How to perform an optimisation

In optimisation mode, the user selects certain inputs to be optimised and an optimisation goal, where the goal includes an ‘output metric’ and whether that metric should be maximised or minimised. SunSolve then determines the values of the selected inputs that meet the optimisation goal. This requires SunSolve to conduct multiple runs under the same simulation.

To perform an optimisation

Switch to optimisation mode: Select optimisation on the Solver tab.

Activate wand: Click the wand icon located on the toolbar. When the wand is highlighted, inputs can be added or removed.

Select inputs to optimise: Eligible inputs are highlighted in yellow. Click eligible inputs to add or remove them from the optimisation input list. Up to three inputs can be selected. Unlike sweep mode, materials are not permitted as inputs.

Define optimisation input parameters: In the input section of the Solver tab, set the upper and lower bounds, the best guess, and the resolution of each optimisation input. The solver will determine the optimal values for these inputs to within the user-defined resolution.

Define optimisation goal: In the goal section of the Solver tab, select the output metric and whether the goal is to maximise or minimise that output.

The output metric can be

• the optical absorption in one or more layers (integrated over the incident spectrum),
• the cell or module electrical current, or
• the cell or module electrical power.

When the output metric is optical absorption, it is necessary to select the relevant optical layers. The output metric is the sum of the equivalent photon current absorbed in the selected layers. If a layer includes free-carrier absorption (FCA), select whether the photon current should represent the layer’s total absorption current (includes FCA) or its generation current (excludes FCA).

Set optimisation options: In the options section of the Solver tab, set limits on the optimisation routine and set inputs for the genetic algorithm. These are described below.

Run optimisation solver: Click play ▶. The solver applies a genetic algorithm to find the optimal input values that meet the optimisation goal (within the constraints of the input bounds and the optimisation options).

Download results: Click ‘Download optimisation results’ on the Solver tab to download the inputs and output metric of every run in the simulation. Click the download icon to download the various output files for the optimised run.