Skip to main content
. 2021 Jan 21;9:20208–20222. doi: 10.1109/ACCESS.2021.3053351

TABLE 5. Comparison of Algorithm Performance.

Route Route 1 Route 2 Route 3
Training metrics Iterations Solving time/s Iterations Solving time/s Iterations Solving time/s
Traditional Q-learning 153 187 246 266 162 196
Improved Q-learning 95 109 192 212 106 131
Optimization rate 37.9% 41.7% 22% 20.3% 34.6% 33.2%