Evolve better control solutions with genetic algorithms

in compilers by DP | 4 comments

Brian writes:

I wrote an article over at Microcontrollercentral about using Genetic Algorithms to optimize microcontroller algorithms; in this case a PID control loop. It was a new and fun way of thinking about programming, and I included some source code for anyone interested!

Via the contact form.

This entry was posted in compilers and tagged , .

Comments

  1. Matseng says:

    Genetic algorithms is an interesting subject. I’ve been tempted to read up about them a bit and actually try doing something myself, but I’ve never gotten around actually doing it yet….

  2. Drone says:

    Nice post DP. You’re not “dumbing us down” by including nice content like this – thanks…

    Evolutionary Algorithms, (aka EA’s; e.g., Genetic and/or Neural networks are a couple of common examples) are not heuristic. Therefore things can get way out of control (hardware releases the proverbial magic blue smoke) unless limits are employed by one or more methods at the input and/or output (e.g., PID, Kalman, Bayesian to name but a few, of many).

    The strength of EA’s really shines when you have multiple inputs and need to optimize them to a single (or more) outputs depending on the rules with which decisions are made, often in a cooperative manner.

    Parallel sequential logic “processors” and/or multi-combinational logic “engines” are well suited to this type of thing, especially for real-time applications. But keep in-mind, memory requirements and speed with EA’s do NOT scale linearly (that’s putting it mildly).

    Below is a link to a dated but decent paper comparing Genetic vs. Neural networks. The research space is deep like an abyss with respect to this stuff – just the tip of a very large iceberg, regarding this topic – with most of which is well submerged!

    agsm.edu.au/bobm/papers/jena.pdf

    Enjoy…

Leave a Comment

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Notify me of followup comments via e-mail. You can also subscribe without commenting.