• Smart Structures and Systems
• /
• v.22 no.2
• /
• pp.175-183
• /
• 2018

Time-series data often contain one of the most valuable pieces of information in many fields including manufacturing. Because time-series data are relatively cheap to acquire, they (e.g., vibration signals) have become a crucial part of big data even in manufacturing shop floors. Recently, deep-learning models have shown state-of-art performance for analyzing big data because of their sophisticated structures and considerable computational power. Traditional models for a machinery-monitoring system have highly relied on features selected by human experts. In addition, the representational power of such models fails as the data distribution becomes complicated. On the other hand, deep-learning models automatically select highly abstracted features during the optimization process, and their representational power is better than that of traditional neural network models. However, the applicability of deep-learning models to the field of prognostics and health management (PHM) has not been well investigated yet. This study integrates the "residual fitting" mechanism inherently embedded in the wavelet transform into the convolutional neural network deep-learning structure. As a result, the architecture combines a signal smoother and classification procedures into a single model. Validation results from rotor vibration data demonstrate that our model outperforms all other off-the-shelf feature-based models.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.6
• /
• pp.9-21
• /
• 2013

Tracking is a process which explores user equipment (UE) in the area of tracking in terms of cells. In this paper, two tracking schemes based on macrocell-microcell tiers in hierarchical LTE networks, PMMT and IMMT, are evaluated. In this network, UE can receive a signal from macrocells and overlapping microcells, and can be called from each macrocell or microcell-tier in the PMMT. Also, the UE can be called from the combined macrocell-tier and microcell-tier in the IMMT. Finally, we analyze the optimization of load balance between marcocell-tier and microcell-tier, and an analytical model is developed to evaluate those two arrangements.

• Tunnel and Underground Space
• /
• v.21 no.3
• /
• pp.174-180
• /
• 2011

An anisotropic version of Mohr-Coulomb failure criterion is proposed in order to provide a strength criterion for transversely isotropic rock. The concept of fabric tensor introduced by Pietruszczak & Mroz (2001) is employed to define the friction angle and cohesion as scalar functions of the fabric tensors. The anisotroy in these two strength parameters are calculated in association with the consideration of the relative rotation between the principal stress coordinate and the principal material triad. The critical plane on which the anisotropic function maximized is found by an optimization technique based on the Lagrange multiplier method. To demonstrate the performance of the anisotropic failure criterion, conventional triaxial tests on the samples having various inclinations of weakness plane are simulated and the resulting triaxial strength and dip angle of failure plane are discussed.

• Journal of Korean Society of Water and Wastewater
• /
• v.28 no.5
• /
• pp.601-608
• /
• 2014

A coagulation-flocculation (CF) process using aluminum sulfate as a coagulant was employed to treat highly suspended solids in tunnel wastewater. Response surface methodology (RSM) based on a Box-Behnken design was applied to evaluate the effects of three factors (coagulant dosage, pH and temperature) on total suspended solids (TSS) removal efficiency as well as to identify optimal values of those factors to maximize removal of TSS. Optimal conditions of coagulant dosage and pH for maximum TSS removal changed depending on the temperature ($4{\sim}24^{\circ}C$). As temperature increased, the amount of coagulant dosage and pH level decreased for maximum TSS removal efficiency during the CF process. Proper adjustment of optimal pH and coagulant dosage to accommodate temperature fluctuations can improve TSS removal performance of the CF process.

• Journal of Ocean Engineering and Technology
• /
• v.29 no.5
• /
• pp.351-358
• /
• 2015

This paper presents an effective damage detection method using a meta model. A meta model is an approximation model that uses the relations between the design and response variables. It eliminates the need for repetitive analyses of computationally expensive models during the optimization process. In this study, a response surface model was employed as the meta model. The surface model was estimated using the correlation of the stiffness and natural frequencies of the structures. The locations and values of the damages were identified using a meta model-based damage detection method. Two numerical examples (a cantilever beam and jacket structure) were considered to verify the performance of the proposed method. As a result, the damages to the structures were accurately detected.

• Journal of the Korean Institute of Electrical and Electronic Material Engineers
• /
• v.25 no.1
• /
• pp.33-36
• /
• 2012

In this study, the thickness effects of $Al_2O_3$ layer on the sensing properties of $SiO_2/Al_2O_3$ (OA) stacked membrane were investigated using electrolyte-insulator-semiconductor (EIS) structure for high quality pH sensor. The $Al_2O_3$ layers with a respective thickness of 5 nm, 15 nm, 23 nm, 50 nm, and 100 nm were deposited on the 5-nm-thick $SiO_2$ layers. The electrical characteristics and sensing properties of each OA membranes were investigated using metal-insulator-semiconductor (MIS) and EIS devices, respectively. As a result, the OA stacked membrane with 23-nm-thick $Al_2O_3$ layer shows the excellent characteristics as a sensing membrane of EIS sensor, which can enhance the signal to noise ratio.

• Transactions on Electrical and Electronic Materials
• /
• v.15 no.3
• /
• pp.149-154
• /
• 2014

A new CMOS analog design methodology using an independently optimized self-cascode (SC) is proposed. This idea is based on the concept of the dual-workfunction-gate MOSFETs, which are equivalent to SC structures. The channel length of the source-side MOSFET is optimized, to give higher transconductance ($g_m$) and output resistance ($r_{out}$). The highest $g_m$ and $r_{out}$ of the SC structures are obtained by independently optimizing the channel length ratio of the SC MOSFETs, which is a critical design parameter. An operational amplifier (OPAMP) with the proposed design methodology using a standard digital $0.18-{\mu}m$ CMOS technology was designed and fabricated, to provide better performance. Independently $g_m$ and $r_{out}$ optimized SC MOSFETs were used in the differential input and output stages, respectively. The measured DC gain of the fabricated OPAMP with the proposed design methodology was approximately 18 dB higher, than that of the conventional OPAMP.

• Proceedings of the KIEE Conference
• /
• 1997.11a
• /
• pp.600-602
• /
• 1997

In this study, a hybrid neural net consisting of an Adaptive LVQ(ALVQ) algorithm and MLP is proposed to perform speaker identification task. ALVQ is a new learning procedure using adaptively feature vector sequence instead of only one feature vector in training codebooks initialized by LBG algorithm and the optimization criterion of this method is consistent with the speaker classification decision rule. ALVQ aims at providing a compressed, geometrically consistent data representation. It is fit to cover irregular data distributions and computes the distance of the input vector sequence from its nodes. On the other hand, MLP aim at a data representation to fit to discriminate patterns belonging to different classes. It has been shown that MLP nets can approximate Bayesian "optimal" classifiers with high precision, and their output values can be related a-posteriori class probabilities. The different characteristics of these neural models make it possible to devise hybrid neural net systems, consisting of classification modules based on these two different philosophies. The proposed method is compared with LBG algorithm, LVQ algorithm and MLP for performance.

• Journal of Electrical Engineering and Technology
• /
• v.10 no.3
• /
• pp.1383-1388
• /
• 2015

Proportional integral derivative (PID) control is one of the conventional control strategies. Industrial PID control has many options, tools, and parameters for dealing with the wide spectrum of difficulties and opportunities in manufacturing plants. It has a simple control structure that is easy to understand and relatively easy to tune. Injection mold is warming up to the idea of cycling the tool surface temperature during the molding cycle rather than keeping it constant. This “heating and cooling” process has rapidly gained popularity abroad. However, it has discovered that raising the mold wall temperature above the resin’s glass-transition or crystalline melting temperature during the filling stage is followed by rapid cooling and improved product performance in applications from automotive to packaging to optics. In previous studies, optimization methods were mainly selected on the basis of the subjective experience. Appropriate techniques are necessary to optimize the cooling channels for the injection mold. In this study, a digital signal processor (DSP)-based PID control system is applied to injection molding machines. The main aim of this study is to optimize the control of the proposed structure, including a digital PID control method with a DSP chip in the injection molding machine.

• Journal of Electrical Engineering and Technology
• /
• v.10 no.3
• /
• pp.820-831
• /
• 2015

In a deregulated electricity market, congestion of the transmission lines is a major problem the independent system operator (ISO) would face. Rescheduling of generators is one of the most practiced techniques to alleviate the congestion. However, not all generators in the system operate deterministically and independently, especially wind power generators (WTGs). Therefore, a novel optimal rescheduling model for congestion management that accounts for the uncertain and correlated power sources and loads is proposed. A probabilistic power flow (PPF) model based on 2m+1 point estimate method (PEM) is used to simulate the performance of uncertain and correlated input random variables. In addition, the impact of direct electricity purchase contracts on the congestion management has also been studied. This paper uses artificial bee colony (ABC) algorithm to solve the complex optimization problem. The proposed algorithm is tested on modified IEEE 30-bus system and IEEE 57-bus system to demonstrate the impacts of the uncertainties and correlations of the input random variables and the direct electricity purchase contracts on the congestion management. Both pool and nodal pricing model are also discussed.