• The Journal of the Acoustical Society of Korea
• /
• v.19 no.2E
• /
• pp.27-31
• /
• 2000

Evaluating log-likelihood of Gaussian mixture density is major computational burden for the keyword spotting system using continuous HMM. In this paper, we employ the bucket box intersection (BBI) algorithm to reduce the computational complexity of keyword spotting. We make some modification in implementing BBI algorithm in order to increase the discrimination ability among the keyword models. According to our keyword spotting experiments, the modified BBI algorithm reduces 50% of log-likelihood computations without performance degradation, while the original BBI algorithm under the same condition reduces only 30% of log-likelihood computations.

• Proceedings of the Computational Structural Engineering Institute Conference
• /
• 1998.04a
• /
• pp.383-389
• /
• 1998

A numerical study is conducted to examine the wave scattering at infilled trenches which may be constructed to reduce the ground-transmitted vibration. The finite element method is used for the simulation of the wave propagation in the semi-infinite region. In order to keep the computational burden manageable, the absorbing boundaries are employed. The numerical technique is validated by modeling a published problem. The results are shown to be in good agreement with the published data. The screening effectiveness of the infilled trenches is then studied for different trench dimensions and material properties.

• ETRI Journal
• /
• v.35 no.5
• /
• pp.943-946
• /
• 2013

Instead of the extended Kalman filter, the unscented Kalman filter (UKF) has been used in nonlinear systems without initial accurate state estimates over the last decade because the UKF is robust against large initial estimation errors. However, in a multirate integrated system, such as an inertial navigation system (INS)/Global Positioning System (GPS) integrated navigation system, it is difficult to implement a UKF-based navigation algorithm in a low-grade or mid-grade microcontroller, owing to a large computational burden. To overcome this problem, this letter proposes a modified UKF that has a reduced computational burden based on the basic idea that the change of probability distribution for the state variables between measurement updates is small in a multirate INS/GPS integrated navigation filter. The performance of the modified UKF is verified through numerical simulations.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.2627-2631
• /
• 2003

With the development of RAIM techniques for single failure, there has been increasing interest in the multiple failure problem. There have been many approaches to tackle the problem from various points of view. This paper approaches to two failure problem with total least squares (TLS) technique, which has rarely been addressed because TLS requires a great number of computations. In this paper, the special form of the observation matrix H, that is, one column is exactly known, is exploited so as to develop an algorithm in a sequential form, which reduces computational burden. The algorithm makes us enjoy the advantages of TLS without much computational burden. The proposed algorithm is verified through a numerical simulation.

• Transactions of the Korean Society of Automotive Engineers
• /
• v.19 no.4
• /
• pp.84-90
• /
• 2011

In active noise control filter, LMS algorithms which used for control filter, assure the convergence property, and computational burden of these algorithms are proportionate to the filter taps. The convergence speed of LMS algorithms is mainly determined by value of the convergence coefficient, so optimal selection of the value of convergence coefficient is very important. In this paper, We proposed novel adaptive fuzzy logic LMS algorithms with FIR filter structure which has better convergence speed and less computational burden than conventional LMS algorithms, for single channel active noise control with ill conditioned signal case. Computer simulations were performed to show the effectiveness of a proposed algorithms.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.32 no.1C
• /
• pp.34-42
• /
• 2007

The sphere decoder (SD) has recently been proposed to perform maximum likelihood (ML) decoding for multi-input multi-output systems. Employing a 'breadth-first' searching algorithm for closet points in a lattice, we propose a novel ML decoding scheme for multi-input multi-output systems. Simulation results show that the proposed scheme has the same bit error rate performance as the conventional ML decoders while allowing significantly lower computational burden than the SD.

• Communications for Statistical Applications and Methods
• /
• v.28 no.6
• /
• pp.673-679
• /
• 2021

The moment calculation in a truncated multivariate normal distribution is a long-standing problem in statistical computation. Recently, Kan and Robotti (2017) developed an algorithm able to calculate all orders of moment under different types of truncation. This result was implemented in an R package MomTrunc by Galarza et al. (2021); however, it is difficult to use the package in practical statistical problems because the computational burden increases exponentially as the order of the moment or the dimension of the random vector increases. Meanwhile, Lee (2021) presented an efficient numerical method in both accuracy and computational burden using Gauss-Hermit quadrature. This article introduces trunmnt implementation of Lee's work as an R package. The Package is believed to be useful for moment calculations in most practical statistical problems.

• Journal of Power Electronics
• /
• v.17 no.3
• /
• pp.674-685
• /
• 2017

The conventional model predictive direct torque control (MPDTC) method uses all of the voltage vectors available from a two level voltage source inverter for the prediction of the stator flux and stator current, which leads to a heavy computational burden. This paper proposes an improved model predictive direct torque control method. The stator flux predictive controller is obtained from an analysis of the relationship between the stator flux and the torque, which can be used to calculate the desired voltage vector based on the stator flux and torque reference. Then this method only needs to evaluate three voltage vectors in the sector of the desired voltage vector. As a result, the computational burden of the conventional MPDTC is effectively reduced. The time delay introduced by the computational time causes the stator current to oscillate around its reference. It also increases the current and torque ripples. To address this problem, a delay compensation method is adopted in this paper. Furthermore, the switching frequency of the inverter is significantly reduced by introducing the constraint of the power semiconductor switching number to the cost function of the MPDTC. Both simulation and experimental results are presented to verify the validity and feasibility of the proposed method.

• Nuclear Engineering and Technology
• /
• v.44 no.2
• /
• pp.151-160
• /
• 2012

The advantages for using Monte Carlo methods to analyze full-core reactor configurations include essentially exact representation of geometry and physical phenomena that are important for reactor analysis. But this substantial advantage comes at a substantial cost because of the computational burden, both in terms of memory demand and computational time. This paper focuses on the challenges facing full-core Monte Carlo for keff calculations and the prospects for Monte Carlo becoming a routine tool for reactor analysis.

• Journal of Mechanical Science and Technology
• /
• v.20 no.3
• /
• pp.376-381
• /
• 2006

In this paper, the quaternion, the dual Euler, and the direction cosine methods are numerically compared using a non-aerodynamic 6 degree-of-freedom rigid model at all-attitude angles of an aircraft. The dual Euler method turns out to be superior to the others in the applications because it shows better numerical accuracy, stability, and robustness in integration step sizes. The dual Euler method is affordably less efficient than the quaternion method in terms of computational cost. Numerical accuracy and stability, which allow larger integration step sizes, are more critical in modern real-time applications than computational efficiency because of today's increased computational power. If the quaternion method is required because of constraints in computation time, then a suppression mechanism should be provided for algebraic constraint errors which will eventually add computational burden.