Open Gate 2011

Notables:

• I'm no David (David 2.0?)
• Only three blisters (putting up a trellis for Eli's grapes)
• We need more green and yellow

Now thinking about modeling the backing up of a baler attached to a wagon - 1989 was the first time someone solved the problem of backing up a semi into a dock.   I think if I was real clever I could solve this problem using fuzzy logic and modeling the response with neural networks.  For the training set I could simulate the problem using a remote controlled tractor.

This has practical applications - there is a 12 year old out there who could back up a trailer better than me at 21 when he was 10.  Maybe a through understanding of the system would allow me to beat Eli (I doubt it though - that kid is good.)

I stole the following figures from Jayarathna et. al., Fuzzy Logic and Neural Network Control Systems for Backing up a Truck and Trailer.  The convention they establish is that theta describes the relative angle of the steering wheels relative to the direction of the cab, while phi describes the angle from the y-axis.

State Variables of Truck Docking

Change in state variables

I think this is the configuration that I am interested in; I'm want to write out the state equations after I think about it a little more - it just seems to be translations / rotations of matrices, similar to robotic arms where the joint between the arms isn't set by a motor.

General Idea - the front wheel's of the tractor don't have to be parallel to the direction of travel, the same goes for the hay wagon.  The baler's wheel's are fixed to be parallel to the direction of travel.  Tongues between them can be any angle.

After I get a working model it would be interesting to see the effects of adding the differential lock of the tractor - but I expect this would just make the turns tighter.

As an aside, one of my biggest fears is that when I leave graduate school I won't be able to pursue projects like this.  By then my focus might change.