• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.41 no.3
• /
• pp.173-184
• /
• 2013

A multi-level design optimization framework for aerodynamic design of rotary wing such as propeller and helicopter rotor blades is presented in this study. Strategy of the proposed framework is to enhance aerodynamic performance by sequentially applying the planform and sectional design optimization. In the first level of a planform design, we used a genetic algorithm and blade element momentum theory (BEMT) based on two-dimensional aerodynamic database to find optimal planform variables. After an initial planform design, local flow conditions of blade sections are analyzed using high-fidelity CFD methods. During the next level, a sectional design optimization is conducted using two dimensional Navier-Stokes analysis and a gradient based optimization algorithm. When optimal airfoil shape is determined at the several spanwise locations, a planform design is performed again. Through this iterative design process, not only an optimal flow condition but also an optimal shape of an EAV propeller blade is obtained. To validate the optimized propeller-blade design, it is tested in wind-tunnel facility with different flow conditions. An efficiency, which is slightly less than the expected improvement of 7% predicted by our proposed design framework but is still satisfactory to enhance the aerodynamic performance of EAV system.

• Journal of Intelligence and Information Systems
• /
• v.20 no.4
• /
• pp.43-58
• /
• 2014

As the use of trading systems has increased rapidly, many researchers have become interested in developing effective stock market prediction models using artificial intelligence techniques. Stock market prediction involves multifaceted interactions between market-controlling factors and unknown random processes. A successful stock prediction model achieves the most accurate result from minimum input data with the least complex model. In this research, we develop a combination model of ${\pi}$-fuzzy logic and support vector machine (SVM) models, using a genetic algorithm to optimize the parameters of the SVM and ${\pi}$-fuzzy functions, as well as feature subset selection to improve the performance of stock market prediction. To evaluate the performance of our proposed model, we compare the performance of our model to other comparative models, including the logistic regression, multiple discriminant analysis, classification and regression tree, artificial neural network, SVM, and fuzzy SVM models, with the same data. The results show that our model outperforms all other comparative models in prediction accuracy as well as return on investment.

• Transactions of the Korean Society of Mechanical Engineers B
• /
• v.40 no.9
• /
• pp.597-604
• /
• 2016

A parallel algorithm of bi-conjugate gradient method was developed based on CUDA for parallel computation of the incompressible Navier-Stokes equations. The governing equations were discretized using splitting P2P1 finite element method. Asymmetric stenotic flow problem was solved to validate the proposed algorithm, and then the parallel performance of the GPU was examined by measuring the elapsed times. Further, the GPU performance for sparse matrix-vector multiplication was also investigated with a matrix of fluid-structure interaction problem. A kernel was generated to simultaneously compute the inner product of each row of sparse matrix and a vector. In addition, the kernel was optimized to improve the performance by using both parallel reduction and memory coalescing. In the kernel construction, the effect of warp on the parallel performance of the present CUDA was also examined. The present GPU computation was more than 7 times faster than the single CPU by double precision.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.15 no.3
• /
• pp.357-373
• /
• 2013

In this paper, a back analysis program for analyzing the behavior of tunnel-ground system and evaluating the material properties and tunnel design parameters was developed. This program was designed to be able to implement the back analysis of underground structure by combination of using FLAC and optimized algorithm as direct method. In particular, Rosenbrock method which is able to do direct search without obtaining differential coefficient was adopted for the back analysis algorithm among optimization methods. This back analysis program was applied to the site to evaluate the design parameters. The back analysis was carried out using field measurement results from 5 sites. In the course of back analysis, nonlinear regression analysis was carried out to identify the optimum function of the measured ground displacement. Exponential function and fractional function were used for the regression analysis and total displacement calculated by optimum function was used as the back analysis input data. As a result, displacement recalculated through the back analysis using measured displacement of the structure showed 4.5% of error factor comparing to the measured data. Hence, the program developed in this study proved to be effectively applicable to tunnel analysis.

• Journal of Korean Tunnelling and Underground Space Association
• /
• v.20 no.1
• /
• pp.211-224
• /
• 2018

It is important to design tunnel support system considering the various loads acting on the tunnel because they have a direct impact on the stability of tunnels. In Korea, standardized support patterns are defined based on the rock mass classification system depending on the project, and it is stated that it should be modified appropriately considering the behavior of tunnel during construction. In this study, the tunnel support pattern optimization method is suggested based on the convergence-confinement method, earth pressure, axial force of rock bolt, and moment acting on the shotcrete. The length and spacing of the rock bolts and the thickness of the shotcrete were optimized by using the differential evolution algorithm (DEA) and the results were compared to the standard support pattern III for railway tunnel. Rock bolt length can be reduced and the installation interval can be widened for shallow tunnel. As the depth of tunnel increases, the thickness of shotcrete increases linearly. Therefore, the thickness of shotcrete should be thicker than the standard support pattern as the depth of tunnel increases to secure the stability of tunnel.

• The KIPS Transactions:PartC
• /
• v.14C no.7
• /
• pp.583-588
• /
• 2007

It is easy to implement a temporary network with a mobile ad-hoc network in which mobile nodes have without using a infrastructure network. They depend on their limited power. Recently, it is a hot issue to save the energy in a mobile ad-hoc network because a mobile nodes have a limited energy. Research of IEEE 802.11 PSM was proposed in a single hop ad-hoc assumption. If IEEE 802.11 PSM is applied to multi hop ad-hoc network, non-optimal paths will be generated by the mobile nodes which didn't receive a message of routing request. Non-optimal paths increase not only a network latency but also energy consumption of mobile nodes. Reconfiguring algorithm of non-optimal paths caused by the mobile nodes which didn't receive a message of routing request is proposed in this paper. A mobile node can overhear the data in his range. A wireless medium is shared by all mobile nodes using the same bandwidth. All mobile nodes lookout the non-optimal paths with these properties of a medium, if non-optimal path is generated, optimal reconfiguring will be accomplished by modifying routing table of itself or sending a request message of routing update to nearby nodes. By reconfiguring the non-optimal paths to optimized ones, network latency and energy consumption was decreased. It is confirmed to ignore the overhead caused by a algorithm presented in this paper through the result of the simulation.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.39C no.4
• /
• pp.351-361
• /
• 2014

Compared to a terrestrial communication, the high BER(Bit Error Ratio) and low channel bandwidth are the major factor of throughput degradation due to characteristics of underwater channel. Therefore, a MAC protocol must be designed to solve this problem in UWASNs(Underwater Acoustic Sensor Networks). MAC protocols for UWASNs can be classified into two major types according to the contention scheme(Contention-free scheme and Contention-based scheme). In large scale of sensor networks, a Contention-based scheme is commonly used due to time-synchronize problem of Contention-free scheme. In the contention-based scheme, Each node contends with neighbor nodes to access network channel by using Back-off algorithm. But a Slot-Time of Back-off algorithm has long delay times which are cause of decrease network throughput. In this paper, we propose a new scheme to solve this problem. The proposed scheme uses variable Slot-Time instead of fixed Slot-Time. Each node measures propagation delay from neighbors which are used by Slot-time. Therefore, Slot-Times of each node are optimized by considering node deployment. Consequently, the wasted-time for Back-off is reduced and network throughput is improved. A new mac protocol performance in throughput and delay is assessed through NS3 and compared with existing MAC protocol(MACA-U). Finally, it was proved that the MAC protocol using the proposed scheme has better performance than existing MAC protocol as a result of comparison.

• Journal of the Korean Institute of Telematics and Electronics C
• /
• v.36C no.12
• /
• pp.78-89
• /
• 1999

Previous wavelet network structures are determined by considering the relationship between wavelet windows distribution of training patterns that are transformed into time-frequency space. Because it is separated two algorithms that determines wavelet network structure and that modifies parameters of network, learning process that minimizes output error of network is executed after the network structure is determined. But this method has some weakness that training patterns must be transformed into time-frequency space by additional preprocessing and the network structure should be fixed during learning process. In this paper, we propose a new constructing method for wavelet network structure by using differences between the output and the desired response without preprocessing. Because the algorithm perform network construction and error minimizing process simultaneously, it can determine the number of hidden nodes adaptively as with the complexity of problems. In addition, the network structure is optimized by inserting new hidden nodes in the area that has maximum error and extracting hidden nodes that has no effect to the output of network. This algorithm has no constraint condition that all training patterns must be known, because it removes preprocessing procedure for training patterns and it can be applied effectively to systems that has time varying outputs.

• International Journal of Aeronautical and Space Sciences
• /
• v.8 no.1
• /
• pp.10-20
• /
• 2007

For spacecraft attitude control, reaction wheel (RW) steering laws with more than three wheels for three-axis attitude control can be derived by using a control allocation (CA) approach.1-2 The CA technique deals with a problem of distributing a given control demand to available sets of actuators.3-4 There are many references for CA with applications to aerospace systems. For spacecraft, the control torque command for three body-fixed reference frames can be constructed by a combination of multiple wheels, usually four-wheel pyramid sets. Multi-wheel configurations can be exploited to satisfy a body-axis control torque requirement while satisfying objectives such as minimum control energy.1-2 In general, the reaction wheel steering laws determine required torque command for each wheel in the form of matrix pseudo-inverse. In general, the attitude control command is generated in the form of a feedback control. The spacecraft body angular rate measured by gyros is used to estimate angular displacement also.⁵ Combination of the body angular rate and attitude parameters such as quaternion and MRPs(Modified Rodrigues Parameters) is typically used in synthesizing the control command which should be produced by RWs.¹ The attitude sensor signals are usually corrupted by noise; gyros tend to contain errors such as drift and random noise. The attitude determination system can estimate such errors, and provide best true signals for feedback control.⁶ Even if the attitude determination system, for instance, sophisticated algorithm such as the EKF(Extended Kalman Filter) algorithm⁶, can eliminate the errors efficiently, it is quite probable that the control command still contains noise sources. The noise and/or other high frequency components in the control command would cause the wheel speed to change in an undesirable manner. The closed-loop system, governed by the feedback control law, is also directly affected by the noise due to imperfect sensor characteristics. The noise components in the sensor signal should be mitigated so that the control command is isolated from the noise effect. This can be done by adding a filter to the sensor output or preventing rapid change in the control command. Dynamic control allocation(DCA), recently studied by Härkegård, is to distribute the control command in the sense of dynamics⁴: the allocation is made over a certain time interval, not a fixed time instant. The dynamic behavior of the control command is taken into account in the course of distributing the control command. Not only the control command requirement, but also variation of the control command over a sampling interval is included in the performance criterion to be optimized. The result is a control command in the form of a finite difference equation over the given time interval.⁴ It results in a filter dynamics by taking the previous control command into account for the synthesis of current control command. Stability of the proposed dynamic control allocation (CA) approach was proved to ensure the control command is bounded at the steady-state. In this study, we extended the results presented in Ref. 4 by adding a two-step dynamic CA term in deriving the control allocation law. Also, the strict equality constraint, between the virtual and actual control inputs, is relaxed in order to construct control command with a smooth profile. The proposed DCA technique is applied to a spacecraft attitude control problem. The sensor noise and/or irregular signals, which are existent in most of spacecraft attitude sensors, can be handled effectively by the proposed approach.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.13 no.2
• /
• pp.173-182
• /
• 1993

This paper presents an algorithm to derive the representative unit hydrograph for the real environment of a watershed. For a given watershed, the conventional methods give several different unit hydrographs by storm events. In this study the LP model is somewhat modified based on the previous study by Mays et also as follows: the objective function is designed to minimize the sum of weighted residuals. An additional constraint of moving average is added to prevent the unit hydrograph from the occurence of oscillation which was not active in Mays's paper. Configuration of rainfall matrix was improved to reduce its dimension in accordance with Diskin's review point. In spite of the superiority of LP approach in terms of representativeness, all the methods were very sensitive to the validity of baseflow separation and rainfall-loss. Several methods of the separations for rainfall excesses and direct runoffs were applied and no preferred methods were identified. This is the matter of judgement considering catchment and rainfall characteristics. This algorithm was applied to a real watershed of the Wi stream in the Nak-dong river. Compared with the IHP results by conventional methods, this optimized representative unit hydrograph demonstrated relatively smaller and shorter values in terms of the peak discharge and the basin lag respectively, and the oscillation of its falling limb successfully eliminated owing to the additional constraints of moving averages.