• 디지털콘텐츠학회 논문지
• /
• 제18권8호
• /
• pp.1671-1676
• /
• 2017

빅데이터 시대의 도래로 다양한 데이터들이 발생되고 있다. 많은 산업 부분에서는 이러한 데이터들을 수집하여 분석하고자 한다. 하지만 사용자 정보 수집은 직접적인 개인정보 유출을 초래할 수 있다. 구글(Google) 사에서 제안한 지역 차분 프라이버시 기법은 데이터 변조를 통해 사용자 정보 수집에 있어 발생할 수 있는 개인정보 유출을 방지한다. 이러한 데이터 변조를 통한 개인정보 유출 방지는 그 변조되는 정도가 높을수록 개인정보를 강력히 보장하지만 이와 반대로 데이터의 활용도는 현저히 떨어진다. 그래서 데이터 변조의 정도를 데이터 수집목적에 적합하게 설정해야한다. 본 논문에서 제시하는 시뮬레이션 도구는 지역 차분 프라이버시를 만족하는 사용자 정보 수집에 있어 설정해야하는 다양한 변수값을 데이터 수집환경에 맞게 적용함으로써 데이터 수집가가 자신의 환경에 맞는 데이터 수집을 할 수 있도록 지원한다.

• 정보보호학회논문지
• /
• 제32권4호
• /
• pp.647-660
• /
• 2022

최근 정상 데이터와 일부 비정상 데이터를 보유한 환경에서 딥러닝 기반 준 지도 학습 이상 탐지 기법이 매우 효과적으로 동작함이 알려져 있다. 하지만 사이버 보안 분야와 같이 실제 시스템에 대한 알려지지 않은 공격 등 비정상 데이터 확보가 어려운 환경에서는 비정상 데이터 부족이 발생할 가능성이 있다. 본 논문은 비정상 데이터가 정상 데이터보다 극히 작은 환경에서 준 지도 이상 탐지 기법에 적용 가능한 섭동을 활용한 초구 기반 비정상 데이터 증강 기법인 ADA-PH(Abnormal Data Augmentation Method using Perturbation based on Hypersphere)를 제안한다. ADA-PH는 정상 데이터를 잘 표현할 수 있는 초구의 중심으로부터 상대적으로 먼 거리에 위치한 샘플에 대해 적대적 섭동을 추가함으로써 비정상 데이터를 생성한다. 제안하는 기법은 비정상 데이터가 극소수로 존재하는 네트워크 침입 탐지 데이터셋에 대하여 데이터 증강을 수행하지 않았을 경우보다 평균적으로 23.63% 향상된 AUC가 도출되었고, 다른 증강 기법들과 비교했을 때 가장 높은 AUC가 또한 도출되었다. 또한, 실제 비정상 데이터에 유사한지에 대한 정량적 및 정성적 분석을 수행하였다.

• International Journal of Control, Automation, and Systems
• /
• 제6권4호
• /
• pp.506-514
• /
• 2008

This paper presents an intelligent model; named as free model, approach for a closed-loop system identification using input and output data and its application to design a power system stabilizer (PSS). The free model concept is introduced as an alternative intelligent system technique to design a controller for such dynamic system, which is complex, difficult to know, or unknown, with input and output data only, and it does not require the detail knowledge of mathematical model for the system. In the free model, the data used has incremental forms using backward difference operators. The parameters of the free model can be obtained by simultaneous perturbation stochastic approximation (SPSA) method. A linear transformation is introduced to convert the free model into a linear model so that a conventional linear controller design method can be applied. In this paper, the feasibility of the proposed method is demonstrated in a one-machine infinite bus power system. The linear quadratic regulator (LQR) method is applied to the free model to design a PSS for the system, and compared with the conventional PSS. The proposed SPSA-based LQR controller is robust in different loading conditions and system failures such as the outage of a major transmission line or a three phase to ground fault which causes the change of the system structure.

• 정보처리학회논문지B
• /
• 제9B권3호
• /
• pp.277-286
• /
• 2002

양자역학 섭동이론과 유전자프로그래밍(GP) 기법을 접목시킴으로써 실세계(Real-world)에서 발생하는 카오스 시계열에 대하여 수학모델을 구축, 예측하기 위한 새로운 알고리즘을 개발하였다. 시계열 분석과 양자역학 파동방정식의 해를 구하는 섭동이론과의 절차적 유사성을 논하고, 이것을 GP로 구현하는 전형적 접근방안을 제시한다. 함수집합(Function Set)으로서 직교함수(Orthogonal Functions)를 이용하고 병렬 집단을 사용하는 GP를 이용하여 원 시계열에 대한 초기 수학모델을 구하고, 원 시계열 데이터로부터 모델의 평가값을 뺀 나머지로 구성되는 잔여 시계열에 대하여 다시 GP를 적용하는 과정을 일정한 종료조건이 충족될 때가지 반복함으로써 실세계 카오스 시계열에 대한 정확성 높은 수학모델을 구축하는데 성공하였다. 타 방법론과의 비교와 향후 해결과제에 대하여도 소개한다.

• 말소리와 음성과학
• /
• 제9권2호
• /
• pp.85-94
• /
• 2017

Conventional speaker verification systems using time delay neural network, identity vector and probabilistic linear discriminant analysis (TDNN-Ivector-PLDA) are known to be very effective for verifying long-duration speech utterances. However, when test utterances are of short duration, duration mismatch between enrollment and test utterances significantly degrades the performance of TDNN-Ivector-PLDA systems. To compensate for the I-vector mismatch between long and short utterances, this paper proposes to use probabilistic linear discriminant analysis (PLDA) model adaptation with augmented data. A PLDA model is trained on vast amount of speech data, most of which have long duration. Then, the PLDA model is adapted with the I-vectors obtained from short-utterance data which are augmented by using vocal tract length perturbation (VTLP). In computer experiments using the NIST SRE 2008 database, the proposed method is shown to achieve significantly better performance than the conventional TDNN-Ivector-PLDA systems when there exists duration mismatch between enrollment and test utterances.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1992년도 하계학술대회 논문집 A
• /
• pp.330-332
• /
• 1992

This paper tries to evaluate the control performance of systeme in cases of parameter perturbations. Six cases of the root distribution and location changes of each characteristic roots by the parameter perturbation are considered as evaluation factors. A qualitative evaluation is performed through several simulations. These results will be used as a basic data for the complete analysis of the control performance.

• 한국정밀공학회:학술대회논문집
• /
• 한국정밀공학회 2004년도 추계학술대회 논문집
• /
• pp.1160-1165
• /
• 2004

The purpose of this study was to calculate three dimensional angular displacements, moments and joint reaction forces of the ankle joint during the waist pulling, and to assess the ankle joint reaction forces according to different perturbation modes and different levels of perturbation magnitude. Ankle joint model was assumed 3-D ball and socket joint which is capable of three rotational movements. We used 6 cameras, force plate and waist pulling system. Two different waist pulling systems were adopted for forward sway with three magnitudes each. From motion data and ground reaction forces, we could calculate 3-D angular displacements, moments and joint reaction forces during the recovery of postural balance control. From the experiment using falling mass perturbation, joint moments were larger than those from the experiment using air cylinder pulling system with milder perturbation. However, JRF were similar nevertheless the difference in joint moment. From this finding, we could conjecture that the human body employs different strategies to protect joints by decreasing joint reaction forces, like using the joint movement of flexion or extension or compensating joint reaction force with surrounding soft tissues. Therefore, biomechanical analysis of human ankle joint presented in this study is considered useful for understanding balance control and ankle injury mechanism.

• 한국연소학회:학술대회논문집
• /
• 한국연소학회 2004년도 제29회 KOSCI SYMPOSIUM 논문집
• /
• pp.152-160
• /
• 2004

Studies to investigate the influence on hydraulic acoustic wave were conducted using pressure swirl atomizer under making frequency range from 0 to 60Hz using water as a propellant. Pressure oscillation from hydraulic sources gives a strong influences on atomization and mixing processes. The ability to drive these low frequency pressure oscillations makes spray characteristics changeable. The effect of pressure perturbation and its spray characteristics showed that low injector pressure with pressure pulsation gives more significantly than high injector pressure with pressure perturbation in SMD, spray cone angle, breakup length. Moreover, this data could be used for prediction of low combustion instability getting G factor.

• Journal of Electrical Engineering and Technology
• /
• 제3권1호
• /
• pp.14-19
• /
• 2008

The paper presents an intelligent time series model to predict uncertain electricity market price in the deregulated industry environment. Since the price of electricity in a deregulated market is very volatile, it is difficult to estimate an accurate market price using historically observed data. The parameter of an intelligent time series model is obtained based on the simultaneous perturbation stochastic approximation (SPSA). The SPSA is flexible to use in high dimensional systems. Since prediction models have their modeling error, an error compensator is developed as compensation. The SPSA based intelligent model is applied to predict the electricity market price in the Pennsylvania-New Jersey-Maryland (PJM) electricity market.

• Smart Structures and Systems
• /
• 제16권3호
• /
• pp.435-457
• /
• 2015

The most widely known form of multifunctional aircraft structure is smart structures for structural health monitoring (SHM). The aim is to provide automated systems whose purposes are to identify and to characterize possible damage within structures by using a network of actuators and sensors. Unfortunately, environmental and operational variability render many of the proposed damage detection methods difficult to successfully be applied. In this paper, an original robust damage detection approach using output-only vibration data is proposed. It is based on independent component analysis and matrix perturbation analysis, where an analytical threshold is proposed to get rid of statistical assumptions usually performed in damage detection approach. The effectiveness of the proposed SHM method is demonstrated numerically using finite element simulations and experimentally through a conformal load-bearing antenna structure and composite plates instrumented with piezoelectric ceramic materials.