1. query v$log and see group which has INACTIVE status.
the group which has a status of CURRENT should be the last to be recreated(group 2 for example).
select * from v$log;
2. delete the logfile group which has an INACTIVE status.
Assumed that group 3 has INACTIVE status.
alter database drop logfile group 3;
3. recreate again the logfile group 3 with bigger size.
This is for mirrored redologs(common)
alter database add logfile group 3 ('/u06/oradata/PRD/log03a.dbf','/u04/oradata/PRD/log03b.dbf') size 50M reuse;
4. repeat step 2 and 3 for group 1.
5. force the database to switch to new online redo logs. To force group 2 to become INACTIVE.
alter system switch logfile.
6. repeat step 2 and 3.

You can add more group if you wanted to.
Do this when no user is using your db. Or open your db in restrict mode.

Hope this help.