KINGSTON, R.I. – April 28, 2011 – Researchers have spent decades trying to develop machines and technologies that can mimic the human brain and make intelligent decisions. It’s a field that Haibo He, assistant professor of electrical engineering at the University of Rhode Island, is beginning to make headway in.
His research received a significant boost this month when He learned that he was awarded a $400,000 research grant from the National Science Foundation through its Faculty Early Career Development program, the most prestigious NSF grant program supporting junior faculty members.
A resident of North Kingstown, He said that most research in the field to date has focused on developing technologies to achieve one of two brain-like objectives – making the right decision (optimization) or predicting what will happen next. Instead, his research is aimed at integrating the two objectives into one technology.
He said he will use a methodology called adaptive dynamic programming to develop algorithms and models that can replicate the brain’s intelligence and then design computer hardware to apply these algorithms to real-world needs.
For instance, He said that intelligent power systems (i.e. smart grid) could be designed to maintain the power grid and “control the mechanisms to deliver power to the right place at the right time in the most efficient way while maintaining system stability.” If a fault occurred in a grid-connected wind farm, the technology could make predictions and take appropriate control actions to maintain the stability of the power system.
This technology could also be applied in wireless communication networks or even helicopter controls, among many other applications.
“It could be used in almost any complex system where multiple components are required to work together,” He said. “It’s fundamental research that starts with understanding the brain and then developing the architectures and models for computational intelligence.”
During the five-year grant period, He will study the mathematical foundation of the principles of adaptive dynamic programming, create the models and architectures to replicate certain levels of brain-like intelligence, and finally demonstrate its use in simulated applications.
He hopes to use the proposed wind energy farm off the coast of Rhode Island as one example of how his research can be applied. He is collaborating with URI ocean engineers and Deepwater Wind on this component of the project.
His research has already generated enthusiasm among his colleagues. He has been invited to give presentations on the topic to researchers in China, France and Northern Ireland, among other places.
An important element of the National Science Foundation’s Faculty Early Career Development grants is to link research with teaching and outreach. To fulfill these requirements, He will design new courses based on his research on adaptive dynamic programming and develop teaching modules that can be shared online with other universities. He will also host a summer workshop for academic faculty and industrial engineers and scientists from around the world so they can share their experience and latest research results in this field.
In addition, He will work with professors at the Community College of Rhode Island to arrange opportunities to visit with students there in an effort to generate interest in this discipline. He also plans to demonstrate the principles of his research to students at Times Squared Academy in Providence to encourage them to pursue careers in science, technology, engineering and mathematics.
URI Department of Communications & Marketing photo by Michael Salerno Photography.