Developing a web based PV simulation platform (targeting at machine learning combined with advanced device and process simulation to support process optimization)

Abstract

We demonstrate the current state-of-the-art of a web-based photovoltaic simulation platform, which is developed within the Solar Energy Research Institute of Singapore (SERIS). This platform shall allow us to (1) dynamically upload a variety of measurements from a measurement tool within our cleanroom into a user centric database, (2) dynamically upload and execute internally and externally developed computer programs to simulate such measurements, (3) compare simulation to measurement, using automated fitting, (4) Test/deploy machine learning applications.Using calibrated computer simulations, a large number of simulated measurement data can be generated, as required for the computation extensive training of an artificial neural network. However, the deployment of the neural network (prediction in a forward modelling as well as in a reverse modelling direction) is fast. Thus, using artificial neural networks, predictions and optimizations deploying machine learning will be possible.