Slashdot Mirror

← Back to Stories (view on slashdot.org)

Linux Controls a Gasoline Engine With Machine Learning

Posted by Soulskill on from the penguin-engine dept.

An anonymous reader writes: Here's a short (2 min) video of PREEMPT_RT Linux controlling a gasoline engine from one burn to the next using a Raspberry Pi. It's using an adaptive machine learning algorithm that can predict near chaotic combustion in real-time. A paper about the algorithm is available at the arXiv.

1 of 89 comments (clear)

  1. Re:I didn't see any mention of efficiency. by Anonymous Coward · · Score: 5, Informative

    In HCCI, you operate the engine *without* the throttle (for practical purpose, they probably left the throttle in but operated at WOT: pedal to the metal). This improves the efficiency for sure right there.

    HCCI *is* about efficiency: without the throttle in the way, things go much better thermally in the machine. Torque output (thus power) is then controlled by fuel injection and the motor autoignites (like a diesel).

    Combustion at every cycle starts according to current conditions (pressure and temperature). This means that earlier cycles are essentially what determines the current combustion... You only get to control fuel timming and quantity supply...

    This is hard people...

    The article is about a *control* strategy. It is normal to seek worst case conditions (hence the chaos, variability, etc) and to try to demonstrate robustness... Efficiency will certainly be evaluated next, once robustness is set. Given the article was written in october 2013, they probably have this figured-out by now ;-)