Rudolf Kálmán Has Died • Hacker News Board

Re: Rudolf Kálmán Has Died

#31

Post by tnecniv » Wed, Jul 06, 2016, 3:35 PM UTC

mnw21cam wrote: post #19

I'm always surprised when people talk about Kalman filters without mentioning the killer app, which is weather prediction. Quite a few weather prediction organisations are at least experimenting with Kalman filters, and some are running whole ensemble forecasts using the method. It may sound strange, but the Kalman filter can be a less CPU-intensive (or at least more parallelisable) way of calculating atmospheric sta…

That's an interesting application that I didn't consider, but I'd say the real killer app was the fact that it helped put a man on the moon...

Re: Rudolf Kálmán Has Died

#32

Post by xchip » Wed, Jul 06, 2016, 3:36 PM UTC

joezydeco wrote: post #6

Just last week I needed to smooth out a display reading on an oven controller. The RTD was being read way too fast so I'd get a lot of flicker between values due to ADC resolution. In the back of my head I remembered one word: Kalman. This line of code fixed it right up: static float display_temp = 0; display_temp += 0.04 * (adc_temp - display_temp);

LOL if there's ever a zoolander movie for signal processing I want that line in.

Re: Rudolf Kálmán Has Died

#33

Post by tgb » Wed, Jul 06, 2016, 3:46 PM UTC

xchip wrote: post #29

bendykstra wrote: post #18

Edit: This is wrong. This is a particle filter, another type of Bayesian filter. I can't delete now, so please downvote to hide. I made a Kalman Filter visualization[1] last year to learn more about them. It's amazing to see how good a result you can get from very poor sensor data. In the visualization, a lawnmower (green dot) is tracked (blue circle) using triangulation. The distance sensors have very low accuracy (…

That is not a Kalman filter :) A Kalman filter uses matrices, a physical model of the system its trying to filter data and lots of stuff I haven't found in that code

It looks like a particle filter to me. https://en.wikipedia.org/wiki/Particle_filter

Re: Rudolf Kálmán Has Died

#34

Post by texan » Wed, Jul 06, 2016, 3:49 PM UTC

It's interesting that there is close to none of any personal background or information available on the interwebs.

Re: Rudolf Kálmán Has Died

#35

Post by joezydeco » Wed, Jul 06, 2016, 3:56 PM UTC

xchip wrote: post #32

joezydeco wrote: post #6

Just last week I needed to smooth out a display reading on an oven controller. The RTD was being read way too fast so I'd get a lot of flicker between values due to ADC resolution. In the back of my head I remembered one word: Kalman. This line of code fixed it right up: static float display_temp = 0; display_temp += 0.04 * (adc_temp - display_temp);

LOL if there's ever a zoolander movie for signal processing I want that line in.

Alrighty. I'll stick to what I know best. Thanks for the reality check.

Re: Rudolf Kálmán Has Died

#36

Post by bendykstra » Wed, Jul 06, 2016, 4:00 PM UTC

xchip wrote: post #29

bendykstra wrote: post #18

Edit: This is wrong. This is a particle filter, another type of Bayesian filter. I can't delete now, so please downvote to hide. I made a Kalman Filter visualization[1] last year to learn more about them. It's amazing to see how good a result you can get from very poor sensor data. In the visualization, a lawnmower (green dot) is tracked (blue circle) using triangulation. The distance sensors have very low accuracy (…

That is not a Kalman filter :) A Kalman filter uses matrices, a physical model of the system its trying to filter data and lots of stuff I haven't found in that code

Oh, you are right. That is actually a particle filter. I'm sorry, I studied both on the same day and misremembered which one I had implemented.

Re: Rudolf Kálmán Has Died

#37

Post by ramisama » Wed, Jul 06, 2016, 4:05 PM UTC

I remember using the Kalman Filter in my Advanced Macroeconomics class to calculate economic cycles out of the GDP data of my country!

Re: Rudolf Kálmán Has Died

#38

Post by reachtarunhere » Wed, Jul 06, 2016, 4:08 PM UTC

RogerL wrote: post #21

If you are interested in learning about them in depth, I'll toot my own horn and point you to my interactive book on them: https://github.com/rlabbe/Kalman-and-Bayesian-Filters-in-Pyt... It uses Jupyter Notebooks to run code in the browser. Check out the book, or run online using Binder.

I highly recommend this.

Re: Rudolf Kálmán Has Died

#39

Post by ArkyBeagle » Wed, Jul 06, 2016, 4:31 PM UTC

joezydeco wrote: post #6

Just last week I needed to smooth out a display reading on an oven controller. The RTD was being read way too fast so I'd get a lot of flicker between values due to ADC resolution. In the back of my head I remembered one word: Kalman. This line of code fixed it right up: static float display_temp = 0; display_temp += 0.04 * (adc_temp - display_temp);

I've heard that called an exponential moving average:

https://en.wikipedia.org/wiki/Moving_average#Exponential_mov...

Re: Rudolf Kálmán Has Died

#40

Post by zbuf » Wed, Jul 06, 2016, 4:40 PM UTC

mnw21cam wrote: post #19

I'm always surprised when people talk about Kalman filters without mentioning the killer app, which is weather prediction. Quite a few weather prediction organisations are at least experimenting with Kalman filters, and some are running whole ensemble forecasts using the method. It may sound strange, but the Kalman filter can be a less CPU-intensive (or at least more parallelisable) way of calculating atmospheric sta…

The reason the Kalman filter is so significant is that there isn't a standout "killer app"; there's a whole bunch of them.

It's much more fundamental than any of its applications, which are all around us.