• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2003년도 하계종합학술대회 논문집 Ⅲ
• /
• pp.1565-1568
• /
• 2003

In this paper, we propose modular network to solve difficult and complex problems that are seldom solved with multi-layer neural network. The structure of modular neural network in researched by Jacobs and Jordan is selected in this paper. Modular network consists of several expert networks and a gating network which is composed of single-layer neural network or multi-layer neural network. We propose modular network structure using recurrent neural network, since the state of the whole network at a particular time depends on an aggregate of previous states as well as on the current input. Finally, we show excellence of the proposed network compared with modular network.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2003년도 ISIS 2003
• /
• pp.694-697
• /
• 2003

In this paper, we propose modular network to solve difficult and complex problems that are seldom solved with Multi-Layer Neural Network(MLNN). The structure of Modular Neural Network(MNN) in researched by Jacobs and jordan is selected in this paper. Modular network consists of several Expert Networks(EN) and a Gating Network(CN) which is composed of single-layer neural network(SLNN) or multi-layer neural network. We propose modular network structure using Recurrent Neural Network(RNN), since the state of the whole network at a particular time depends on aggregate of previous states as well as on the current input. Finally, we show excellence of the proposed network compared with modular network.

• 토지주택연구
• /
• 제10권3호
• /
• pp.67-75
• /
• 2019

The modular construction means that more than 70% of the parts such as walls, windows, electrical wiring, facility piping, bathrooms, and kitchen appliances are pre-assembled at the factory and transported to the site. It is possible to shorten the construction period than general construction work and to secure high quality through modular mass production since the modular construction works in the field at the same time as the modular production. However, there are only four domestic modular manufacturers, and each company's modular components and construction methods are different, so it is necessary to standardize them. Therefore, this study investigated the construction process centering on the stacking method of modular housing construction work applied to D site in Cheonan-si, and proposed the key points of quality management by construction stage. As the project was conducted as a pilot project for government R&D projects, some differences may occur from general modular housing construction. However, the construction process and quality control focus of each unit box type modular house analyzed in this study can be used as basic data in the future of modular housing construction. In addition, the results of this study can be used to establish construction standards, such as the development of checklists and establishment of standard processes.

• 한국정밀공학회지
• /
• 제15권2호
• /
• pp.35-44
• /
• 1998

This paper describes multiple fault diagnosis method in complex system with hierarchical structure. Complex system is divided into subsystem, item and component. For diagnosing this hierarchical complex system, it is necessary to implement special neural network. We introduced Modular Artificial Neural Network(MANN) for this purpose. MANN consists of four level neural network, first level for symptom classification, second level for item fault diagnosis, third level for component symptom classification, forth level for component fault diagnosis. Each network is multi layer perceptron with 7 inputs, 30 hidden node and 7 outputs trained by backpropagation. UNIX IPC(Inter Process Communication) is used for implementing MANN with multitasking and message transfer between processes in SUN workstation. We tested MANN in reactor system.

• IEIE Transactions on Smart Processing and Computing
• /
• 제5권1호
• /
• pp.55-62
• /
• 2016

In this paper, we propose a hybrid multi.system-on-chip (H-MSoC) architecture that provides a high-flexibility system in a rapid development time. The H-MSoC approach provides a flexible system-on-chip (SoC) architecture that is easy to configure for physical- and application-layer development. The physical- and application-layer aspects are dynamically designed and modified; hence, it is important to consider a design methodology that supports rapid SoC development. Physical layer development refers to intellectual property cores or other modular hardware (HW) development, while application layer development refers to user interface or application software (SW) development. H-MSoC is built from multi-SoC architectures in which each SoC is localized and specified based on its development focus, either physical or application (hybrid). Physical HW development SoC is referred to as physical-SoC (Phy-SoC) and application SW development SoC is referred to as application-SoC (App-SoC). Phy-SoC and App-SoC are connected to each other via Ethernet. Ethernet was chosen because of its flexibility, high speed, and easy configuration. For prototyping, we used a LEON3 SoC as the Phy-SoC and a ZYNQ-7000 SoC as the App-SoC. The proposed design was proven in real-time tests and achieved good performance.

• 한국정보과학회 언어공학연구회:학술대회논문집(한글 및 한국어 정보처리)
• /
• 한국정보과학회언어공학연구회 1989년도 한글날기념 학술대회 발표논문집
• /
• pp.267-271
• /
• 1989

최근 Scofield는 coulomb energy network에 적용할 수 있는 learning algorithm(supervised learning algorithm)을 제안하였다. 이 learning algorithm은 multi-layer network에도 쉽게 적용이 가능하고 한 layer 에서 발생한 error가 다른 layer에 영향을 주지 않아서 system을 modular하게 구성할 수가 있으며 각 layer를 독립적으로 learning 시킬 수 있는 특징이 있다. 본 논문에서는 coulomb energy network를 이용하여 한글인식을 위한 neural network를 구현하여 인식실험을 한 결과와 구현한 network 에서 인식율을 높이기 위한 방안 (2 stage learning) 을 제시한다.

• ETRI Journal
• /
• 제41권6호
• /
• pp.863-872
• /
• 2019

Elliptic curve cryptography is a relatively lightweight public-key cryptography method for key generation and digital signature verification. Some lightweight curves (eg, Curve25519 and Curve Ed448) have been adopted by upcoming Transport Layer Security 1.3 (TLS 1.3) to replace the standardized NIST curves. However, the efficient implementation of Curve Ed448 on Internet of Things (IoT) devices remains underexplored. This study is focused on the optimization of the Curve Ed448 implementation on low-end IoT processors (ie, 8-bit AVR and 16-bit MSP processors). In particular, the three-level and two-level subtractive Karatsuba algorithms are adopted for multi-precision multiplication on AVR and MSP processors, respectively, and two-level Karatsuba routines are employed for multi-precision squaring. For modular reduction and finite field inversion, fast reduction and Fermat-based inversion operations are used to mitigate side-channel vulnerabilities. The scalar multiplication operation using the Montgomery ladder algorithm requires only 103 and 73 M clock cycles on AVR and MSP processors.

• BioChip Journal
• /
• 제12권4호
• /
• pp.280-286
• /
• 2018

Hydrogel scaffolds composed of multiple components are promising platform in tissue engineering as a transplantation materials or artificial organs. Here, we present a new fabrication method for implementing multi-layered macroscopic hydrogel scaffold composed of multiple components by controlling height of hydrogel layer through precise control of ultraviolet (UV) energy density. Through the repetition of the photolithography process with energy control, we can form several layers of hydrogel with different height. We characterized UV energy-dependent profiles with single-layered PEGDA posts photocrosslinked by the modular methodology and examined the optical effect on the fabrication of multi-layered, macroscopic hydrogel structure. Finally, we successfully demonstrated the potential applicability of our approach by fabricating various macroscopic hydrogel constructs composed of multiple hydrogel layers.

• 전자공학회논문지CI
• /
• 제40권6호
• /
• pp.56-68
• /
• 2003

범용 신경망 연산기를 디지털 회로로 구현함에 있어 가장 까다로운 문제들 중 하나는 시냅스의 확장과 해당 네트워크에 맞게 뉴런들을 재배치하는 재구성 문제일 것이다. 본 논문에서는 이러한 문제들을 해결하기 위한 새로운 하드웨어 구조를 제안한다. 제안된 구조는 시냅스의 확장과 네트워크 구조의 변경을 위해 오리지날 디자인의 변경이 필요치 않으며, 모듈러 프로세싱 유니트의 확장을 통한 뉴런의 개수 및 레이어의 확장이 가능하다. 이 구조의 범용성 및 확장성에 대한 검증을 위해 다양한 종류의 다층 퍼셉트론 및 코호넨 네트워크를 구성하여 HDL 시뮬레이터를 통한 결과와 C 언어로 작성된 소프트웨어 시뮬레이터 결과를 비교하였으며 그 결과 성능이 거의 일치함을 확인하였다.

• 전기전자학회논문지
• /
• 제27권4호
• /
• pp.430-438
• /
• 2023

The Modular Multilevel Converter (MMC) has emerged as a key component in HVDC systems due to its ability to efficiently transmit large amounts of power over long distances. In such systems, accurate estimation of the MMC capacitor voltage is of utmost importance for ensuring optimal system performance, stability, and reliability. Traditional methods for voltage estimation may face limitations in accuracy and robustness, prompting the need for innovative approaches. In this paper, we propose a novel distributed neural network observer specifically designed for MMC capacitor voltage estimation. Our observer harnesses the power of a multi-layer neural network architecture, which enables the observer to learn and adapt to the complex dynamics of the MMC system. By utilizing a distributed approach, we deploy multiple observers, each with its own set of neural network layers, to collectively estimate the capacitor voltage. This distributed configuration enhances the accuracy and robustness of the voltage estimation process. A crucial aspect of our observer's performance lies in the meticulous initialization of random weights within the neural network. This initialization process ensures that the observer starts with a solid foundation for efficient learning and accurate voltage estimation. The observer iteratively updates its weights based on the observed voltage and current values, continuously improving its estimation accuracy over time. The validity of proposed algorithm is verified by the result of estimated voltage at each observer in capacitor of MMC.