• Transactions of Materials Processing
• /
• v.24 no.5
• /
• pp.332-339
• /
• 2015

In the current paper, a cold forging sequence was developed to manufacture a precisely cold forged H/Lower, which is used as the air back unit in commercial automobiles. The preform shape of the H/Lower influences the dimensional accuracy and stiffness of the final product. The shape factor (SF) ratio and shape of the tools are considered as the design parameters to achieve adequate backward extrusion height and maintain appropriate thickness variations. The optimal conditions of the design parameters were determined by using an artificial neural network (ANN). To experimentally verify the optimal preform and tool shapes, the experiments of the backward extrusion of the H/Lower were executed. The process design methodology proposed in the current paper, can provide a more systematic and economically feasible means for designing the preform and tool shapes for cold forging.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2005.10a
• /
• pp.946-949
• /
• 2005

It is very a time-consuming and error-prone process to obtain the optimal injection condition, which can produce good injection molding products in some operational variation of facilities, from a seed injection condition. This study proposes a new approach to search the optimal injection molding condition using a neural network and a genetic algorithm. To estimate the defect type of unknown injection conditions, this study forces the neural network into learning iteratively from the injection molding conditions collected. Major two parameters of the injection molding condition - injection pressure and velocity are encoded in a binary value to apply to the genetic algorithm. The optimal injection condition is obtained through the selection, cross-over, and mutation process of the genetic algorithm. Finally, this study compares the optimal injection condition searched using the proposed approach. with the other ones obtained by heuristic algorithms and design of experiment technique. The comparison result shows the usability of the approach proposed.

• Journal of the military operations research society of Korea
• /
• v.23 no.2
• /
• pp.171-191
• /
• 1997

In this paper, we design and construct a high speed LAN(Local Area Network) using FDDI(Fiber Distributed Data Interface) and Frame Relay in order to support our multimedia communication services. A program of this communication network is divided into requirement analysis, design, establishment and test. First, we propose an optimal communication method that compares various network techniques in the requirement analysis phase. Second, we design the physical network configuration, secure method, and address in the LAN and WAN. Finally, we establish and test the communication devices and lines. Ultimately, we minimized mistakes and satisfied user requirements using this program. We constructed efficiently a high speed data communication network using FDDI and Frame Relay.

• Journal of Information Processing Systems
• /
• v.18 no.6
• /
• pp.755-762
• /
• 2022

The existing electrical engineering design method and the dynamic objective function in the application process of automatic control system fail to meet the unbounded condition, which affects the control tracking accuracy. In order to improve the tracking control accuracy, this paper studies the electrical engineering design method based on neural network and the application of automatic control system. This paper analyzes the structure and working mechanism of electrical engineering automation control system by an automation control model with main control objectives. Following the analysis, an optimal solution of controllability design and fault-tolerant control is figured out. The automatic control power coefficient is distributed based on an ideal control effect of system. According to the distribution results, an automatic control algorithm is based on neural network for accurate control. The experimental results show that the electrical automation control method based on neural network can significantly reduce the control following error to 3.62%, improve the accuracy of the electrical automation tracking control, thus meeting the actual production needs of electrical engineering automation control system.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.42 no.1
• /
• pp.87-94
• /
• 2019

A wireless sensor network is emerging technology and intelligent wireless communication paradigm that is dynamically aware of its surrounding environment. It is also able to respond to it in order to achieve reliable and efficient communication. The dynamical cognition capability and environmental adaptability rely on organizing dynamical networks effectively. However, optimally clustering the cognitive wireless sensor networks is an NP-complete problem. The objective of this paper is to develop an optimal sensor network design for maximizing the performance. This proposed Ranking Artificial Bee Colony (RABC) is developed based on Artificial Bee Colony (ABC) with ranking strategy. The ranking strategy can make the much better solutions by combining the best solutions so far and add these solutions in the solution population when applying ABC. RABC is designed to adapt to topological changes to any network graph in a time. We can minimize the total energy dissipation of sensors to prolong the lifetime of a network to balance the energy consumption of all nodes with robust optimal solution. Simulation results show that the performance of our proposed RABC is better than those of previous methods (LEACH, LEACH-C, and etc.) in wireless sensor networks. Our proposed method is the best for the 100 node-network example when the Sink node is centrally located.

• Journal of the Korean Institute of Telematics and Electronics S
• /
• v.35S no.4
• /
• pp.67-75
• /
• 1998

In general, Evoluationary Algorithm(EAs) are refered to as methods of population-based optimization. And EAs are considered as very efficient methods of optimal sytem design because they can provice much opportunity for obtaining the global optimal solution. This paper presents a co-evolution scheme of artifical neural networks, which has two different, still cooperatively working, populations, called as a host popuation and a parasite population, respectively. Using the conventional generatic algorithm the host population is evolved in the given environment, and the parastie population composed of schemata is evolved to find useful schema for the host population. the structure of artificial neural network is a diagonal recurrent neural netork which has self-feedback loops only in its hidden nodes. To find optimal neural networks we should take into account the structure of the neural network as well as the adaptive parameters, weight of neurons. So we use the genetic algorithm that searches the structure of the neural network by the co-evolution mechanism, and for the weights learning we adopted the evolutionary stategies. As a results of co-evolution we will find the optimal structure of the neural network in a short time with a small population. The validity and effectiveness of the proposed method are inspected by applying it to the stabilization and position control of the invered-pendulum system. And we will show that the result of co-evolution is better than that of the conventioal genetic algorithm.

• Journal of Korean Association for Spatial Structures
• /
• v.23 no.2
• /
• pp.29-36
• /
• 2023

Reinforcement learning (RL) is widely applied to various engineering fields. Especially, RL has shown successful performance for control problems, such as vehicles, robotics, and active structural control system. However, little research on application of RL to optimal structural design has conducted to date. In this study, the possibility of application of RL to structural design of reinforced concrete (RC) beam was investigated. The example of RC beam structural design problem introduced in previous study was used for comparative study. Deep q-network (DQN) is a famous RL algorithm presenting good performance in the discrete action space and thus it was used in this study. The action of DQN agent is required to represent design variables of RC beam. However, the number of design variables of RC beam is too many to represent by the action of conventional DQN. To solve this problem, multi-agent DQN was used in this study. For more effective reinforcement learning process, DDQN (Double Q-Learning) that is an advanced version of a conventional DQN was employed. The multi-agent of DDQN was trained for optimal structural design of RC beam to satisfy American Concrete Institute (318) without any hand-labeled dataset. Five agents of DDQN provides actions for beam with, beam depth, main rebar size, number of main rebar, and shear stirrup size, respectively. Five agents of DDQN were trained for 10,000 episodes and the performance of the multi-agent of DDQN was evaluated with 100 test design cases. This study shows that the multi-agent DDQN algorithm can provide successfully structural design results of RC beam.

• Journal of Korean Society of Water and Wastewater
• /
• v.13 no.3
• /
• pp.73-82
• /
• 1999

In this study, the optimal analysis for pipe network is performed for the combined ideal pipe network system(CASE 1, CASE 2 and CASE 3) which is composed of 25 nodes, 41 elements, and 1 fixed nodal head with evaluating pressure variation distribution of main and branch in grid composed drainage pipe network. The linear analysis technique used as the analysis method in this study, the KYPIPE being used extensively as the linear technique to design and analysis of pipe network is applied. Firstly, in the analysis of pipe network, the CASE 2 and CASE 3 supply same thing(value) in the result of considering the total flow provided each pipeline, but in the general intension in the case of CASE 2, relative width of supply is more large than CASE 1 and CASE 3. Secondly, in the analysis technique of pipe network, CASE 3 is analysed largest as a result of comparing with same heads, and in the order of their size CASE 2 and CASE 1 were determined but the difference doesn't appear to be obvious. Thirdly, as the result of determining main factor, pressure in the design and analysis of net work. CASE 3 is from Node 3 to 25 than CASE 1 and CASE 2 and it is determined in the order of their size, CASE 2 and CASE 1. Finally, in this study, discharge flow distribution is evaluated in the same condition with 3-type CASE in the case of branch position for designing optimal composed drainage pipe network. As the result of that, branch pipe perform. Therefore, it is thought that the efficient and reasonable management of water supply and sewerage design will be possible if it give all our energies to study at the pipe system design in and out of country in the future.

• 한국정보통신설비학회:학술대회논문집
• /
• 2009.08a
• /
• pp.68-71
• /
• 2009

All-IP network requires change of the existing IP network engineering methods as the convergence service market between communication and broadcasting industries using IP network is growing rapidly. Especially the video services like IPTV require more strict transmission quality and higher bandwidth than the existing data services. So it is difficult to design All-IP network by the over-provisioning method which used to be used for the existing IP network design. It also requires a heavy investment which becomes one of big obstacles to the IPTV service expansion. In order to reduce the investment costs, it is required to design an optimized network by maximizing the utilization of the network resources and at the same time maintaining the customer satisfaction in terms of service quality. In this paper, we first analyze the effects of IPTV traffic on the existing internet. Then we compare two traffic engineering technologies, which are dimensioning without admission control and dimensioning with admission control, on the All-IP network design by simulation. Finally, we suggest cost effectiveness of traffic engineering technologies for designing the All-IP network.

• Proceedings of the Korean Society for Emotion and Sensibility Conference
• /
• 1998.11a
• /
• pp.189-194
• /
• 1998

Design factors are very important and deterministic in determining the first impression of products and environment. The final 30 number of channel button were chosen as a design factors at the Audio Unit. Then, we made the 8 types of prototype. with combining the design factors for experiment. Subjects rated the SD(Semantic Differential) evaluation sheets which have the 30 adjectives after watching each prototype. With the evaluated values, we simulated to identify the relation between the design factors and the adjectives using Neural Network. As a results, we could abstract the affective adjectives on each 8 types. Therefore, this research suggested the possibilities that we can infer the optimal design factors and types using Neural Network, if adjectives were given.