first take a look at the `journalctl -r` output (the complete system journal in reversed order .
If that does not go back far enough (only present since boot) - you need to enable persistent journalling
i.e. you need to run:
if you don't have a persistent journal you can also check /var/log/syslog* for hints to what causes the reboots (however somtimes the journal contains more information if a host crashes)