• Journal of KSNVE
• /
• v.9 no.1
• /
• pp.131-140
• /
• 1999

A method to reconstruct the sound field around a Korean bell is developed. The sound radiation problem is formulated based on the boundary element method by using the algorithm of the acoustical holography. Sound pressures at the hologram surface are measured and used as input data for the analysis program that was developed in this study. An error minimization scheme is presented to overcome difficulties that arise in the backward reconstruction of the BEM-based acoustical holography In the model fictitious source surfaces were also introduced to reduce the complexity stemmed from the source shape. The sound field associated with the (4.0) vibrational mode of the Korean bell was visualized and verified experimentally.

• Smart Structures and Systems
• /
• v.33 no.2
• /
• pp.119-131
• /
• 2024

Drift angle is a significant index for diagnosing post-event structures. A common way to estimate this drift response is by using modal parameters identified under natural excitations. Although the modal parameters of shear structures cannot be identified accurately in the real environment, the identification error has little impact on the estimation when measurements from several floors are used. However, the estimation accuracy falls dramatically when there is only one accelerometer. This paper describes the susceptibility of single sensor identification to modelling error and simulations that preliminarily verified this characteristic. To make a robust evaluation from measurements of one floor of shear structures based on imprecisely identified parameters, a novel scheme is devised to approximately correct the mode shapes with respect to fictitious frequencies generated with a genetic algorithm; in particular, the scheme uses constrained minimization to take both the mathematical aspect and the realistic aspect of the mode shapes into account. The algorithm was validated by using a full-scale shear building. The differences between single-sensor and multiple-sensor estimations were analyzed. It was found that, as the number of accelerometers decreases, the error rises due to insufficient data and becomes very high when there is only one sensor. Moreover, when measurements for only one floor are available, the proposed method yields more precise and appropriate mode shapes, leading to a better estimation on the drift angle of the lower floors compared with a method designed for multiple sensors. As well, it is shown that the reduction in space complexity is offset by increasing the computation complexity.

• Proceedings of the KIEE Conference
• /
• 1999.07g
• /
• pp.2932-2934
• /
• 1999

In this paper, a scheme for recognition of handwritten digits using a multilayer neural network trained with the back-propagation algorithm using generalized delta rule is proposed. The neural network is trained with hand written digit data of different writers and different styles. One of the purpose of the work with neural networks is the minimization of the mean square error(MSE) between actual output and desired one. The back-propagation algorithm is an efficient and very classical method. The back-propagation algorithm for training the weights in a multilayer net uses the steepest descent minimization procedure and the sigmoid threshold function. As an error rate is reduced, recognition rate is improved. Therefore we propose a method that is reduced an error rate.

• Journal of KIISE:Software and Applications
• /
• v.28 no.9
• /
• pp.681-687
• /
• 2001

In order to raise a class discrimination power by combining multiple classifiers under the Bayesian decision theory, the upper bound of a Bayes error rate bounded by the conditional entropy of a class variable and decision variables obtained from training data samples should be minimized. Wang and Wong proposed a tree dependence first-order approximation scheme of a high order probability distribution composed of the class and multiple feature pattern variables for minimizing the upper bound of the Bayes error rate. This paper presents an extended high order product approximation scheme dealing with higher order dependency more than the first-order tree dependence, based on the minimization of the upper bound of the Bayes error rate. Multiple recognizers for unconstrained handwritten numerals from CENPARMI were combined by the proposed approximation scheme using the Bayesian formalism, and the high recognition rates were obtained by them.

• Journal of Electrical Engineering and Technology
• /
• v.9 no.2
• /
• pp.478-488
• /
• 2014

Torque pulsation mechanism and highly nonlinear magnetic characterization of switched reluctance motors(SRM) lead to unfavorable torque ripple and limit the variety of applications in industry. In this paper, a modification method proposed for torque ripple minimization of SRM based on conventional torque sharing functions(TSF) to improve maximum speed of torque ripple-free operation considering converter limitations. Due to increasing phase inductance in outgoing phase during the commutation region, reference current tracking can be deteriorated especially when the speed increased. Moreover, phase torque production in incoming phase may not be reached to the reference value near the turn-on angle in which the incremental inductance would be dramatically decreased. Torque error for outgoing phase can cause increasing the resultant motor torque while it would be negative for incoming phase and yields reducing the motor torque. In this paper, a modification method is proposed in which phase torque tracking error for each phase under the commutation added to the other phase so that the resultant torque remained in constant level. This yields to extend constant torque region and reduce peak phase current when the speed increased. Simulation and experimental results for four phase 4 KW, 8/6 SRM validate the effectiveness of the proposed scheme.

• Journal of Electrical Engineering and Technology
• /
• v.7 no.4
• /
• pp.546-557
• /
• 2012

In Conventional Combined Vector and Direct Controls (VC-DTC) of induction motor, stator current is very rich in harmonic components. It leads to high torque ripple of induction motor in high and low speed region. To solve this problem, a control method based on the concept of fuzzy logic approach is used. The control scheme proposed uses stator current error as variable. Through the fuzzy logic controller rules, the choice of voltage space vector is optimized and then torque and speed are controlled successfully with a less ripple level in torque response, which improve the system's performance. Simulation results trough MATLAB/SIMULINK${(R)}$ software gave results that justify the claims.

• Transactions on Control, Automation and Systems Engineering
• /
• v.2 no.1
• /
• pp.56-61
• /
• 2000

Majority of industrial robots are controlled by a simple independent joint control of joint actuators rather than complex controllers based on the nonlinear dynamic model of the robot manipulator. In this independent joint control scheme, the performance of actuator control is influenced significantly by the joint disturbance torques including gravity, Coriolis and centrifugal torques, which result in the trajectory tracking error in the joint control system. The control performance of a redundant manipulator under independent joint control can be improved by minimizing this joint disturbance torque in resolving the kinematic redundancy. A 3 DOF planar robot is studied as an example, and the dynamic programming method is used to find the globally optimal joint trajectory that minimize the joint disturbance torque over the entire motion. The resulting solution is compared with the solution obtained by the conventional joint torque minimization, and it is shown that joint disturbance can be reduced using the kinematic redundancy.

• Journal of the Optical Society of Korea
• /
• v.13 no.3
• /
• pp.367-372
• /
• 2009

In this paper we propose an electronic domain timing phase selection scheme for the optical communication systems suffering from inter-symbol-interference (ISI) distortion due to chromatic dispersion (CD) or polarization mode dispersion (PMD). In the presence of CD/PMD a proper timing phase selection is important for discrete time domain equalizers, since different timing phases produce different nonlinear ISI channels of different severity. The proposed timing phase recovery scheme based on dispersion minimization (DM) practically approximates the optimal minimum mean squared error (MMSE) timing phase without training signals which reduces overall throughput substantially, especially in time-varying channels such as PMD. The simulation results show that the proposed DM timing agrees with MMSE timing phase, under proper normalization of the received signals, for various dispersion and OSNR.

• International Journal of Computer Science & Network Security
• /
• v.21 no.11
• /
• pp.312-320
• /
• 2021

Increase in computational cost and exhaustive search can lead to more complexity and computational energy. Thus, there is need for effective and efficient scheme to reduce the complexity to achieve optimal energy utilization. This will improve the energy efficiency and enhance the proficiency in terms of the resources needed to achieve convergence. This paper primarily focuses on the development of hybrid swarm intelligence scheme for reducing the computational complexity in binary optimization. In order to reduce the complexity, both artificial bee colony (ABC) and particle swarm optimization (PSO) have been employed to effectively minimize the exhaustive search and increase convergence. First, a new approach using ABC and PSO has been proposed and developed to solve the binary optimization problem. Second, the scout for good quality food sources is accomplished through the deployment of PSO in order to optimally search and explore the best source. Extensive experimental simulations conducted have demonstrate that the proposed scheme outperforms the ABC approaches for reducing complexity and energy consumption in terms of convergence, search and error minimization performance measures.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.2 no.3
• /
• pp.87-95
• /
• 2009

As information theoretic learning techniques, error entropy minimization criterion (MEE) and maximum cross correntropy criterion (MCC) have been studied in depth for supervised learning. MEE criterion leads to maximization of information potential and MCC criterion leads to maximization of cross correlation between output and input random processes. The weighted combination scheme of these two criteria, namely, minimization of Error Entropy with Fiducial points (MEEF) has been introduced and developed by many researchers. As an approach to unsupervised, blind channel equalization, we investigate the possibility of applying constant modulus error (CME) to MEE criterion and some problems of the method. Also we study on the application of CME to MEEF for blind equalization and find out that MEE-CME loses the information of the constant modulus. This leads MEE-CME and MEEF-CME not to converge or to converge slower than other algorithms dependent on the constant modulus.