• Journal of the Korean Institute of Telematics and Electronics C
• /
• v.36C no.11
• /
• pp.75-83
• /
• 1999

We propose a PfSGA(parameter-free species genetic algorithm) to learn the topological structure of MLP classifiers being adequate to given applications. The PfSGA is a combinational method of SGA(species genetic algorithm) and PfGA(parameter-free genetic algorithm). In SGA, we divide the total search space into several subspaces(species) according to the number of hidden units, and reduce the unnecessary search by eliminating the low promising species from the evolutionary process. However the performances of SGA classifiers are readily affected by the values of parameters such as mutation ratio and crossover ratio. In this paper, therefore, we combine SGA with PfGA, for which it is not necessary to determine the learning parameters. Experimental results on benchmark data and sign language words show that PfSGA can reduce the learning time of SGA and is not affected by the selection parameter values on structural learning. The results also show that PfSGA is more efficient than the exisiting methods in the aspect of misclassification ratio, learning rate, and complexity of MLP structure.

• Journal of the Korean Geographical Society
• /
• v.44 no.2
• /
• pp.187-205
• /
• 2009

Failures on network components of a public transportation system can give rise to the severe degradation of entire system functionality. This paper aims at exploring how potential failures can affect the system flows and reliability of subway network systems in Metropolitan Seoul. To evaluate the range of impacts of disruptions, this research employs a probabilistic approach, network reliability. Network reliability measures the network resiliency and probability of flow loss under a variety of simulated disruptions of critical network components, transfer stations in subway system. By identifying the best and worst scenarios associated with geographical impact, as well as evaluating the criticality of transfer stations, this research presents some insights for protecting current subways systems.

• Composites Research
• /
• v.17 no.2
• /
• pp.28-33
• /
• 2004

Failure detection in a cross-ply laminated composite beam under tensile loading were performed using a fiber Bragg grating (FBG) sensor. A Passive Mach-Zehnder interferometric demodulator was proposed to enhance sensitivity and bandwidth. The proposed FBG sensor system without active device such as a phase modulator is very simple in configuration, easy to implement and enables the measurement of high-frequency vibration with low strain amplitude such as impact or failure signal. Failure signals detected by a FBG sensor had offset value corresponding to the strain shift with vibration at a maximum frequency of several hundreds of kilohertz. at the instant of transverse crack propagation in the 90 degree layer of composite beam.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.3 no.3
• /
• pp.214-221
• /
• 1998

A new control approach using neural network for the robust position control of a BRUSHLESS direct drive(BLDD) motor is presented. The linear quadratic controller plus feedforward neural network is employed to obtain the robust BLDD motor system approximately linearized using field-orientation method for an AC servo. The neural network is trained in on-line phases and this neural network is composed by a feedforward recall and error back-propagation training. Since the total number of nodes are only eight, this system will be easily realized by the general microprocessor. During the normal operation, the input-output response is sampled and the weighting value is trained by error back-propagation at each sample period to accommodate the possible variations in the parameters or load torque. And the state space analysis is performed to obtain the state feedback gains systematically. In addition, the robustness is also obtained without affecting overall system response.

• Journal of Korean Society of Occupational and Environmental Hygiene
• /
• v.22 no.2
• /
• pp.134-139
• /
• 2012

Objectives: This study was conducted to study the analytical variability of A & B counting rules in counting using a phase contrast microscope airborne fibers collected on filters in man-made mineral or vitreous fibers (MMMFs) industries. Methods: Fibers in filters were counted using A & B rules of NIOSH Method 7400. Intra-counter and inter-counter variations by fiber type and density were obtained. The types of MMMFs analyzed were glass wool fiber, rock wool fiber, slag wool fiber, and refractory ceramic fibers. The densities of fibers classified were <20 $fibers/mm^2,$ 20 - <50 $fibers/mm^2$, 50 - <100 $fibers/mm^2,$ and ${\geq}100$ $fibers/mm^2,$ respectively. Results: Intra-counter relative standard deviations by rule A were 0.084, 0.102, 0.071 for glass wool fibers, rock wool fibers and refractory ceramic fibers, and those by rule B were 0.139, 0.120 and 0.142, respectively. Inter-counter relative standard deviations by rule A were 0.281, 0.296, 0.180 for glass wool fibers, rock wool fibers and refractory ceramic fibers, and those by rule B were 0.396, 0.337 and 0.238, respectively. Conclusions: Intra-counter variation was not different significantly among fiber types (p>0.05), but B rule variation for ceramic fibers approximately 2 times greater than corresponding A rule estimates, and intra-counter and inter-counter variations were higher in the low fiber density.