• Nuclear Engineering and Technology
• /
• v.53 no.12
• /
• pp.3944-3951
• /
• 2021

Several reasons such as no free lunch theorem indicate that there is not a universal Feature selection (FS) technique that outperforms other ones. Moreover, some approaches such as using synthetic dataset, in presence of large number of FS techniques, are very tedious and time consuming task. In this study to tackle the issue of dependency of estimation accuracy on the selected FS technique, a methodology based on the heterogeneous ensemble is proposed. The performance of the major learning algorithms of neural network (i.e. the FFNN-BR, the FFNN-LM) in combination with the diverse FS techniques (i.e. the NCA, the F-test, the Kendall's tau, the Pearson, the Spearman, and the Relief) and different combination techniques of the heterogeneous ensemble (i.e. the Min, the Median, the Arithmetic mean, and the Geometric mean) are considered. The target parameters/transients of Bushehr nuclear power plant (BNPP) are examined as the case study. The results show that the Min combination technique gives the more accurate estimation. Therefore, if the number of FS techniques is m and the number of learning algorithms is n, by the heterogeneous ensemble, the search space for acceptable estimation of the target parameters may be reduced from n × m to n × 1. The proposed methodology gives a simple and practical approach for more reliable and more accurate estimation of the target parameters compared to the methods such as the use of synthetic dataset or trial and error methods.

• Smart Structures and Systems
• /
• v.29 no.2
• /
• pp.337-350
• /
• 2022

This paper focuses on predicting the post-heating mechanical properties of cementitious composites reinforced with multi-scale additives using the Artificial Neural Network (ANN) approach. A total of four different feed-forward ANN models are developed using 261 data sets collected from 18 published sources. The models are optimized using 12 input parameters selected based on a comprehensive literature review to predict the residual compressive strength, the residual flexural strengths, elastic modulus, and fracture energy of heat-damaged cementitious specimens. Furthermore, the ANN is employed to predict the impact of several variables including; the content of polypropylene (PP) microfibers and carbon nanotubes (CNTs) used in the concrete, mortar, or paste mix design, length of PP fibers, the average diameter of CNTs, and the average length of CNTs. The influence of the studied parameters is investigated at different heating levels ranged from 25℃ to 800℃. The results demonstrate that the developed ANN models have a strong potential for predicting the mechanical properties of the heated cementitious composites based on the mixing ingredients in addition to the heating conditions.

• Korean Chemical Engineering Research
• /
• v.61 no.2
• /
• pp.328-339
• /
• 2023

Methanol steam reforming (MSR) is a promising method for hydrogen supplying as a critical step in hydrogen fuel cell commercialization in mobile applications. Modelling and understanding of the reactor behavior is an attractive research field to develop an efficient reformer. Three-layer feed-forward artificial neural network (ANN) and Box-Behnken design (BBD) were used to modelling of MSR process using the Cu-SiO₂ aerogel catalyst. Furthermore, impacts of the basic operational variables and their mutual interactions were studied. The results showed that the most affecting parameters were the reaction temperature (56%) and its quadratic term (20.5%). In addition, it was also found that the interaction between temperature and Steam/Methanol ratio is important on the MSR performance. These models precisely predict MSR performance and have great agreement with experimental results. However, on the basis of statistical criteria the ANN technique showed the greater modelling ability as compared with statistical BBD approach.

• International conference on construction engineering and project management
• /
• 2015.10a
• /
• pp.195-199
• /
• 2015

Many conventional management methods have emphasized the minimization of required resources along the supply chain. Accordingly, this paper presents a proposed method called the Virtual Design and Construction (VDC)-aided system. It is based on object-oriented resource control, in order to accomplish a feed-forward control monitoring supply chain logistics. The system is supported by two main parts: (1) IT-based Technologies; and (2) VDC Models. They enable the system to convey proactive information from the detection technology to its linked visualization. The paper includes a field study as the system's pre-test: the Scaffolding Works in a LNG Mega Project. The study demonstrates a system of real-time productivity monitoring by use of the RFIDbased Mobile Information Hub. The on-line 'productivity dashboard' provides an opportunity to display the continuing processes for each work-package. This research project offers the observed opportunities created by the developed system. Future work will entail research experiments aimed towards system validation.

• Journal of Drive and Control
• /
• v.21 no.1
• /
• pp.31-37
• /
• 2024

In this study, to prevent accidents in underground facilities during excavation, we developed a Lv.3 automated control system that can be configured as an electronic control system without changing the existing hydraulic system in a general excavator and utilized digital map information of underground facilities. We aimed to develop a strategy to prevent accidents caused by operator error. To implement this, a real-time excavator bucket end position recognition and control system was developed through angle measurement of the boom, arm, and bucket using an electronic joystick, RTK-GPS, and angle sensors. In addition, excavators are large, machine-based equipment, and it is difficult to control overshoot due to inertia with feedback control using position recognition information of the bucket tip. Therefore, feed-forward control is used to calculate the moving speed of the bucket tip in real-time to determine the target position. We developed a technology that can converge and verified the performance of the developed system through actual vehicle installation and field tests.

• Journal of Information Technology Applications and Management
• /
• v.30 no.3
• /
• pp.1-13
• /
• 2023

Demand forecasting is a crucial task for an online retail where has to manage daily fresh foods effectively. Failing in forecasting results loss of profitability because of incompetent inventory management. This study investigated the optimal performance of different forecasting models for a very short shelf-life product. Demand data of 13 perishable items with aging of 210 days were used for analysis. Our comparison results of four methods: Trivial Identity, Seasonal Naïve, Feed-Forward and Autoregressive Recurrent Neural Networks (DeepAR) reveals that DeepAR outperforms with the lowest MAPE. This study also suggests the managerial implications by employing coefficient of variation (CV) as demand variation indicators. Three classes: Low, Medium and High variation are introduced for classify 13 products into groups. Our analysis found that DeepAR is suitable for medium and high variations, while the low group can use any methods. With this approach, the case can gain benefit of better fill-rate performance.

• Journal of the Korean Institute of Gas
• /
• v.18 no.1
• /
• pp.46-51
• /
• 2014

The purpose of this paper is to study of the vehicle lamp manufacturing system composing ultrasonic waves connection process. Making lamp assembly plant, it was produced in the separate process as the injection molding, ultrasonic waves bonding, annealing in the constant temperature, lamp assembling and packing. But the improvement method producing the lamp was added with one-step process by one automation technique. As a result, welding with ultrasonic waves process, the method decreased the energy consumption and noise during ultrasonic waves welding. Therefore, this method used the mathematics modeling for checking validity, it selected the stability and suitable controller using transfer function of plant and bode chart. In this study, the $180^{\circ}$ revolution control system to turn injection part upside down was $M_{eq}\;lcos{\theta}(t)$ because of gravity influence. It effected to unstable condition a system. For solving this problem, it aimed the linearization and stabilization of system by elimination $M_{eq}\;lcos{\theta}(t)$ as applying Free-forward control technique.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.24 no.12A
• /
• pp.2015-2024
• /
• 1999

Presented in this paper are a new complex-umber filter architecture, which is suitable for an efficient implementation of baseband signal processing of digital communication systems, and a chip-set design of adaptive decision-feedback equalizer (ADFE) employing the proposed structure. The basic concept behind the approach proposed in this paper is to apply redundant binary (RB) arithmetic instead of conventional 2’s complement arithmetic in order to achieve an efficient realization of complex-number multiplication and accumulation. With the proposed way, an N-tap complex-number filter can be realized using 2N RB multipliers and 2N-2 RB adders, and each filter tap has its critical delay of $T_{m.RB}+T_{a.RB}$ (where $T_{m.RB}, T_{a.RB}$are delays of a RB multiplier and a RB adder, respectively), making the filter structure simple, as well as resulting in enhanced speed by means of reduced arithmetic operations. To demonstrate the proposed idea, a prototype ADFE chip-set, FFEM (Feed-Forward Equalizer Module) and DFEM (Decision-Feedback Equalizer Module) that can be cascaded to implement longer filter taps, has been designed. Each module is composed of two complex-number filter taps with their LMS coefficient update circuits, and contains about 26,000 gates. The chip-set was modeled and verified using COSSAP and VHDL, and synthesized using 0.8- μm SOG (Sea-Of-Gate) cell library.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.1489-1494
• /
• 2003

A neuromuscular control system of a myoelectric prosthetic hand (PH) constitutes a nonlinear system with a dead zone whose magnitude is equal to its joint angle when the PH just grasps an object. This is because the neuromuscular control system remains an open-loop system until the PH grasps the object but it constitutes a feedback control system after the PH griped the object in which a torque induced in the fingers of the PH is fed back. To improve the transient performance of the control system, it is desirable to make the feed-forward gain as large as possible, so long as the stability of the system is not impaired. It is also desired that the control system remains stable even when the PH lifts a heavy or rigid object, because this makes the closed loop gain large and leads to the closed system unstable. According to the theory of stability analysis of nonlinear systems, we can only know the sufficient conditions that the system should be stable. Thus the nonlinear theory on stability is insufficient to be used to design the neuromuscular control system for improving its transient responses. This paper shows that the nonlinear system with a dead zone can be approximated to a linear feedback system and that well-known methods of analysis and design on linear control systems can be applicable. It is also shown through various simulation results that errors induced by approximation are practically negligible and thus the design methods are quite accurate.

• International Journal of Control, Automation, and Systems
• /
• v.5 no.6
• /
• pp.643-651
• /
• 2007

The various blood pressure simulators have been proposed to evaluate and improve the performance of the automatic sphygmomanometer. These have some problems such as the deviation of the actual blood pressure waveform, limitation in the blood pressure condition of the simulator, or difficulty in displaying the blood flow. An improved simulator using disturbance observer is proposed to supplement the current problems of the blood pressure simulator. The proposed simulator has an artificial arm model capable of feeding appropriate fluids that can generate the blood pressure waveform to evaluate the automatic sphygmomanometer. A controller was designed and thereafter, simulation was performed to control the output signal with respect to the reference input in the fluid dynamic model using the proposed proportional control valve. To minimize the external fluctuation of pressure applied to the artificial arm, a disturbance observer was designed on the plant. A hybrid controller combined with a proportional controller and feed-forward controller was fabricated after applying a disturbance observer to the control plant. Comparison of the simulations between the conventional proportional controller and the proposed hybrid controller indicated that even though the former showed good control performance without disturbance, it was affected by the disturbance signal induced by the cuff. The latter exhibited an excellent performance under both situations.