[1P08-12] Robot, Instrumentation, Support System

[1P11] Electrical Isolation Supporting System

(2)Automatic Isolation Planning with Deep Learning

*Susumu Naito¹, Kei Takakura¹, Hiroki Shiba² (1. Power and Industrial Systems Research and Development Center, Toshiba Corporation, 2. Isogo Nuclear Engineering Center, Toshiba Corporation)

Keywords:data mining, deep learning, deep Q-network, artificial intelligence

Currently, a skilled engineer plans electrical isolation procedure with hundreds of the circuit diagrams and the related documents, taking man-hours. If this task becomes automatic, it is very efficient. The automatic planning has two issues. One is much calculation time of electrical circuit simulator, searching billions of electrical conducting paths. The other is interpretation of human fuzzy information, such as ease of work. Deep learning is helpful to resolve them. We performed a model case study of the electrical isolation. We applied a deep neural network (DNN) for dropping in the calculation time. We trained the circuit diagrams and the responses of the circuit simulator to the DNN, constructing an efficient path search algorithm in the DNN. The calculation time of the DNN was shorter by a factor of 32 for the model case, compared with that of the electrical circuit simulator. We applied a deep Q-network (DQN) for the interpretation of fuzzy information. The DQN optimized isolation procedure from the viewpoint of the work load with the equipment layout, related doses, and values of wet bulb globe temperature (WBGT) to prevent heatstroke.