SPbPU: Algorithm for instant detection of accident cause at NPP

Main article: Information technology in power

2023: Algorithm Development

SPBPUS scientists have developed an algorithm for predictive analysis, detection and classification of accidents at nuclear power plants at an early stage using artificial intelligence without the participation of a human operator.

This algorithm is applicable in cases of emergency situations with loss of coolant - artificial intelligence within two seconds determines the cause of the accident and offers actions to be taken to eliminate it.

The solution to the problem was obtained on the basis of an analogy with the gene expression algorithm (GEP) - an evolutionary algorithm that creates computer programs and models.

The algorithm processed a set of dimensionless source data to develop relationships at different points in time after the start of the accident. The results showed that the GEP algorithm can be used as a powerful tool to determine progress in a rapid emergency at a nuclear power plant.

This technique is proposed for the first time in the field of nuclear power plants. The algorithm is able to analyze complex functions and is suitable for stations with a large number of parameters that are constantly changing, which allows you to control the situation at the nuclear power plant, - comments the leading participant of the project Hashayar Sadegi.

The results of the project are published in the journal of the first quartile Energies.

The project is attended by scientists from Iran who have already completed graduate school at the Polytechnic University and work as senior teachers. The team develops at the Polytechnic University the center "Promising atomic technologies in the direction of sustainable development and decarbonization of power."