As a research tool, just about any analysis one can imagine is possible with CHARMM, either directly, or by exporting the needed data and writing a program.
If the membrane is set up with the Z axis as the bilayer normal, that's a good first step; vector projections on the Z axis can be used to measure rotation. For any vector, if it is normalized to a unit vector, the z component is the cosine of the angle it makes with the Z axis.
Choosing the right vectors for the peptide can pose a challenge, however, and the least-squares plane normal may only be appropriate for a flat peptide, such as a beta hairpin config. For a helix, that is probably not a good choice. You must first establish how much the peptide conformation changes during the simulation in order to pick the right vector(s).
There are a number of ways to define vectors within CORREL, and a couple more COOR commands, such as COOR LSQP and COOR HELIX