Prognosticator.

The Fermostat has been called many things, mostly by me, and well, mostly during childish temper-tantrums and sometimes even while throwing things.  But now, a few more names can be added to it’s cache of monikers: a prognosticator, a seer, maybe even soothsayer… a prophet? OK, maybe a little too far.  I know.. you’re thinking “Wait, could it be? Can it see into the future??“.  And your answer: a resounding “Well, kinda.”

First, a little background on PID Controllers. (buzzkill)  PID controllers create an output based on feedback from an external source (in our case, the temperature of the system).  However, the feedback being used to help control the output is based on 3 factors: a Porportional factor (the current temperature value), an Integral factor (how much the temperature has changed over a period of time in the past), and a Derivitive factor (the current rate of change of the temperature).  These are all basically engineering terms that give the controller a better picture of what it’s output is doing to the system it is attempting to control.

Armed with this information, the controller is now able to predict that the temperature of the system is going to, for instance, drop 0.1 degree every 10 seconds for 50 seconds after the controller shuts itself off.  This allows it to shut itself off before the system has actually reached it’s desired temperature. In this particular example, the controller would shut itself off at 0.5 degrees before it hits it’s target temperature.  This is just an example, but the point is, the number of temperature overshoots, undershoots and oscillations are minimized with the new algorithm that is running the Fermostat.  This allows the Fermostat to keep the temperature of the system closer to the desired setpoint.  Which is much better than the simple set point controls of any other generic off-the-shelf controller that is currently out there.

Soooo… we got that going for us.