• International journal of advanced smart convergence
• /
• v.9 no.2
• /
• pp.20-27
• /
• 2020

Sound event detection in real-world environments suffers from the interference of non-stationary and time-varying noise. This paper presents an adaptive noise reduction method for sound event detection based on non-negative matrix factorization (NMF). In this paper, we proposed a deep learning model that integrates Convolution Neural Network (CNN) with Non-Negative Matrix Factorization (NMF). To improve the separation quality of the NMF, it includes noise update technique that learns and adapts the characteristics of the current noise in real time. The noise update technique analyzes the sparsity and activity of the noise bias at the present time and decides the update training based on the noise candidate group obtained every frame in the previous noise reduction stage. Noise bias ranks selected as candidates for update training are updated in real time with discrimination NMF training. This NMF was applied to CNN and Hidden Markov Model(HMM) to achieve improvement for performance of sound event detection. Since CNN has a more obvious performance improvement effect, it can be widely used in sound source based CNN algorithm.

• 제어로봇시스템학회:학술대회논문집
• /
• 1997.10a
• /
• pp.1489-1492
• /
• 1997

To increase reliability and performance of an IMS(Intelligent Manufacturing System), fault tolerant control based on an accurate fault diagnosis is needed. In this paper, robot FDI(fault detection and identification) is proposed for IMS where the robot is controlled with state estimates of a nonlinear filter using a mathematical robot model. The Chi-square distribution is applied fault detection and fault size is estimated by a proposed bias filter. Performance of the proposed algorithm is tested by simulation for studies.

• Journal of Institute of Control, Robotics and Systems
• /
• v.4 no.5
• /
• pp.666-673
• /
• 1998

To increase reliability and performance of an IMS(Intelligent Manufacturing System), fault tolerant control based on an accurate fault diagnosis is needed. In this paper, robot FDI(fault detection and identification) is proposed for IMS where the robot is controlled with state estimates of a nonlinear filter using a mathematical robot model. The Chi-square test and GLR(General likelihood ratio) test are applied for fault detection and fault size is estimated by a proposed bias filter. Performance of the proposed algorithm is tested by simulation for studies.

• Journal of Haehwa Medicine
• /
• v.24 no.1
• /
• pp.25-36
• /
• 2015

Objective : This study aims to evaluate a risk of bias by Risk of Bias tool and RoBANS(Risk of Bias Assessment tool for Non-randomized Study) tool for clinical trial papers proving treatment effect of herbs to alopecia and provides the newest reason of effectiveness of herbs to alopecia. Methos : Data were collected through electronic database including NDSL, KISS, KMBASE, Koreantk, OASIS, KoreaMed, KISTI, Pubmd, Cochrane CENTRAL and CINAHL. Two experts in Oriental Medince assessed risk of bias of randomized controlled trials by Cochrane group's Risk of Bias tool and non-randomized controlled trials by RoBANS tool after searching, reviewing and selecting papers. Results : Total number of selected trials is 20 including 4 randomized controlled trials, 13 non-randomized controlled trials and 3 case reports. This study evaluates the risk of bias of 17 papers including 4 randomized controlled trials and 13 non-randomized controlled trials except 3 case reports by risk of bias tool and RoBANS tool. All papers of randomized controlled trials are evaluated unclear for random sequence generation and allocation concealment as there are no word on them. And all papers of non-randomized controlled trials are evaluated unclear for blinding of outcome assessments and relatively low for others. Conclusion : We must try to specify concretely methods of allocation concealment after planning and practicing it for reducing a selection bias in randomized controlled trials. Also report a reason of missing value and blinding outcome assessments. And we have to agonize and mention methods of blinding of researchers for reducing a detection bias in non-randomized controlled trials.

• Journal of the Society of Naval Architects of Korea
• /
• v.30 no.4
• /
• pp.8-16
• /
• 1993

Auto-Pilot System uses heading angle information via the position sensor and the rudder device to control the ship direction. Most of the control logics are composed of the state estimation and control algorithms assuming that the measurement device and the actuator have no fault except the measurement noise. But such asumptions could bring the danger in real situation. For example, if the heading angle measuring device is out of order the control action based on those false position information could bring serious safety problem. In this study, the control system including improved method for processing the position information is applied to the Auto-Pilot System. To show the difference between general state estimator and F.D.F., BJDFs for the sensor and the actuator failure detection are designed and the performance are tested. And it is shown that bias error in sensor could be detected by state-augmented estimator. So the residual confined in the 2-dim in the presence of the sensor failure could be unidirectional in output space and bias sensor error is much easier to be detected.

• Current Optics and Photonics
• /
• v.8 no.5
• /
• pp.456-462
• /
• 2024

We propose a novel fade-free optical fiber interferometric vibration sensor using a simple setup with a 90° optical hybrid. The interferometer consists of all-optical components without the phase modulators and complex demodulation processes that were previously used to compensate for signal fading induced by phase bias change. Fade-free output was successfully obtained by in-phase and quadrature detection with a π/2 phase shifting scheme. Theoretical analysis and measurement results showed that the proposed interferometric vibration sensor operates independently of the phase bias state of interfering waves.

• Journal of Ocean Engineering and Technology
• /
• v.31 no.2
• /
• pp.177-182
• /
• 2017

This paper describes an unscented Kalman filter approach to estimate the bias in magnetic field measurements. A microelectromechanical systems attitude heading reference system (MEMS AHRS) was used to measure the magnetic field, together with the acceleration and angular rate. A magnetic field is usually used for yaw detection, while the acceleration serves to detect the roll and pitch. Magnetic field measurements are vulnerable to distortion due to hard-iron effect and soft-iron effect. The bias in the measurement accounts for the hard-iron effect, and this paper focuses on an approach to estimate this bias. The proposed method is compared with other methods through experiments that implement the navigation of an underwater robot using an AHRS and Doppler velocity log. The results verify that the compensation of the bias by the proposed method improves the navigation performance more than or comparable to the compensation by other methods.

• 제어로봇시스템학회:학술대회논문집
• /
• 1991.10a
• /
• pp.71-75
• /
• 1991

This paper deals with designing a real-time fault and accommodation system. The LQG controller is adopted in the normal state and the output of LQG controller is corrected using Separated Bias Estimator in the faulty state. The proposed scheme has been applied to the two-tank control system and showed satisfactory performance.

• Journal of Korean Society for Atmospheric Environment
• /
• v.21 no.2
• /
• pp.269-274
• /
• 2005

In this work, analytical bias arising from the gas chromatographic determination of sulfur compounds was evaluated by the application of direct loop injection method to the GC/PFPD detection of four sulfur compounds including H$_{2}$S, CH$_{3}$SH, DMS, and DMDS. For the proper evaluation of analytical uncertainties involved in GC calibration, we employed two comparative techniques of calibration at fxed concentration injection (CFCI) vs calibration at fixed volume injection (CFVI) method. The results of our study indicate that CFCI method exhibits very poor sensitivity due to the matrix effect with varying injection volumes. On the other hand, as CFVI method overcomes such limitation, it can be used to obtain very accurate quantification of S compounds at their high concentration levels above a few to a few tens ppb.

• Journal of Hydrospace Technology
• /
• v.1 no.1
• /
• pp.75-88
• /
• 1995

Auto-Pilot System uses heading angle information via the position sensor and the rudder device to control the ship's direction. Most of the control logics are composed of the state estimation and control algorithms assuming that the measurement device and the actuator have no fault except the measurement noise. But such asumptions could bring the danger in real situation. For example, if the heading angle measuring device is out of order the control action based on those false position information could bring serious safety problem. In this study, the control system including improved method for processing the position information is applied to the Auto-Pilot System. To show the difference between general state estimator and F.D.F., BJDFs for the sensor and the actuator failure detection are designed and the performance are tested. And it is shown that bias error in sensor could be detected by state-augmented estimator. So the residual confined in the 2-dimension in the presence of the sensor failure could be unidirectional in output space and bias sensor error is much easier to be detected.