[2K01-05] Advancement of the Maintenance and Preservation Technology of the Plant 2

[2K01] Development of Abnormal Sign Detection System using AI for nuclear power plant

(3)Overview of Abnomal Sign Detection System

*Isaku Nagura¹, Shinya Tominaga¹, Ryota Miyake¹, Toshio Aoki¹, Susumu Naito², Yasunori Taguchi², Yuichi Kato², Kouta Nakata² (1. TOSHIBA ESS, 2. TOSHIBA)

Keywords:early detection of anomaly signs, AI, deep learning, autoencoder, Plant Operation Monitoring

Thousands of process values are measured in nuclear power plants for stable and safe operation, but it is difficult for operators to constantly monitor all process values. And sufficient knowledge and experience is required to identify the subtle signal abnormality from a lot of measurement data. Therefore, using AI technology, which is being utilized with the progress of computer technology in recent years, the development of abnormal sign detection system is expected. To achieve this detection system, we have developed a detection algorithm based on an auto encoder, which is one of the deep learning algorithms. In addition, in order to apply this detection system to a nuclear power plant, we examined the user interface and equipment configuration. This time, we will report the detection method of the anomaly sign detection system and the system configuration for nuclear power plants as an overview of these development results.