Path Selection in Disaster Response Management Based on Q-learning.pdfVIP

P．sueaj．／PathSelectioninDisasterResponseManagementBasedonQ—learning 103 maximum rewardasofllows Step 3．Obtainanext： arg m xa Q(s，a，)． ， ． u (s)：mxa{ s)u (s，))． (6) Step 4．Executeaction anext，and drivetheagentto a r+A』∑-一P(s， states rfom s． s ∈S