backtop


Print 11 comment(s) - last by Bubbacub.. on Jan 29 at 4:48 PM

Genetic algorithm discovered unusual pattern was best at trapping light

A common problem in the design of thin film solar cells is how to develop optical materials and nanopatterns for those materials.  The ultimate goal is to trap and capture as much solar energy as possible.  The problem is far too challenging for naive searches based solely on human creativity.  Thus artificial intelligence is increasingly being employed to search for the optimal cell materials and nanopatterns.

Northwestern University Mechanical Engineering Professor Wei Chen and her graduate student Cheng Sun have published a new paper on how genetic algorithms -- an artificial intelligence technique based on evolutionary biology -- can be used to develop high performance nano-patterns.

Much like evolution and genetic processes serve in the real world serve to select creatures giving rise to fit species like humans or drug-resistant bacteria, genetic algorithms weed out bad candidates, while preserving and mixing elements of the fittest performers.

Starting with dozens of random designs, Prof. Chen's team "bred" the nanopatterns through 20 generations, employing genetic algorithm techniques like mutation and crossover.  

The result was a strike nanopattern that outperformed other designs.
 
Nanopatterned thin fillm
The "fittest" pattern [Image Source: Northwestern University]

The optimized 100-nanometer-thick organic dielectric "scattering layer" appears superb in simulations at trapping photons and transmitting them into the active layer.  In fact, the simulation results predict that the layer will surpass three-fold the Yablonovitch Limit; a thermodynamic limit developed in the 1980s that statistically describes how long a photon can be trapped in a semiconductor.

Current organic solar cells have traditional been, in a word, bad.  While relatively cheap to produce compared to their rare-metal thin film counterparts, their low efficiencies make them a disappointing dead end.  But the new design is intriguing as it may boost organic thin film cells into a regime in which they would actually be cost effective -- perhaps more so than rare-metal designs.

Comments Prof. Chen, "Due to the highly nonlinear and irregular behavior of the system, you must use an intelligent approach to find the optimal solution.  Our approach is based on the biologically evolutionary process of survival of the fittest."

Northwestern Univ. is currently working with Argonne National Laboratory to produce a prototype of the nanopatterned cell, for real world testing.

In the mean time, the work has been published in the peer-reviewed journal Scientific Reports, a Nature journal.

Sources: Northwestern University, Scientific Reports



Comments     Threshold


This article is over a month old, voting and posting comments is disabled

RE: evolution
By Bubbacub on 1/29/2013 2:07:21 PM , Rating: 2
apologies with regard to this article.

i would put it to you that you do perhaps have a slight tendency to sensationalise the unremarkable on occasion.


"I'm an Internet expert too. It's all right to wire the industrial zone only, but there are many problems if other regions of the North are wired." -- North Korean Supreme Commander Kim Jong-il














botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki