• ETRI Journal
• /
• v.32 no.5
• /
• pp.795-800
• /
• 2010

Embedding a large vocabulary speech recognition system in mobile devices requires a reduced acoustic model obtained by eliminating redundant model parameters. In conventional optimization methods based on the minimum description length (MDL) criterion, a binary Gaussian tree is built at each state of a hidden Markov model by iteratively finding and merging similar mixture components. An optimal subset of the tree nodes is then selected to generate a downsized acoustic model. To obtain a better binary Gaussian tree by improving the process of finding the most similar Gaussian components, this paper proposes a new distance measure that exploits the difference in likelihood values for cases before and after two components are combined. The mixture weight of Gaussian components is also introduced in the component merging step. Experimental results show that the proposed method outperforms MDL-based optimization using either a Kullback-Leibler (KL) divergence or weighted KL divergence measure. The proposed method could also reduce the acoustic model size by 50% with less than a 1.5% increase in error rate compared to a baseline system.

• Proceedings of the KSME Conference
• /
• 2007.05a
• /
• pp.538-543
• /
• 2007

A procedure is proposed to generate optimal grid with minimal user intervention while keeping a prescribed level of accuracy, using an adaptive X-FEM and applied to shape optimization. In spite of various advantages of X-FEM, however, there are several obstacles for practical applications. Because of using a uniform background mesh and additional degree of freedoms for enrichment, an X-FEM is usually computationally more expensive than traditional finite element method. Furthermore, there are often accuracy problems. For an automatic procedure of optimal mesh generation, an h-adaptive scheme and a posteriori error estimation obtained by a post-processing process are utilized. The procedure is shown by 2-D shape optimization examples.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.61 no.4
• /
• pp.650-655
• /
• 2012

This paper proposed PSO-Fuzzy controller design method. We could improve the learning performance of fuzzy controller by using PSO algorithm, which had recently showed its robust of performance while solving various difficult optimization problems. In other words, our aim was to forward the controller is performance by deciding fuzzy model structure that had good performance on optimization of the controller, based on PSO. During a simulation, we could see whether the mobile robot could convergence on the final goal or not, and also see the error, and through this process, we found out that this controller is more robust than the conventional controller.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2008.04a
• /
• pp.202-205
• /
• 2008

This paper presents an optimization of permanent magnet (PM) in a brushless dc (BLDC) motor using the response surface methodology (RSM). Size and angle of the PM are optimized to minimize the cogging torque, while reducing the magnitude of harmonic at a dominant frequency and maintaining the operating torque. A fitted RS model is constructed by verifying the high reliability of the total variation and the variation of estimated error. The optimized design is validated by carrying out the reanalysis and comparing to the initial model using the nonlinear transient finite element analysis.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.12 no.3
• /
• pp.272-281
• /
• 2009

In this paper, three-dimensional optimal evasive maneuver patterns for air-to-surface attack missiles against proportionally navigated anti-air defense missiles were investigated. An interception error of the defense missile is produced by an evasive maneuver of the attack missile. It is assumed that the defense missiles are continuously launched during the flight of attack missile. The performance index to be minimized is then defined as the negative square integral of the interception errors. The direct parameter optimization technique based on SQP and a co-evolution method based on the augmented Lagrangian formulation are adopted to get the attack missile's optimal evasive maneuver patterns. The overall shape of the resultant optimal evasive maneuver is represented as a deformed barrel-roll.

• Journal of Advanced Marine Engineering and Technology
• /
• v.31 no.1
• /
• pp.58-66
• /
• 2007

Genetic Algorithms (GA) have some difficulties in practical applications because of too many function evaluations. To overcome these limitations, an approximated modeling method such as Response Surface Modeling(RSM) is coupled to GAs. Original RSM method predicts linear or convex problems well but it is not good for highly nonlinear problems cause of the average effect of the least square method(LSM). So the locally approximated methods. so called as moving least squares method(MLSM) have been used to reduce the error of LSM. In this study, the efficient evolutionary GAs tightly coupled with RSM with MLSM are constructed and then a 2-dimensional inviscid airfoil shape optimization is performed to show its efficiency.

• Structural Engineering and Mechanics
• /
• v.51 no.2
• /
• pp.299-313
• /
• 2014

In the first part of this paper, the influences of some of crack parameters on natural frequencies of a cracked cantilever Functionally Graded Beam (FGB) are studied. A cantilever beam is modeled using Finite Element Method (FEM) and its natural frequencies are obtained for different conditions of cracks. Then effect of variation of depth and location of cracks on natural frequencies of FGB with single and multiple cracks are investigated. In the second part, two Multi-Layer Feed Forward (MLFF) Artificial Neural Networks (ANNs) are designed for prediction of FGB's Cracks' location and depth. Particle Swarm Optimization (PSO) and Back-Error Propagation (BEP) algorithms are applied for training ANNs. The accuracy of two training methods' results are investigated.

• Proceedings of the Korean Society for Noise and Vibration Engineering Conference
• /
• 2001.11a
• /
• pp.52-57
• /
• 2001

In this paper, the optimum designs of air-bearing surface(ABS) are achieved effectively by using reduced basis concept which can reduce the number of design variables although the design space is distended. Generally, the optimization method is more effective than the trial and error. However, the efficiency of the former is largely dependent on the number of the design variables. In order to reduce the number of design variables and increase the efficiency, reduced basis concept is applied. We can define the desired design as a linear combination of basis designs using this concept. From this optimization method with reduced basis concept, we easily obtain the optimum designs of ABS whose target flying heights are 25, 20, 15 nm.

• Proceedings of the KSME Conference
• /
• 2008.11a
• /
• pp.822-826
• /
• 2008

A linkage mechanism is a device to convert an input motion into a desired output motion. Traditional linkage mechanism designs are based on trial and error approaches so that size or shape changes of an original mechanism often result in improper results. In order to resolve these problems, an improved automatic mechanism synthesis method that determines the linkage type and dimensions by using an optimization method during the synthesis process has been proposed. For the synthesis, a planar linkage is modeled as a set of rigid blocks connected by zero-length translational springs with variable stiffness. In this study, the sizes of rigid blocks were also treated as design variables for more general linkage synthesis. The values of spring stiffness and the size of rigid block yielding a desired output motion at the end-effecter are found by using an optimization method.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.24 no.7 s.178
• /
• pp.1753-1762
• /
• 2000

An inverse method is presented to obtain material's flow properties by using small punch test. This procedure employs, as the objective function of inverse analysis, the balance of measured load-di splacement response and calculated one during deformation. In order to guarantee convergence to global minimum, simulated annealing method was adopted to optimize the current objective function. In addition, artificial neural network was used to predict the load-displacement response under given material parameters which is the most time consuming and limits applications of global optimization methods to these kinds of problems. By implementing the simulated annealing for optimization along with calculating load-displacement curve by neural network, material parameters were identified irrespective of initial values within very short time for simulated test data. We also tested the present method for error-containing experimental data and showed that the flow properties of material were well predicted.