• Title/Summary/Keyword: 지역 및 전역적 시계열 특징

Search Result 3, Processing Time 0.02 seconds

Passive sonar signal classification using attention based gated recurrent unit (어텐션 기반 게이트 순환 유닛을 이용한 수동소나 신호분류)

  • Kibae Lee;Guhn Hyeok Ko;Chong Hyun Lee
    • The Journal of the Acoustical Society of Korea
    • /
    • v.42 no.4
    • /
    • pp.345-356
    • /
    • 2023
  • Target signal of passive sonar shows narrow band harmonic characteristic with a variation in intensity within a few seconds and long term frequency variation due to the Lloyd's mirror effect. We propose a signal classification algorithm based on Gated Recurrent Unit (GRU) that learns local and global time series features. The algorithm proposed implements a multi layer network using GRU and extracts local and global time series features via dilated connections. We learns attention mechanism to weight time series features and classify passive sonar signals. In experiments using public underwater acoustic data, the proposed network showed superior classification accuracy of 96.50 %. This result is 4.17 % higher classification accuracy compared to existing skip connected GRU network.

Fusion of Multi-Scale Features towards Improving Accuracy of Long-Term Time Series Forecasting (다중 스케일 특징 융합을 통한 트랜스포머 기반 장기 시계열 예측 정확도 향상 기법)

  • Min, Heesu;Chae, Dong-Kyu
    • Annual Conference of KIPS
    • /
    • 2022.11a
    • /
    • pp.539-540
    • /
    • 2022
  • 본 논문에서는 정확한 장기 시계열 예측을 위해 시계열 데이터의 다양한 스케일 (시간 규모)에서 표현을 학습하는 트랜스포머 모델을 제안한다. 제안하는 모델은 시계열의 다중 스케일 특징을 추출하고, 이를 트랜스포머에 반영하여 예측 시계열을 생성하는 구조로 되어 있다. 스케일 정규화 과정을 통해 시계열의 전역적 및 지역적인 시간 정보를 효율적으로 융합하여 종속성을 학습한다. 3 가지의 다변량 시계열 데이터를 이용한 실험을 통해 제안하는 방법의 우수성을 보인다.

A Study on the Hydraulic Factors of Groundwater Level Fluctuation by Region in Jeju Island (제주도 지역별 지하수위 변동 요인에 대한 고찰)

  • Jeong, Jiho;Park, Jaesung;Koh, Eun-hee;Park, Won-bae;Jeong, Jina
    • The Journal of Engineering Geology
    • /
    • v.32 no.2
    • /
    • pp.257-270
    • /
    • 2022
  • This study evaluated the hydraulic factors contributing to the decreasing groundwater levels across Jeju island. Time-series data for groundwater level, precipitation, and groundwater usage and information on land use were acquired, and the correlations among them were analyzed to evaluate the causes of the decreasing groundwater. The effects of precipitation and groundwater usage on the fluctuations of groundwater level were quantified using response surface analysis and sensitivity analysis, and methods for groundwater quantity management by region were proposed. The results showed that the rate of groundwater decrease in the western region was larger than that in the eastern region. For the eastern region, the influence of precipitation was large and the rate of decrease in the groundwater level was relatively small. The geological formation of this part of the island and continuous seawater intrusion suggest that although the absolute amount of groundwater extracted for use was large, the decrease in the groundwater level was not seen to be great due to an increase in pressure by seawater intrusion. Overall, precipitation and groundwater usage had the greatest effect on the amount of groundwater in the western region, and thus their data would be most useful for informing groundwater management, whereas other factors (e.g., sea level and the location of the freshwater-seawater transition zone) must be considered when understanding Jeju's eastern region. As the characteristics of groundwater level fluctuations in the eastern and western regions are distinct, an optimal management plan for each region should be proposed to ensure the efficient management of groundwater quantity.