Filter years later. thanks a lot for your amazing work!

[@MichelvanBiezen]

Glad to hear that

[@lengooi6125]

Very well done for illuminating a otherwise confusing subject matter. Many thanks for posting these videos. You make learning difficult topics a joy to learn !!.

[@wanderfra42]

Many have already said that, but I want to emphasize that this is one of the best explanation ever done of the Kalman filter. All principles of teaching are perfectly applied: step by step, intuition first, visual clues, repetition of concepts and a nice smile. Thanks a lot!! (just a question: in the "Current Becomes Previous" block, the two variables should be Pk and Xk (not Xp), right?)

[@MichelvanBiezen]

Thank you for your comment. You are correct it should be Xk (not Xp) (Xk is defined in the next video). Good catch. Thank you.

[@morakinyostephenz7003]

You sure deserve a Nobel Prize for this wonderful explanation.

[@chendonggua6506]

Dear sir, thanks for your work, making the complex thing so easy to understand. really feel educated in your course!!!!

[@tongwu8467]

You saved me from my undergraduate physics class. Now you are saving me from my graduate times series class again.

[@eldoprano]

I'm really happy with the previous videos, this one should have taken in consideration that the viewer saw those previous ones, so that you can explain like "This is variable was this in the previous equations, and now it is...". It feels like jumping into another topic, and my brain is desperately trying to make connections :P Still! Im excited to watch the next videos, I love how you explain it

[@bryanbocao4906]

Thanks for the video! 3:32 should a 3 demensional position and velocity matrix be 6x2? It would be great to clarify that.

[@lilmoesk899]

Thank you very much for taking the time and effort to break this down for non-engineering people like me. I can't say I understand it all, but this explanation definitely makes things seem manageable. Thanks!

[@MichelvanBiezen]

Glad it helped!

[@vuthee529]

Dear Sir! Your explanation is really amazing! I really appreciate.

[@MichelvanBiezen]

Thank you. Glad you found our Kalman Filter videos. 🙂

[@olaoluwapopoola8023]

You are a great teacher. Thanks for this

[@lingyaozhang6104]

Your video is really wonderful! I am new to Kalman filter and your explanation just makes me more interested into it!

[@awais_arshad]

You are an amazing teacher Michel. Thumbs up and lots of respect from Pakistan. Bravo

[@MichelvanBiezen]

Thank you and welcome to the channel!

[@joaosimoes5052]

Amazing video series so far! Great lessons from a great teacher

[@MichelvanBiezen]

Glad you like them!

[@rubenguerrerorivera7462]

So neat! The art of making complicated things, easy!

[@MichelvanBiezen]

Thank you! 😊

[@martinroa]

Thank you very much for the lectures Michel. I spotted an error that I think nobody has referenced here in the chat yet. In the "UPDATE WITH NEW MEASUREMENT AND KALMAN GAIN" box, when calculating "K", the "H" in the numerator should be transposed.

[@mickjoris4039]

I think you're right.

[@jiemin3065]

Thanks for this great series! It is really crystal-clear flow chart and really helps a lot in understanding. One place I think is a typo: In the "Update with New Measurement and Kalman Gain" box, the nominator in the Kalman Gain is Pkp*H, while I thought it might be Pkp * H.T, transpose of H.

[@gergelytakacs]

Yeah, there should be a transpose of H.

[@mickjoris4039]

This overview helped me a lot, thanks!

[@MichelvanBiezen]

Glad it helped! 🙂

[@sakuranooka]

3:30 Why in 3 dimensions X is a 6 x 6 matrix? We have 3 positional coordinates, and 3 velocity components, which gives a 6 x 1 vector, no?

[@nikocheng2437]

Although there are some notations different from what I have learned in the course, it still very helpful for me to understand. Thank you

[@hayemaker7331]

Thanks, you have an amazing skill to explain difficult things in a such easy way

[@MichelvanBiezen]

Glad it was helpful!

[@lard_dork]

You are a miracle worker..Thanks a ton.

[@UserUser-pv2wo]

Hello! I tried several manuals and papers before on this topic, but I'm kind of very stupid students, so did no significant progress :) But with this videos, so far, things don't look so complicated any more and I have really started making a clue of how KF operates! Cross my fingers, and go further. Thank you for outstanding explanation!

[@MichelvanBiezen]

Don't think you are "stupid" because you can't understand the literature on the Kalman Filter. Most of it is very difficult to understand, not because it is a hard topic, but because it is not written in a way for someone who doesn't already know it to understand. (That is why we started the playlist)

[@1098tony]

What a great teacher!!

[@priyankarao7966]

Sir, these videos are very good. Thank you so much

[@manishbhanu07503]

thanks a lot for such an easy way explanation of this topic

[@rameshkumartamilarasivelus470]

you are simply amazing. I have been trying to learn this for a long time but couldn't get a proper understanding about what and when it is happening. thank you so much.

[@gergelytakacs]

Excellent lectures series, thank you. As others have noted: a.) A transpose is missing in the Kalman gain b.) Mixing the H/C notation for the measurement matrix is confusing.

[@MichelvanBiezen]

Thanks for that! We appreciate the input.

[@francisbaffour-awuahjunior3099]

Is the H numerator of the Kalman filter equation supposed to be H transpose?

[@hoerbschmidt618]

Great explanations so far, thank you very much!

[@MichelvanBiezen]

Glad you find it helpful. 🙂

[@hayfahvytsen]

Awesome explanation. Great stuff!

[@muhammadshariqabbas7665]

salute to you sir May God bless you...awsome explaination

[@MichelvanBiezen]

He already has. Thanks.

[@grasuh]

There is a typo. The "measurement input" expression should use H instead of C, like this Yk = H * X_km + Zk This is a great video series. Keep it up!!

[@klam77]

LOL: that's why no response....too late to change the whiteboard! (human nature....LOL). Good catch

[@joaosimoes5052]

thanks!

[@amksprofa3110]

God bless you for your efforts

[@abdelazizchakouri5108]

Sir , You are the best .Thank you very much

[@MichelvanBiezen]

Glad you liked the videos! 🙂

[@aaron.protein]

Thank you for these lectures!

[@justearthur8666]

Hello first of all a big thank you for those vids ! Just can you explain what the matrix H represents please

[@MichelvanBiezen]

H is a conversion matrix in order to make the sizes the same so you can multiply the matrices.

[@stevegilbert3067]

It seems like the C matrix used in the lower right hand box should be H instead.

[@subhamdey8954]

Is steves answer correct?

[@SomasundharamMuthumanickam]

Yes, Steve's answer is correct. C = H

[@ACTUATOR24]

Michael, this is the best explanation I have found, thankyou. I'm sure you'll explain later but from this video, I'm not sure what the purpose of matrix H is? It is not listed on the slide and so far you have not mentioned it in the presentation, can you help?

[@johnnywerner]

very nice content and best explanation ever. Thank you. I just have to say your H matrix in the numerator of K expression should be transposed.

[@fethibencherki3708]

awesome lecture series !

[@ThePgp03]

Sir, very nice lectures. Kudos...

[@jyotiranjansahoo8833]

Dear Sir, could you please tell, what the symbol "H" stands for.

[@redd8551]

It is the C from the Measutement

[@mohanrajg4168]

See at 5.30 minute in the video kzbin.info/www/bejne/fKbcaGaka8SVeMU

[@mohanrajg4168]

@@redd8551 Its just for a reshape operation (refer kzbin.info/www/bejne/fKbcaGaka8SVeMU)

[@visavou]

top! got me 10 points in the exam! thank you very much

[@Speedfinders]

In 3:30, could it be that you mean 1x6 Matrix instead of 6x6?

[@RaviKant-kp9fs]

great lecture Sir,Thank you for your lectures series.

[@NaveedAhmad-nx8yq]

Very well explained ....

[@oleholgerson3416]

Great videos, very good job!

[@sohamprajapati8784]

Awesome Lecture!

[@MichelvanBiezen]

Thank you!

[@ColinBroderickMaths]

Do Q, R, and w ever change? This is the one point that I'm stuck on. Are they positive constants? Are they random samples from a noise distribution function? Are they the stdev of the noise, or some other measure? Thanks

[@tobiascang1717]

wow this is another level, suddenly

[@LuizBitencourt]

Great series of videos!! I think the H matrix on numerator of the Kalman Gain calculation is transpose. Ok?

[@منوعاتثقافيةتعليمية]

Hello Thank you for these amazing lectures, I have a question What are the differences between the Kalman filter and Model predictive control??

[@MichelvanBiezen]

The Kalman filter takes in a measured value and calculates a predicted value and during each iteration determines how much weight to put on each of the 2, combines the weighted average and continues to refine the prediction.

[@منوعاتثقافيةتعليمية]

@@MichelvanBiezen Thank you ...and MPC what the difference???

[@MichelvanBiezen]

MPC is used more for process control, where you vary the input parameters and determine the result of those changes and then you keep varying the inputs until the outputs are exactly (or close) what you want. Here you typically know what the output should be and you vary the input parameters until you get the desired outcome.

[@socrates8562]

Thank you Michel, you are amazing!

[@飛鴻-q1c]

Excellent explanation！

[@MichelvanBiezen]

Thank you. Glad you found our videos.

[@imeldaduma2275]

Hello Mr. Biezen. thank you for your help. Now, I understand about Kalman Filter clearly. I need your published paper about multi dimension model of kalman filter. So I can citation that. Thank you Mr.

[@DXOan]

thật là tuyệt vời

[@MichelvanBiezen]

Glad you found it helpful.

[@TheAlistrawberry]

Thank you so much , you made these stuffs clear and great !

[@pengboli1253]

great presentation

[@agathaniwomugizi2433]

Thank you sir. Very very helpful

[@marius10ster]

Thank you for the video. Truly appreciate it.

[@AdrianKurono02]

Why does he say in the video @3:30 "6x6 matrix"? In that example his considering only velocity and position in all 3 axis. Am I losing something? Should not be a 2x3 matrix? It confuses me.

[@MichelvanBiezen]

+Adrian Tejada If you watch some of the examples in the following videos, it will become clear. (In 3-D, you will have position in the 3 directions and velocity in 3 dimensions, thus you'll have a 6 x 6 matrix)

[@AdrianKurono02]

+Michel van Biezen Sure. I'll be patient and check. Thanks.

[@evanrfraser]

EXCELLENT! Thank you.

[@Yasser2652]

Thank you very much!! I am just wondering how would you calculate the process noise co-variance matrix, if we assumed it is not given Thank you again

[@MichelvanBiezen]

I am planning on making a series of a real multi-dimensional example so you can see how to put all the matrices together. That will be during the summer.

[@bryanbocao4906]

Should H in the PH in the numerater in the Kalman Gain equation be its transpose -- PH^{T}?

[@peter9910]

Great explanation. Thanks!

[@karangusani5919]

Simply Awesome :)

[@Wombatan]

I wonder how is the professor going to divide one matrix by another? (when calculating the Kalman Gain) Does he mean multiplying by an inverse matrix?

[@mahdizytoon8293]

thanks for your very helpful lectures ... if you can please put a link for pdf lectures that you suggest the interest persons to read and follow on . Best wishes

[@vijayshejal4322]

Amazing ..thanks

[@ShahramTaba]

Thank you sir, you are the man.

[@kris533d]

Man, I love you

[@MichelvanBiezen]

Thank you. Glad you found us.

[@amksprofa3110]

How the KF behave in case we have a maneuvering object. will it detect the maneuver and how. THANKS

[@MichelvanBiezen]

+Amksprof A The Kalman filter will be able to track a maneuvering object and smooth out the errors in the observation data.

[@gubijic]

What if the u_k is also random variable vector?

[@ofirshalvi2615]

Isn't there a typo/confusion between C and H on the white board? (H should be C)

[@SuperKreyszig]

You are awesome.

[@3idet]

Thank you very much.

[@Felipe-hi6nh]

Thank you for the huge help!

[@MichelvanBiezen]

Happy to help!

[@mohamedrameez4470]

How can i get the other lucters from 43 to 55 because i couldn't find it can you help me Plz

[@MichelvanBiezen]

We still have to make the rest. We are hoping to get a chance to complete the series this summer.

[@JohnDemetriou]

Still not there though

[@jeevanraajan3238]

You are god.THanks a ton

[@LuizBitencourt]

The numerator of the Kalman Gain equation is corret? I mean, isn't the H matrix transposed on the numerator?

[@WadoNeil]

I believe you are correct - numerator should be P_{k_{p}} H^{T}. Videos are still great! :)