• Korean Journal of Optics and Photonics
• /
• v.14 no.1
• /
• pp.72-79
• /
• 2003

We proposed a noble algorithm using DOP (degree of polarization) not only as feedback signal but alto as feed-forward signal for the compensation of higher-order PMD effect. In the proposed algorithm, DOP after the first-order PMD compensation is considered as the indicator of the amount of residual higher-order PMD. This algorithm has the merit that DOF (degree of freedom) of the system can be limited to the level of the first-order PMD compensation system. Owing to the limited DOF, the reliability of the system can be enhanced and the complexity of the implementation can be degraded. For the analysis of the algorithm, we simulated the 10Gbps NRZ transmission system and obtained the result that the system outage probability can be reduced as much as three times with respect to the only first-order PMD compensation case.

• Asian Journal of Innovation and Policy
• /
• v.1 no.1
• /
• pp.92-115
• /
• 2012

The problem Japanese universities face is exactly the same as that of German universities: no international recognition in world rankings of universities despite their high levels of postwar economic and technological developments. This was indeed one reason why world-class Japanese firms, such as Toyota and Sony, have avoided working closely with Japanese universities for R&D partnership and new technology commercialization. To resolve this problem, the Japanese government has continuously implemented aggressive policies of the internationalization, privatization, liberalization, and privatization of universities since the onset of the economic recession in 1989 in order to revitalize the Japanese economy through radical innovation projects between universities and firms. National projects of developing medical robots for Japan's ageing society are some of the ambitious examples that emphasize feed-forward learning in innovation. However, this paper argues that none of these programs of fostering university-firm alliances toward feed-forward learning has been successful in promoting the world ranking of Japanese universities, although they showed potentials of reinforcing their conventional strength of introducing $kaizen$ through feedback learning of tacit knowledge. It is therefore argued in this paper that Japanese universities and firms should focus on feedback learning as a way to motivate firm-university R&D alliances.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.15 no.3
• /
• pp.226-234
• /
• 2010

This paper proposes an improved feed-forward controller that calculates the gain value by estimating the changed boost inductance in practical operating condition of transformer. The boost inductance is estimated by the measurement of input current and voltage. The estimated boost inductance is optimized by the least square method. The proposed feed-forward controller can be achieved the robust control through the gain value calculating the estimated boost inductance despite of the changed condition of transformer and can minimize the interference phenomenon by reducing the harmonics of input current. The validity of proposed technique is verified through the simulation and experiment.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.19 no.5
• /
• pp.143-149
• /
• 2009

The hash function ARIRANG is one of the 1st round SHA-3 candidates. In this paper, we present preimage attacks on ARIRANG with step-reduced compression functions. Our attack finds a preimage of the 33-step OFF(Original FeedForward1) variants of ARIRANG, and a preimage of the 31-step MFF(Middle FeedForward1) variants of ARIRANG. Its time complexity is about $2^{241}$ for ARIRANG-256 and $2^{481}$ for ARIRANG-512, respectively.

• Journal of Digital Contents Society
• /
• v.18 no.1
• /
• pp.183-190
• /
• 2017

This paper proposes an audio event detection method using Deep Neural Networks (DNN). The proposed method applies Feed Forward Neural Network (FFNN) to generate output probabilities of twenty audio events for each frame. Mel scale filter bank (FBANK) features are extracted from each frame, and its five consecutive frames are combined as one vector which is the input feature of the FFNN. The output layer of FFNN produces audio event probabilities for each input feature vector. More than five consecutive frames of which event probability exceeds threshold are detected as an audio event. An audio event continues until the event is detected within one second. The proposed method achieves as 71.8% accuracy for 20 classes of the UrbanSound8K and the BBC Sound FX dataset.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.22 no.12
• /
• pp.29-35
• /
• 2008

In this paper, The author present a load current feed-forward compensator by method that improve voltage controller of Step-down Chopper to get stable output voltage to sudden change of load current. To confirm the characteristics of a presented load current feed-forward compensator compared each transfer function of whole system that load current feed-forward compensator is added with transfer function of whole system that existent voltage controller is included using Mason gains formula in Root locus and Bode diagram. As a result the pole of system is improved, extreme point of the wave and system improves, and size of peak value and phase margin of break frequency in resonance frequency confirmed that is good. Therefore, presented control technique could confirm that reduce influence by perturbation and improves stationary state and dynamic characteristics in output of Step-down Chopper.

• Journal of Magnetics
• /
• v.22 no.1
• /
• pp.34-42
• /
• 2017

Material defects in the Steam Generator Tubes (SGT) of sodium cooled fast breeder reactor (PFBR) can lead to leakage of water into sodium. The water and sodium reaction will lead to major accidents. Therefore, the examination of steam generator tubes for the early detection of defects is an important requirement for safety and economic considerations. In this work, the Magnetic Flux Leakage (MFL) based Non Destructive Testing (NDT) technique is used to perform the defect detection process. The rectangular notch defects on the outer surface of steam generator tubes are modeled using COMSOL multiphysics 4.3a software. The obtained MFL images are de-noised to improve the integrity of flaw related information. Grey Level Co-occurrence Matrix (GLCM) features are extracted from MFL images and taken as input parameter to train the neural network. A comparative study on characterization have been carried out using feed-forward back propagation (FFBP) and cascade-forward back propagation (CFBP) algorithms. The results of both algorithms are evaluated with Mean Square Error (MSE) as a prediction performance measure. The average percentage error for length, depth and width are also computed. The result shows that the feed-forward back propagation network model performs better in characterizing the defects.

• Journal of Advanced Marine Engineering and Technology
• /
• v.21 no.5
• /
• pp.503-511
• /
• 1997

In 3 phase induction motor control system, the speed control using the load torque observer becomes robust against disturbances by means of a feed-forward control of the estimated load torque component. In case of variation of inertia moment, the estimated load torque has error because the observer uses the nominal inertia to estimate the load torque. And so, it is difficult to obtain good speed control characteristics. This paper has two study target strategy. First, we executes feed-forward control with the load torque observer when motor inertia has nominal value and compare it with conventional PI con¬trol. The second strategy estimates inertia moment error using the load torque observer when inertia moment change. The proposed two strategy is confirmed through the computer simulations and the experimental implementations by TMS320C31 microprocessor.

• Transactions of the Korean Society for Noise and Vibration Engineering
• /
• v.19 no.5
• /
• pp.516-522
• /
• 2009

This paper proposes a new haptic cue vehicle accelerator pedal device using magnetorheological(MR) brake. As a first step, an MR fluid-based haptic cue device is devised to be capable of rotary motion of accelerator pedal. Under consideration of spatial limitation, design parameters are optimally determined to maximize control torque using finite element method. The proposed haptic cue device is then manufactured and integrated with accelerator pedal. Its field-dependant torque is experimentally evaluated. Vehicle system emulating gear shifting and engine speed is constructed in virtual environment and communicated with the haptic cue device. Haptic cue algorithm using the feed-forward control algorithm is formulated to achieve optimal gear shifting in driving. Control performances are experimentally evaluated via feed-forward control strategy and presented in time domain.

• Proceedings of the Korean Society for Technology of Plasticity Conference
• /
• 2003.05a
• /
• pp.168-171
• /
• 2003

Generally RF AGC (Roll Force Automatic Gauge Control) controls the roll gap using the variation of rolling force caused by the roll eccentricity and the entry thickness of material, but RE AGC takes the bad effect of the roll eccentricity. The Feed-forward (FF) AGC method, which controls the next stand roll gap by the estimation of the thickness variation due to skid mark is needed to supplement the shortage of RF AGC. In this paper, an adaptive filtering method which takes account of the kind of material, the final objective thickness and the rolling speed is proposed to predict skid mark thickness variation. In addition, an improved estimation method of control point using a speedometer and looper angle is suggested. Via on line test, the performance improvement of the suggested FF AGC method is verified.