Table 5 Parametric analysis.

From: Deep reinforcement learning-driven multi-objective optimization and its applications on lighting infrastructure operation and maintenance strategy

Parameters

Numerical value

Degradation value (D)

13.20

Failure threshold

10.50

Difference (γ)

2.70

Probability of failure

P{γ ≥ 2.7}

0.03

Expected return on failure

-1.29

Expected return on normal

-2.91

Back to article page