Reinforcement Learning for Computational Guidance of Launch Vehicle Upper Stage

Table 3

Results of RL-guidance and guidance-based optimal control.

Parameter	RL-guidance	Guidance-based optimal control
Final time (s)	299.94	301.15
Final position error (m)	(2.767, 15.614, 128.149)	(47.263, -0.145, -16.306)
Final velocity error (m/s)	(-0.642, -0.259, -0.107)	(0.543, -0.007, -0.101)