表題番号:2022C-486 日付:2023/02/02
研究課題Development of electromagnetic reservoir computing system via wireless power transmission technology for voice classification and word prediction tasks
研究者所属(当時) 資格 氏名
(代表者) 理工学術院 情報生産システム研究センター 助教 アズハリ サマン
(連携研究者) Graduate School of Information, Production and Systems Waseda University Professor Takeo Miyake
(連携研究者) Kyushu Institute of Technology Professor Hirofumi Tanaka
(連携研究者) Graduate School of Information, Production and Systems Waseda University Assistant professor Gabor Mehes
(連携研究者) Kyushu Institute of Technology Assistant professor Yuki Usami
研究成果概要

In the development of an electromagnetic (EM) reservoir computing system for classification tasks, inputting data and measuring interferences in the EM field pose a challenge. We have developed a system that converts high-frequency AC signals to DC to overcome the high-frequency sampling difficulties. The system uses a piezoresistive material to induce changes in the EM field between the transmitter and receiver, which are then observed as a DC voltage drop in the transmitter. These changes serve as a means of measurement for inputting data into the EM reservoir.

 

Our research has successfully addressed the challenges of inputting data and measuring interferences in the EM field for a reservoir computing system. The next step is to increase the number of output nodes to enhance the accuracy of the tasks performed by the system. The results of our work, titled "Integration of wireless power transfer technology with hierarchical multi-walled carbon nanotubes-Polydimethylsiloxane piezo-responsive pressure sensor for remote force measurement", are currently under review by IEEE Sensor.

 

In conclusion, our findings demonstrate the potential of using the EM field as a reservoir computing system to perform time-dependent and classification tasks. The system has numerous potential applications and could have significant implications in various fields.