• Journal of the Korean Solar Energy Society
• /
• v.40 no.3
• /
• pp.1-20
• /
• 2020

In accordance with the Government's policy, renewable power generation is expanding very largely. This leads to increasing uncertainty in the power market and power system owing to the intermittent and fluctuating output characteristics of renewable power generators. Data on the acquisition of renewable power generators can be largely classified according to the operation of the power market and power system. Data on the settlement for the payment for the power amount are acquired in the power market, and real-time data for monitoring the status and output of the generators are acquired in the power system. However, renewable power generators operating in the power market have different acquisition cycles depending on the method of communication of the power meter. They acquire data only for settlement purposes and have no real-time data, which requires improvement. In this paper, the acquisition status is reviewed by classifying the data of renewable power generators into settlement and real-time data. In addition, measures and acquisition criteria for real-time data of renewable power generators for improving the acquisition method are proposed.

• Smart Media Journal
• /
• v.12 no.4
• /
• pp.19-26
• /
• 2023

IoT edge services utilizing neuromorphic hardware architectures are suitable for autonomous IoT applications as they perform intelligent processing on the device itself. However, spiking neural networks applied to neuromorphic hardware are difficult for IoT developers to comprehend due to their complex structures and various hyper-parameters. In this paper, we propose a method for generating spiking neural network (SNN) models that satisfy user performance requirements while considering the constraints of neuromorphic hardware. Our proposed method utilizes previously trained models from pre-processed data to find optimal SNN model parameters from profiling data. Comparing our method to a naive search method, both methods satisfy user requirements, but our proposed method shows better performance in terms of runtime. Additionally, even if the constraints of new hardware are not clearly known, the proposed method can provide high scalability by utilizing the profiled data of the hardware.