Fig 1.
SAR robot movement space.
Fig 2.
Interaction models for reinforcement learning.
Fig 3.
The classical grid map.
Fig 4.
The improved grid map.
Fig 5.
Initialize Q-table with FPA.
Table 1.
Parameter settings of FIQL, CQL, PSO, GWO and DA.
Fig 6.
SAR robot search overall process.
Fig 7.
Iteration times with different learning factor.
Fig 8.
Typical grid map path planning for SAR robot in different environments.
Fig 9.
Mean path length of different algorithms in typical grid map for SAR robot.
Table 2.
Comparison between FIQL, CQL, PSO, GWO and DA in typical grid map.
Fig 10.
The optimal path in the improvement gird map using FIQL algorithm.
Fig 11.
Mean path length of different algorithms in improved grid map for SAR robot.
Table 3.
Comparison between FIQL, CQL, PSO, GWO and DA in improved grid map.
Table 4.
Comparison of performance between FIQL and CQL in typical grid map.
Table 5.
Comparison of performance between FIQL and CQL in improved grid map.