Skip to content
SunSolve

Uncertainty Distributions

Many of the inputs in SunSolve-P90 can be assigned an uncertainty.

Let’s say we have an input value v, and that we are unsure what value v should take. We’d like that uncertainty in v to then flow through to an uncertainty in the energy yield from our PV system.

As described earlier, we can incorporate the input uncertainty into our yield uncertainty by solving the yield multiple times, each with a different value of v. This is the Monte Carlo approach.

In SunSolve-P90, we introduce that variability in v by using a reference value v0 and a multiplier x,

v=v0xv = v_0 x

where x is determined randomly (stochastically) before each simulation.

Thus, the user defines the input to have two components:

  1. a reference value v0, and
  2. a probability density function (PDF) for the multiplier x.

Section Overview

Section titled “Section Overview”

The uncertainty distributions documentation covers:

  • Example: Practical demonstration of applying uncertainty to soiling losses
  • PDF Types: Five supported probability density functions (constant, Gaussian, skewed Gaussian, Weibull, arbitrary)
  • Imposing Limits: How parameter constraints modify PDFs
  • Cumulative Distribution: Construction and use of CDFs
  • Determining Multiplier: Random sampling process for each simulation
  • Syntax: API syntax for defining distributions