• Structural Engineering and Mechanics
• /
• v.24 no.6
• /
• pp.709-725
• /
• 2006

In this paper the buckling and post-buckling behavior of slender bars under self-weight are studied. In order to study the post-buckling behavior of the bar, a geometrically exact formulation for the non-linear analysis of uni-directional structural elements is presented, considering arbitrary load distribution and boundary conditions. From this formulation one obtains a set of first-order coupled nonlinear equations which, together with the boundary conditions at the bar ends, form a two-point boundary value problem. This problem is solved by the simultaneous use of the Runge-Kutta integration scheme and the Newton-Raphson method. By virtue of a continuation algorithm, accurate solutions can be obtained for a variety of stability problems exhibiting either limit point or bifurcational-type buckling. Using this formulation, a detailed parametric analysis is conducted in order to study the buckling and post-buckling behavior of slender bars under self-weight, including the influence of boundary conditions on the stability and large deflection behavior of the bar. In order to evaluate the quality and accuracy of the results, an experimental analysis was conducted considering a clamped-free thin-walled metal bar. As this kind of structure presents a high index of slenderness, its answers could be affected by the introduction of conventional sensors. In this paper, an experimental methodology was developed, allowing the measurement of static or dynamic displacements without making contact with the structure, using digital image processing techniques. The proposed experimental procedure can be used to a wide class of problems involving large deflections and deformations. The experimental buckling and post-buckling behavior compared favorably with the theoretical and numerical results.

• Journal of the Korean Society of Propulsion Engineers
• /
• v.19 no.2
• /
• pp.17-28
• /
• 2015

In this paper, it is accomplished to design a control method for such an actuation system of simplified ON/OFF mechanism with actuation command limit. First of all, based on experimental data, the modeling works for nonlinear/linear actuation dynamics are performed, which are govern by PWM command as a control input. Using the linearized model, a classical PI control method is designed to satisfy the aimed control performance requirements, and a control algorithm is proposed to realize the required control performance in the effective control region through resolving the issue for the PWM command limit which reduces the control performance. Finally, through control simulations, the design method is verified and the corresponding control performance improvement is evaluated.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.39 no.1
• /
• pp.35-47
• /
• 2002

This paper proposes a self organizing RBF neural network equalizer for the equalization of digital communications. It is the most important for the equalizer using the RBF neural network to estimate the RBF centers correctly and quickly, which are the desired channel states. However, the previous RBF equalizers are not used in the actual communication system because of some drawbacks that the number of channel states has to be known in advance and many centers are necessary. Self organizing neural network equalizer proposed in this paper can implement the equalization without prior information regarding the number of channel states because it selects RBF centers among the signals that are transmitted to the equalizer by the new addition and removal criteria. Furthermore, the proposed equalizer has a merit that is able to make a equalization with fewer centers than those of prior one by the course of the training using LMS and clustering algorithm. In the linear, nonlinear and standard telephone channel, the proposed equalizer is compared with the optimal Bayesian equalizer for the BER performance, the symbol decision boundary and the number of centers. As a result of the comparison, we can confirm that the proposed equalizer has almost similar performance with the Bavesian enualizer.

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.20 no.6
• /
• pp.761-767
• /
• 2007

In centralized control system, complicated control systems including sensors, power supply and dampers should be required to satisfy the target response of large-scale structures. The practical applications of the centralized control system, however, is very difficult due to high order finite element model of structures, uncertainty of models, and limitations of the excitation system. In this study, the decentralized response-dependent MR damper of which magnetic field is automatically modulated according to the displacement or velocity transferred to the damper without any sensing and computing systems. this decentralized response-dependent MR damper are investigated according to the ranges of relative magnitude between the control force of MR damper and the story shear force of structures by nonlinear time history analysis. Finally, its performance is compared with centralized LQR algorithm which is used in general centralized control theory for a three story building structure.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.55 no.6
• /
• pp.264-273
• /
• 2006

In this paper, we propose a new architecture of Self-Organizing Fuzzy Polynomial Neural Networks (SOFPNN) by means of consecutive optimization and also discuss its comprehensive design methodology involving mechanisms of genetic optimization. The network is based on a structurally as well as parametrically optimized fuzzy polynomial neurons (FPNs) conducted with the aid of information granulation and genetic algorithms. In structurally identification of FPN, the design procedure applied in the construction of each layer of a SOFPNN deals with its structural optimization involving the selection of preferred nodes (or FPNs) with specific local characteristics and addresses specific aspects of parametric optimization. In addition, the fuzzy rules used in the networks exploit the notion of information granules defined over system's variables and formed through the process of information granulation. That is, we determine the initial location (apexes) of membership functions and initial values of polynomial function being used in the premised and consequence part of the fuzzy rules respectively. This granulation is realized with the aid of the hard c-menas clustering method (HCM). For the parametric identification, we obtained the effective model that the axes of MFs are identified by GA to reflect characteristic of given data. Especially, the genetically dynamic search method is introduced in the identification of parameter. It helps lead to rapidly optimal convergence over a limited region or a boundary condition. To evaluate the performance of the proposed model, the model is experimented with using two time series data(gas furnace process, nonlinear system data, and NOx process data).

• The Journal of the Korea institute of electronic communication sciences
• /
• v.9 no.11
• /
• pp.1221-1226
• /
• 2014

In an effort to improve the efficiency of the speech recognition, we investigate the effect of the number of phoneme clusters. For this purpose, codebooks of varied number of phoneme clusters are prepared by modified k-means clustering algorithm. The subsequent processing is fuzzy vector quantization (FVQ) and hidden Markov model (HMM) for speech recognition test. The result shows that there are two distinct regimes. For large number of phoneme clusters, the recognition performance is roughly independent of it. For small number of phoneme clusters, however, the recognition error rate increases nonlinearly as it is decreased. From numerical calculation, it is found that this nonlinear regime might be modeled by a power law function. The result also shows that about 166 phoneme clusters would be the optimal number for recognition of 300 isolated words. This amounts to roughly 3 variations per phoneme.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.8 no.11
• /
• pp.1749-1754
• /
• 2013

Binary sequences of period $2^n-1$ are widely used in many areas of engineering and sciences. Some well-known applications include coding theory, code-division multiple-access (CDMA) communications, and stream cipher systems. In this paper we analyze different solutions to $x^5+bx^3+b^{2^m}x^2+1=0$ over $GF(2^n)$. The number of different solutions determines frequencies of cross-correlations of nonlinear binary sequences generated by $d=3{\cdot}2^m-2$, n=2m, m=4k($k{\geq}2$). Also we give an algorithm for determination of number of different solutions to the equation.

• Proceedings of the Korean Institute of Navigation and Port Research Conference
• /
• v.2
• /
• pp.255-260
• /
• 2006

Safety-critical navigation systems have to provide 'reliable' position solutions, i.e., they must detect and exclude measurement or system faults and estimate the uncertainty of the solution. To obtain more accurate and reliable navigation systems, various filtering methods have been employed to reduce measurement noise level, or integrate sensors, such as global navigation satellite system/inertial navigation system (GNSS/INS) integration. Recently, particle filters have attracted attention, because they can deal with nonlinear/non-Gaussian systems. In most GNSS applications, the GNSS measurement noise is assumed to follow a Gaussian distribution, but this is not true. Therefore, we have proposed a fault detection and exclusion method using particle filters assuming non-Gaussian measurement noise. The performance of our method was contrasted with that of conventional Kalman filter methods with an assumed Gaussian noise. Since the Kalman filters presume that measurement noise follows a Gaussian distribution, they used an overbounded standard deviation to represent the measurement noise distribution, and since the overbound standard deviations were too conservative compared to the actual distributions, this degraded the integrity-monitoring performance of the filters. A simulation was performed to show the improvement in performance of our proposed particle filter method by not using the sigma overbounding. The results show that our method could detect smaller measurement biases and reduced the protection level by 30% versus the Kalman filter method based on an overbound sigma, which motivates us to use an actual noise model instead of the overbounding or improve the overbounding methods.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.17 no.1
• /
• pp.15-20
• /
• 2016

The existing multiple modulus-based self-recovering equalization type has not been applied to initial equalization. Instead, it was used for steady-state performance improvement. In this paper, for the self-recovering equalization type that considers the multiple modulus as a desired response, the initial convergence performance was improved by extending the dynamics of the errors using error boosting and their characteristics were analyzed. Error boosting in the proposed method was carried out in proportion to a symbol decision for the equalizer output. Furthermore, having the initial convergence capability by extending the dynamics of errors, it showed excellent performance in the initial convergence rate and steady-state error level. In particular, the proposed method can be applied to the entire process of equalization through a single algorithm; the existing methods of switching over or the selection of other operation modes, such as concurrent operating with other algorithms, are not necessary. The usefulness of the proposed method was verified by simulations performed under the channel conditions with multipath propagation and additional noise, and for performance analysis of self-recovering equalization for high-order signal constellations.

• Journal of the Institute of Convergence Signal Processing
• /
• v.18 no.2
• /
• pp.43-48
• /
• 2017

The underwater transients signals contain the characteristics of complexity, time varying, nonlinear, and short duration. So it is very hard to model for these signals with reference patterns. In this paper we separate the whole length of signals into some short duration of constant length with overlapping frame by frame. The 20th LPC(Linear Predictive Coding) coefficients are extracted from the original signals using Durbin algorithm and applied to neural network. The 65% of whole signals were learned and 35% of the signals were tested in the neural network with two hidden layers. The types of the whales for sound classification are Blue whale, Dulsae whale, Gray whale, Humpback whale, Minke whale, and Northern Right whale. Finally, we could obtain more than 83% of classification rate from the test signals.