Thus far the only method I've managed to get working in any sense is a simple forward-shooting setup largely similar to that used in the B-Plane targeting tutorial for Mars, for which I determined the launch angle by manual iteration (also the altitude of the Jupiter pass, as using the value from the database in combination with the imperfect Jupiter arrival led to a Jupiter departure trajectory that fell far short of Uranus). This script (available here), produces the trajectory shown below.
While this is superficially pretty similar to that from the database it should be noted that the magnitude of the burns is substantially higher (most notably the Jupiter TCM), leading to a delta-V requirement more than 3km/s higher than indicated by the database before considering Uranus orbit stabilisation (which would be markedly higher than indicated for a desirable science orbit, though that is not my current issue).
Is there any way I can improve upon this script to get closer to the target trajectory (I have tried to set a desired Jupiter arrival date, but that did not work well)? Or is a different method called for? I have attempted to follow through the multiple-shooting tutorial and, while I believe I understand the principle, I haven't yet been able to get a script using that method to function - if anyone is willing to share a script using that method which works in R2014a I would appreciated the chance to take a look at it.
My thanks in advance for any help.