著者

Takumi Wada Haruhisa Ichikawa Shinji Yokogawa Yoshito Tobe Yuusuke Kawakita

出版者

The Institute of Electronics, Information and Communication Engineers

雑誌

IEICE Communications Express (ISSN:21870136)

巻号頁・発行日

vol.11, no.8, pp.468-473, 2022-08-01 (Released:2022-08-01)

参考文献数

6

---

The study presents a “virtual grid system” currently being developed by the authors. The system consists of a power supply, load, and virtual grid hub (VG-Hub) network connecting them. The VG-Hub network must have a large power distribution capability. Therefore, the challenge is to obtain a graph to evaluate the maximum feasible flow when multiple VG-Hubs are connected to a power source or load. In this paper, we propose a method for generating a graph to derive the maximum feasible flow. A quantitative evaluation of the proposed method is presented.

著者

Minoru ASANO Shinji YOKOGAWA Haruhisa Ichikawa

出版者

The Institute of Electronics, Information and Communication Engineers

雑誌

IEICE Communications Express (ISSN:21870136)

巻号頁・発行日

pp.2022XBL0066, (Released:2022-05-10)

---

Autonomous distributed power grids have attracted attention as a way to utilize renewable energy to achieve a carbon-neutral society. In order to properly operate these grids, it is necessary to obtain sufficient information on the supply and demand power capabilities and battery health of connected devices in a short time. In addition, methods based on direct current are essential to maximizing the use of renewable energy. This study proposes a method for acquiring information about the energy storage of devices connected to the grid via USB power delivery using deep learning techniques. Furthermore, we propose a method to diagnose the embedded battery health of the device based on short-time monitoring.

著者

praveen singh Thakur Masaru Sogabe Katsuyoshi Sakamoto Koichi Yamaguchi Dinesh Bahadur Malla Shinji Yokogawa Tomah Sogabe

出版者

人工知能学会

雑誌

2018年度人工知能学会全国大会(第32回)

巻号頁・発行日

2018-04-12

---

In this paper, for stable learning and faster convergence in Reinforcement learning continuous action tasks, we propose an alternative way of updating the actor (policy) in Deep Deterministic Policy Gradient (DDPG) algorithm. In our proposed Hybrid-DDPG (shortly H-DDPG), at one time step actor is updated similar to DDPG and another time step, policy parameters are moved based on TD-error of critic. Once among 5 trial runs on RoboschoolInvertedPendulumSwingup-v1 environment, reward obtained at the early stage of training in H-DDPG is higher than DDPG. In Hybrid update, the policy gradients are weighted by TD-error. This results in 1) higher reward than DDPG 2) pushes the policy parameters to move in a direction such that the actions with higher reward likely to occur more than the other. This implies if the policy explores at early stages good rewards, the policy may converge quickly otherwise vice versa. However, among the remaining trial runs, H-DDPG performed same as DDPG.