Parameter Extraction of Solar Cells using Genetic Algorithm

Overview

This project implements a Genetic Algorithm (GA) to extract key parameters of solar cells from their raw current-voltage (V-I) data. The model uses both the Single Diode Equation (SDE) and Double Diode Equation (DDE) to accurately represent solar cell behavior, including parasitic effects. This work provides insights into theoretical modeling and solar cell performance, aiding in the optimization of photovoltaic devices for space and terrestrial applications.

Features

• Models solar cells as unbiased pn diodes under illumination using SDE and DDE.

• Incorporates effects of parasitic resistances: Series Resistance (Rs) and Shunt Resistance (Rsh).

• 

• Employs Genetic Algorithm to optimize parameters:

  • SDE: Isc, Io, Rs, Rsh, η.
  • DDE: Isc, Io1, Io2, Rs, Rsh, η1, η2.

• Efficiently handles non-linear, implicit, transcendental equations.

• Supports analysis of single-junction and multi-junction solar cells.

Motivation

Parameters like Isc (short-circuit current), Io (reverse saturation current), and parasitic resistances significantly influence the performance of solar cells but cannot be measured directly. Traditional analytical methods fail to account for all parameters accurately. Heuristic approaches like GA provide a robust alternative for multi-parameter optimization.

Genetic Algorithm (GA)

Key Features

• Initialization: A random population of chromosomes is created.
• Selection: Fittest chromosomes are selected for reproduction.
• Crossover: Top 30% of the population undergoes crossover to produce the next generation.
• Mutation: Random bit-flipping introduces diversity.
• Termination: Process ends when the error falls below a set tolerance or a maximum number of generations is reached.

Fitness Function

The GA minimizes the error between measured and calculated current values:

Implementation Details

• Operating temperature: 300 K.
• Population size: 1000 chromosomes.
• Number of generations: ~100.
• Parameter bounds: Determined via visual inspection or preliminary direct search methods.

Results

The GA was applied to individual GaInP, (In)GaAs, and Ge solar cells:

V-I Curve Validation

The extracted parameters reproduce V-I characteristics closely matching the measured data:

The GA was applied to triple-junction solar cells stack:

How to Use

Prerequisites

• C++ compiler supporting C++11 or higher (e.g., GCC, Clang).

Running the Code

1. Clone this repository:

   git clone https://github.com/your-repo/solar-cell-ga.git

2. Navigate to the project directory:

   cd solar-cell-ga

3. Compile the source code:

   g++ -o sd_ga sd_ga.cpp

4. Run the executable:

   ./sd_ga

References

1. Kassap, S. Principles of Electronic Materials and Devices.
2. Goldberg, D. E. Genetic Algorithms in Search, Optimization, and Machine Learning.
3. Holland, J. H. Adaptation in Natural and Artificial Systems.

Acknowledgments

• Project by Prachi Bisht under the guidance of Shri Suresh E. P., Solar Panel Division, URSC, ISRO.
• Special thanks to the URSC Solar Panel Division for their support and resources.