• Proceedings of the KIEE Conference
• /
• 2003.11c
• /
• pp.621-624
• /
• 2003

For real world applications of reinforcement learning techniques, function approximation or generalization will be required to avoid curse of dimensionality. For this, an improved function approximation-based reinforcement learning method is proposed to speed up convergence by using CA-CMAC(Credit-Assigned Cerebellar Model Articulation Controller). To show that our proposed CACRL(CA-CMAC-based Reinforcement Learning) performs better than the CRL(CMAC-based Reinforcement Learning), computer simulation results are illustrated, where a swing-up control problem of an acrobot is considered.

• Proceedings of the KIEE Conference
• /
• 1998.11a
• /
• pp.70-72
• /
• 1998

Because of using PWM inverters and converters. The problems of insulation breakdown is emerging in both high voltage motors and general motors. In conventional methods, the viewpoint of surge problems is wave propagation with or without cable and inverter. For the purpose of knowing the situation of insulation breakdown, the end-turn coil of windings in the motor winding insulation structure is modelled by FEM, and field analysis of that is done. For first step, only end-turn coil is modelled and the model is simulated with FEM by approximating the resistivity of metallic foil surrounding insulation layers with having nonlinear property. Next, the result of simulation with nonlinear resistivity are compared with the result of linear resistivity. Because of microscope analysis, there is the problem of generalization but the situation of corona discharge in end-turn coil will be explained from this simulation.

• Journal of the Institute of Convergence Signal Processing
• /
• v.7 no.4
• /
• pp.214-220
• /
• 2006

This paper presents the systematic isotropy analysis of a fully actuated caster wheeled omnidirectional mobile robot (COMR) with the steering link offset different from the wheel radius, which can be considered as the generalization of the previous analysis. First with the characteristic length introduced, the kinematic model of a COMR is obtained based on the orthogonal decomposition of the wheel velocities. Second, the necessary and sufficient conditions for the isotropy of a COMR are derived and examined to categorize there different groups, each of which can be dealt with in a similar way. Third, for each group, the isotropy conditions are further explored so as to identify four different sets of all possible isotropic configurations. Fourth, for each set the expressions of the isotropic characteristic length required for the isotropy of a COMR are elaborated.

• Bulletin of the Korean Chemical Society
• /
• v.12 no.5
• /
• pp.529-536
• /
• 1991

In this paper, the question whether the curvature of the minimum energy path can affect the dynamic threshold was tested using the boundary trajectory method developed by Chesnavich and coworkers. For nonreactive system, the MO EXP model potential surface was used with modified equilibrium distance to control the curvature. The results showed that there is no relation between the curvature and the dynamic threshold. In order to study the reactive system, a generalization of the boundary trajectory method was achieved to apply on the nonsymmetric system. We have found no correspondence between the curvature and the dynamic threshold of the system. It was also shown that the fate of the trajectories strongly depends on the shape of potential surface around the turning points along the symmetric stretch line.

• The Magazine of the Society of Air-Conditioning and Refrigerating Engineers of Korea
• /
• v.17 no.4
• /
• pp.466-472
• /
• 1988

Pneumatic transportation systems are operated most economically, if the transportation velocity is somewhat greater than the limit velocity. Therefore it is very important to find out this limit velocity. That is well known for coarse and fine particles, but not yet for long ribbon type material. In this study a formula for the mass ratio of transported material to air in limit load point is derived for ribbon type material from the pneumatic transportation theory and the model theory. A relationship for estimating the limit velocity is also found out through experiments. The possibility of the generalization of this formula, which is obtained from a given pneumatic transportation system, should be estimated through further systematic studies.

• Communications for Statistical Applications and Methods
• /
• v.28 no.3
• /
• pp.307-314
• /
• 2021

We showed that any missing mechanism is reproduced by EMAR or MNAR with equal fit for observed likelihood if there are non-negative solutions of maximum likelihood equations. This is a generalization of Molenberghs et al. (2008) and Jeon et al. (2019). Nonetheless, as MCAR becomes a nested model of MNAR, a natural question is whether or not MNAR and MCAR are testable by using the well-known three statistics, LR (Likelihood ratio), Wald, and Score test statistics. Through simulation studies, we compared these three statistics. We investigated to what extent the boundary solution affect tesing MCAR against MNAR, which is the only testable pair of missing mechanisms based on observed likelihood. We showed that all three statistics are useful as long as the boundary proximity is far from 1.

• Advances in Computational Design
• /
• v.8 no.1
• /
• pp.37-59
• /
• 2023

Squat reinforced concrete (SRC) shear walls are a critical part of the structure for both office/residential buildings and nuclear structures due to their significant role in withstanding seismic loads. Despite this, empirical formulae in current design standards and published studies demonstrate a considerable disparity in predicting SRC wall shear strength. The goal of this research is to develop and evaluate hybrid and ensemble artificial neural network (ANN) models. State-of-the-art population-based algorithms are used in this research for hybrid intelligence algorithms. Six models are developed, including Honey Badger Algorithm (HBA) with ANN (HBA-ANN), Hunger Games Search with ANN (HGS-ANN), fitness-distance balance coyote optimization algorithm (FDB-COA) with ANN (FDB-COA-ANN), Averaging Ensemble (AE) neural network, Snapshot Ensemble (SE) neural network, and Stacked Generalization (SG) ensemble neural network. A total of 434 test results of SRC walls is utilized to train and assess the models. The results reveal that the SG model not only minimizes prediction variance but also produces predictions (with R²= 0.99) that are superior to other models.

• Journal of information and communication convergence engineering
• /
• v.21 no.3
• /
• pp.198-207
• /
• 2023

Responding to changes in artificial intelligence models and the data environment is crucial for increasing data-learning accuracy and inference stability of industrial applications. A learning model that is overfitted to specific training data leads to poor learning performance and a deterioration in flexibility. Therefore, an early stopping technique is used to stop learning at an appropriate time. However, this technique does not consider the homogeneity and independence of the data collected by heterogeneous nodes in a differential network environment, thus resulting in low learning accuracy and degradation of system performance. In this study, the generalization performance of neural networks is maximized, whereas the effect of the homogeneity of datasets is minimized by achieving an accuracy of 99.7%. This corresponds to a decrease in delay time by a factor of 2.33 and improvement in performance by a factor of 2.5 compared with the conventional method.

• The Korean Journal of Applied Statistics
• /
• v.3 no.1
• /
• pp.1-9
• /
• 1990

Engineering approach of statistics processing is defined for the statistical expert system. First, the engineering approach requirement are conceptualized by using an artificial intelligence in statistics, with the extensions being additional statistical knowlege engineering such as software engineering, optinal relationships, and the generalization abstraction. The methodology produces statistical expert system designes that are not only accurate representations of reality but also enough to accommodate future processing requirements. It also representions of knowledge that must be constructed, using the extended engineering processing model conceptualization and proposed engineering approach of the problem.

• Steel and Composite Structures
• /
• v.46 no.3
• /
• pp.435-449
• /
• 2023

Time domain method and frequency domain method are commonly used in the current fatigue life calculation theory. The time domain method has complicated procedures and needs a large amount of calculation, while the frequency domain method has poor applicability to different materials and different spectrum, and improper selection of spectrum model will lead to large errors. Considering that artificial neural network has strong ability of nonlinear mapping and generalization, this paper applied this technique to random fatigue life prediction, and the effect of average stress was taken into account, thereby achieving more accurate prediction result of random fatigue life.