Analyzing the accuracy of a computer chip’s layout will be faster and more efficient, thanks to the work of a UMR Ph. D. student who has developed a way to use fuzzy logic and neural networks to analyze printed circuit layouts.
"We are the first to apply fuzzy logic to this application," says Nian Zhang, a Ph.D. student in computer engineering. "We have the research completed to the point that a company could take this algorithm and reproduce and sell this smart chip."
Using fuzzy logic, Zhang is able to take imprecise and vague information and draw definite conclusions. She won the Best Student Paper Award for this concept at the Institute of Electrical and Electronics Engineers International Conference on Fuzzy Systems in St. Louis in May.
"Nian’s work helps you determine that you’ve done the layout correctly, and she’s found a faster way to do it using neural networks," says Dr. Donald Wunsch, the Mary K. Finley Missouri Distinguished Professor of Computer Engineering at UMR and Zhang’s advisor. "It’s very flexible and general, which makes the possible applications endless."
She has spent the past four years perfecting the fuzzy logic method used to analyze and verify these chips. Zhang says through the use of neural networks the chips can be more quickly checked for correctness. "It can do some fancy things," she says. It just depends what the ideal application may be.
The National Science Foundation and the Mary K. Finley Endowment Fund has helped support Zhang’s research in the area of fuzzy logic and neural networks.
Zhang and Wunsch have made a University of Missouri System invention disclosure seeking a patent based on the results of Zhang’s research, and also submitted a proposal to the National Science Foundation’s Small Business Innovation Research Division.