There is one option called as Auto Failback in F5 which is nothing but preemption. With this setting, you can set a primary member of the cluster to always be the Active member. Also you can define the timeout value that you want the F5 to wait before failing back to specified device. Below article will give you details on how to configure it.
Only concern with such configurations is, it will create problem in case there is issues with active member that is set to auto-failback and its having some critical failures and gets reboots automatically. In this scenario, the cluster will failover to secondary and then immediately failback to primary when it completes reboot. I have seen such issues so highlighted same.