I would try to blend the caster beans and then use a Partition method with pure MDI (Methylene diphenyl diisocyanate) and perhaps ethanol or methanol. There would still be substantial  sample cleanup and the purity of the polyols would also be questionable

Castor oil has ricinoleic acid units as part of the fatty acid triglyceride.  So if you extract castor oil, you have your polyol without further chemistry.  For examples of how to extract castor oil - this is a commercially relevant question, so the patent literature is usually a good first stop.  The link below is one example (found via Google Patents and keywords "castor oil extraction").  It appears that solvent extraction has been the practice, although who knows what the best practice is now....  Also, castor beans have ricin (a toxin) in it, so be warned, this is not a recommendation from me re: how to do the oil extraction.