UROP Openings

Have a UROP opening you would like to submit?

Please fill out the form.

Submit your UROP opening

A solar cell simulator with Automatic Differentiation


Term:

Department:

QI: MIT Quest for Intelligence

Faculty Supervisor:

Dr. Giuseppe Romano

Faculty email:

romanog@mit.edu

Apply by:

Contact:

romanog@mit.edu

Project Description

Photovoltaics are undoubtedly contributing to a sustainable future. Tremendous advances, from both fundamental and technological aspects, have led to high conversion efficiencies, peaking at 47 % [1]. Research in this field has been conducted both at the material and device level. In particular, it has been shown that just moving from single-junction to tandem solar cells, the efficiency has a dramatic increase. This UROP opportunity falls under a project that aims to develop a solar cell simulator specialized to geometry optimization. Specifically, the simulator is based on the concept of “automatic differentiation,” (AD) i.e. each coded operation carries information about derivatives; within this paradigma, the solar cell efficiency can be automatically differentiated with respect to any material parameter via the chain rule. The simulator has been developed in collaboration with colleagues at Google Brain and will be released as an open-source project. We anticipate to use Google’s “JAX” [2] as the AD library. The UROP will help test and develop the simulator, along with other team members. [1] https://www.nrel.gov/pv/cell-efficiency.html [2] https://github.com/google/jax

Pre-requisites

Python fluency and knowledge of Calculus are required.