Award Finalist/Winner

SC Conference - Activity Details

Cerebral Methodology Based Computing for Estimating Validity of Simulation Results

Team Members:
Yoshio Suzuki  (Japan Atomic Energy Agency)
Akemi Nishida  (Japan Atomic Energy Agency)
Tomonori Yamada  (Japan Atomic Energy Agency)
Fumimasa Araya  (Japan Atomic Energy Agency)
Sachiko Hayashi  (Japan Atomic Energy Agency)
Norihiro Nakajima  (Japan Atomic Energy Agency)
Toshio Hirayama  (Japan Atomic Energy Agency)
Challenges Session
SC Analytics Challenge
Tuesday,  04:14PM - 04:36PM
Room 14
Our challenge is to estimate the validity of simulation results by establishing "Cerebral Methodology Based Computing (CMC)". The estimation of the validity is indispensable to accurately predict phenomena by using simulation. It is difficult to estimate the validity from a single simulation because the simulation contains uncertainties and ambiguities. We have proposed CMC as the computing methodology which enables us to estimate the validity of results with both deductive and inductive approaches, similar to the human thinking process. To realize CMC, we have constructed the system which enables us to execute both deductive simulations and inductive simulations, and then combine results of those simulations in an integration process. We have applied the system to a thermal displacement analysis of a nuclear power plant and confirmed the usefulness of the system.
   IEEE Computer Society  /  ACM     2 0   Y E A R S   -   U N L E A S H I N G   T H E   P O W E R   O F   H P C