Costa Concordia: Control System Synthesis using a submarine as a model

In my previous blog about Software Engineering/Architecture/Vaguely Organized

https://blogs.msdn.com/b/socal-sam/archive/2012/01/23/costa-concordia-mathematical-model-of-the-submarine.aspx

In software architecture, at sometime you will have to deal with either an electrical engineering type or mechanical engineering type that has generated the control system synthesis for your system.  This person or person’s will talk about singularities (and not the black hole kinds from Star Trek, although….), linear time invariant plant and so forth.  This means that they have generated a model that will simulate the craft, keep them in billing hours, and allow you as the software architect to have a way to make sure your software doesn’t make an expensive mistake.

 

Here what the submarine’s control system looks like:

image

If you are a hiring manager and the control synthesis is complete, you might want to consider hiring this person for the project.

Examples of an expensive control system mistake

Tacoma Narrows Bridge, an example of what control system synthesis attempts to avoid for large systems.  In this case the vibration caused by the wind caused the “model” or bridge to vibrate at resonance.  If you design a system that uses software to run it, then you need to make sure that input from your software won’t cause a resonant disaster to occur.

B-2 Crash, this was caused by water in a sensor.  In this case the control law model could have been used to detect what happens if there is water in the Angle of Attack sensor (I am guessing, but the aircraft does a sharp pitch up just after take off, there appears to be a pilot input, nose down, nose up, wing over and punch out by the pilots).  The failure to examine what happens if the sensor, let’s say it was the AOA or Air Speed Pitot Tub then the ability for the craft to recover could have been simulated.

There are two B-2s in the video, the first one appears to take off successfully, at 2:00 minutes the second B-2 takes off with a sharp pitch up, a mist is emitted around the engine ducts and so forth.

Where do we go from here?

How do you use C# or similar to do control system synthesis?  Let’s see how we might approach that in next week’s blog.