• 한국음향학회지
• /
• 제42권4호
• /
• pp.313-319
• /
• 2023

본 논문에서는 일부 소음 데이터만 알고 있을 때 결손된 데이터를 예측할 목적으로 수조에서 측정된 기포유동 소음 데이터와 수중 운동체 발사 소음 데이터를 시계열 기계학습 모델인 Long Short Term Memory(LSTM)에 적용해 보았다. 기포유동소음 데이터는 파이프에서 측정된 소음으로 기포소음, 유동소음, 유체기인소음이 혼합되어 있으며 유형별로 3가지로 분류할 수 있다. 수중 운동체 발사소음은 모형 발사튜브에서 수중 운동체가 사출될 때 발생하는 소음으로 순간소음이며 발사 이벤트마다 불규칙하게 변한다. 이러한 종류의 소음 생성을 위해서는 해석적인 모델보다는 데이터 기반 모델이 유용할 수 있다. 본 연구에서는 LSTM을 데이터 기반 모델을 만들었다. 모델에 영향을 주는 LSTM의 은닉유닛의 개수, 입력시퀸스의 개수, 데시메이션 인자에 따른 모델의 성능을 확인하고 최적의 LSTM 모델을 구성했다. 같은 유형은 새로운 데이터에 대해서도 잘 동작하는 것을 보였다.

• 한국소음진동공학회논문집
• /
• 제22권12호
• /
• pp.1213-1219
• /
• 2012

The purpose of this study is to present a guideline to design a short-term manual measurement of environmental noise level, which is more economical and flexible, but less representative than long-term automatic measurement. The proposed guideline can provide the number of measurement times and the length of measurement term required to secure the extent of the representativeness. The data was collected at 4 sites located in Seoul and at 4 sites located outside of Seoul. The probabilities for five-minute equivalent noise levels, Leq, 5min, to stay in an error range from the quarterly representative noise level were used to evaluate sampling techniques. The probability analysis of the daytime period showed that the noise levels measured between 10 am and 2 pm and between 9 pm and 10 pm have the probabilities higher than 60 %. On the other hand, even for the same length of total measurement time, increasing the number of random samplings results in higher probabilities than increasing the length of measurement term.

• 한국음향학회지
• /
• 제29권3호
• /
• pp.200-208
• /
• 2010

본 논문에서는 배경잡음에 강인한 잡음제거 알고리즘 설계를 위해서 minimum statistics (MS) 기법을 codebook driven short-term predictor parameter estimation (CDSTP) 기법에 접목하는 방법을 제안한다. MS는 stationary 배경잡음에는 강인하지만, non-stationary 배경잡음에는 상대적으로 취약하다. CDSTP는 non-stationary 배경잡음에 강인한 특성을 보이지만, 코드북에 없는 배경잡음 환경에는 취약하다. 따라서 non-stationary 배경잡음에 강인한 CDSTP 방법과 별도의 코드북 학습 과정이 필요 없는 MS를 결합해서 다양한 배경잡음에 강인한 알고리즘을 제안한다. 제안방법은 MS나 CDSTP 방법에 비해서 전체적으로 향상된 perceptual evaluation of speech quality (PESQ) 성능을 나타냈으며, 특히 stationary 배경잡음과 non-stationary 배경잡음이 섞여 있는 mixed 배경잡음 환경에서 강인한 특성을 보였다.

• 산업기술연구
• /
• 제27권B호
• /
• pp.21-27
• /
• 2007

The power line channel has time-variant characteristics caused by various kind of electrical devices. This characteristics are synchronized with the main voltage by their own characteristics. The main factors of disturbance are the variation of the channel impedance and noises. In other papers, the synchronous noise modeling has been achieved. But the modeling is not satisfied simultaneously with the time domain and the frequency domain and there are not any discussion about short-term variations of the channel impedance which cause to the signal fading. Therefore, this paper researched to solve problems about the signal fading by analyzing the short-term variation of the channel impedance, and proposed the synchronous noise modeling which is satisfied simultaneously in the time domain and the frequency domain.

• International Journal of Internet, Broadcasting and Communication
• /
• 제12권3호
• /
• pp.220-225
• /
• 2020

This paper investigates the non-intrusive speech intelligibility estimation method in noise environments when the bottleneck feature of autoencoder is used as an input to a neural network. The bottleneck feature-based method has the problem of severe performance degradation when the noise environment is changed. In order to overcome this problem, we propose a novel non-intrusive speech intelligibility estimation method that adds the noise environment information along with bottleneck feature to the input of long short-term memory (LSTM) neural network whose output is a short-time objective intelligence (STOI) score that is a standard tool for measuring intrusive speech intelligibility with reference speech signals. From the experiments in various noise environments, the proposed method showed improved performance when the noise environment is same. In particular, the performance was significant improved compared to that of the conventional methods in different environments. Therefore, we can conclude that the method proposed in this paper can be successfully used for estimating non-intrusive speech intelligibility in various noise environments.

• 대한전기학회논문지
• /
• 제32권5호
• /
• pp.183-188
• /
• 1983

This paper deals with short-term load forecasting. The load model is represented by the state variable form to exploit the Kalman filter technique. The load model is derived from Taylor series expansion and remainder term is considered as noise term. In order to solve recursive filter form, among various algorithm of solving Kalman filter, this paper uses exponential data weighting technique. This paper also deals with the asymptotic stability of filter. Case studies are carried out for the hourly power demand forecasting of the Korea electrical system.

• Journal of Electrical Engineering and Technology
• /
• 제13권4호
• /
• pp.1419-1424
• /
• 2018

Very short-term electric load forecasting is essential for real-time power system operation. In this paper, a very short-term electric load forecasting technique applying the Kalman filter algorithm is proposed. In order to apply the Kalman filter algorithm to electric load forecasting, an electrical load forecasting algorithm is defined as an observation model and a state space model in a time domain. In addition, in order to precisely reflect the noise characteristics of the Kalman filter algorithm, the optimal error covariance matrixes Q and R are selected from several experiments. The proposed algorithm is expected to contribute to stable real-time power system operation by providing a precise electric load forecasting result in the next six hours.

• 한국소음진동공학회:학술대회논문집
• /
• 한국소음진동공학회 2013년도 추계학술대회 논문집
• /
• pp.553-554
• /
• 2013

• Journal of Oral Medicine and Pain
• /
• 제11권1호
• /
• pp.29-35
• /
• 1986

18 TMD patients who received occlusal adjustment in PNUH though Jan.1984 to 1985 were followed up for short-term(2-6yrs.) and long-term(1-2yrs.) evaluation. The obtained results were as follows : 1. Pain index showed gradual decrease after occlusal adjustment and significant change on long-term evaluation. 2. Noise index had no significant change throughout the all follow-up evaluation. 3. Opening limitation index showed gradual decrease after occlusal adjustment and significant change on both long-term and short-term evaluation. 4. Maximum comfortable opening exhibited more and more increase and significant change on long-term evaluation.

• 말소리와 음성과학
• /
• 제8권1호
• /
• pp.63-69
• /
• 2016

The aims of the present study are two-fold: 1) to explore differences on speech perception between younger and older adults according to noise conditions; and 2) to investigate which cognitive domains are correlated with speech perception. Data were acquired from 15 younger adults and 15 older adults. Sentence recognition test was conducted in four noise conditions(i.e., in-quiet, +5 dB SNR, 0 dB SNR, -5 dB SNR). All participants completed auditory and cognitive assessment. Upon controlling for hearing thresholds, the older group revealed significantly poorer performance compared to the younger adults only under the high noise condition at -5 dB SNR. For older group, performance on Seoul Verbal Learning Test(immediate recall) was significantly correlated with speech perception performance, upon controlling for hearing thresholds. In older adults, working memory and verbal short-term memory are the best predictors of speech-in-noise perception. The current study suggests that consideration of cognitive function for older adults in speech perception assessment is necessary due to its adverse effect on speech perception under background noise.