Example

In this example, we’ll apply a soiling loss of (2 ± 1)% or, to be more precise, a soiling loss with a Gaussian distribution centred at 2% and a relative standard deviation of 0.25, as shown in Figure 4.1(a).¹

We would then define a reference value of v₀ = 2%, and the PDF shown in Figure 4.1(b), which is a Gaussian centred on x = 1 with a standard deviation of 0.25.

At the beginning of each simulation, SunSolve will then stochastically determine x from the PDF, multiply it by v₀, and give a different v for every simulation. With enough simulations, the variation in v will match the desired distribution shown in Figure 4.1(a).

Figure 4.1: Example showing (a) the desired distribution in an input v. This distribution is achieved by inserting a reference value of v₀ = 2% and (b) the PDF for the multiplier x.

We’ll describe the syntax later, but for this example one would apply the soiling distribution by including soiling_front=0.02 within build_optical_settings of Step 3, and create_distribution(DistributionInput.SoilingFront, simToSim=["Gaussian", 1, 0.25]) within Step 6.

Footnotes

1. This is equivalent to an absolute standard deviation of 0.5%. ↩