sim.ini

来自「这个代码是policy iteration算法关于强化学习的. 请您用winzi」· INI 代码 · 共 12 行

2                      dimension
0                      deterministic (set to zero for stachastic action selection)
50                     straight_line_steps
20000                  max_steps_to_goal
1000                   max_num_episodes
0.1                    noise_mag
-0.85 -0.85            start state x1 x2 ...
0.85  0.85             goal  state x1 x2 ...
2                      num_of_gaussians  // must be at least 2
0.85 2.0 0.85 2.0      // the first guassian x1 std_x1 x2 std_x2 ...
-0.2 0.8 0.2 0.8       // the second guassian x1 std_x1 x2 std_x2 ...