TABLE 2.

Description of parameters in the initial reward function. The vessel’s maximum speed is denoted in decameters per second (1 dam = 10 m).

Scaling parameter	Interpretation	Value
γ ϵ	Cross-track error scaling	5.0
γ θ	Sensor angle scaling	10.0
γ x	Obstacle distance scaling	0.1
α r	Zero-reward relative speed	0.05
r coll	Collision reward	-10000
γ r	Constant multiplier	1.0
λ	Objective trade-off coefficient	0.5
Sensor parameter
U max	Vessel’s maximum speed	1.0 dam/s
N	Number of sensors	180
θ i	Angle of sensor i	$-\pi +\frac{2\pi }{N}i$
d	Number of sensor sectors	9
S r	Sensor range	1.5 km
ΔLA	Look-ahead distance	3 km