save the applied schedule into the state during heuristic canonicalisation

This is the data we need to later reconstruct the chosen scalarset permutation.

Github: related to #194 "clearer CEX output with scalarsets"