Plans to develop the smart grid – a system that uses intelligent computer networks to manage electric power – cannot succeed without creating “thinking machines.”
These machines will learn and adapt to new situations, from power outages along the grid to fluctuations in the power supply, said Dr. Ganesh Kumar Venayagamoorthy, a power engineering expert at Missouri University of Science and Technology.
They must take on almost human-like intelligent characteristics, such as the ability to make decisions, adapt to unfamiliar situations, learn from changes in their environments and make sense of how all of the electricity flows through the nation’s power grid, said Venayagamoorthy, a professor of electrical and computer engineering at Missouri S&T.
“And those capabilities, in turn, will depend on subsystems that continuously improve their knowledge of grid dynamics, and not just gather data,” Venayagamoorthy said. The machines will arise from research in two fields of computing: Computational intelligence (CI) and adaptive critic designs (ACDs).
“The research is already well under way” to support the development of thinking machines, he said. “As the smart grid evolves over time, and as we become more dependent on intermittent sources of energy such as wind and solar power, we will see that traditional technology will not work.”
Computational intelligence is the successor to artificial intelligence “and the way forward in future computing,” Venayagamoorthy said. Adaptive critic designs are like “a teacher-student framework” where one part of a computer network (the student) learns from another.
Such a computational systems thinking machine would be capable of managing complex power systems, Venayagamoorthy said. Venayagamoorthy is the founder of Missouri S&T’s Real-Time Power and Intelligent Systems Laboratory and a principal investigator in the Brain2Grid project, an effort funded through the National Science Foundation’s Office of Emerging Frontiers in Research and Innovation.
“Several research studies have reported using CI-based technologies to dynamically forecast wind and solar power, monitor voltage stability, and assess real-time stability” of power systems, he said.
“In combination, CI and ACD technologies can provide a smart grid with capabilities for dynamic foresight, sense-making, situational awareness, rapid adaptation, fault-tolerance and robustness,” Venayagamoorthy said. For example, when the demand for electricity is low, a thinking machine should be able to dispatch power from wind farms to storage units that can store the excess power for later use.