• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2003년도 ICCAS
• /
• pp.587-591
• /
• 2003

This study develops a neural network for solving optimization problems. Hopfield network has been used for such problems, but it frequently gives abnormal solutions or non-optimal solutions. Moreover, it takes much time for solving a solution. To overcome such disadvantages, this study adopts a neural network whose output nodes change with a small value at every evolution, and the proposed neural network is applied to solve ALB (Assembly Line Balancing) problems . Given a precedence diagram and a required number of workstations, an ALB problem is solved while achieving even distribution of workload among workstations. Here, the workload variance is used as the index of workload deviation, and is reflected to an energy function. The simulation results show that the proposed neural network yields good results for solving ALB problems with high success rate and fast execution time.

• 전자공학회논문지A
• /
• 제30A권9호
• /
• pp.71-79
• /
• 1993

This paper proposes a new test pattern generation methodology for combinational logic circuits using neural networks based on a modular structure. The CUT (Circuit Under Test) is described in our gate level hardware description language. By conferring neural database, the CUT is compiled to an ATPG (Automatic Test Pattern Generation) neural network. Each logic gate in CUT is represented as a discrete Hopfield network. Such a neual network is called a gate module in this paper. All the gate modules for a CUT form an ATPG neural network by connecting each module through message passing paths by which the states of modules are transferred to their adjacent modules. A fault is injected by setting the activation values of some neurons at given values and by invalidating connections between some gate modules. A test pattern for an injected fault is obtained when all gate modules in the ATPG neural network are stabilized through evolution and mutual interactions. The proposed methodology is efficient for test generation, known to be NP-complete, through its massive paralelism. Some results on combinational logic circuits confirm the feasibility of the proposed methodology.

• 대한기계학회논문집
• /
• 제18권7호
• /
• pp.1881-1897
• /
• 1994

This paper presents a neural network-based computational scheme to generate the optimized robotic assembly sequence for an assembly product consisting of a number of parts. An assembly sequence is considered to be optimal when it meets a number of conditions : it must satisfy assembly constraints, keep the stability of in-process subassemblies, and minimize assembly cost. To derive such an optimal sequence, we propose a scheme using both the Hopfield neural network and the expert system. Based upon the inferred precedence constraints and the assembly costs from the expert system, we derive the evolution equation of the network. To illustrate the suitability of the proposed scheme, a case study is presented for industrial product of an electrical relay. The result is compared with that obtained from the expert system.

• 한국통신학회논문지
• /
• 제22권5호
• /
• pp.881-888
• /
• 1997

The existing method s for curvature estimation have a common problem in determining a unique smoothong factor. we previously proposed two approaches to overcome that problem: a constrained regularization approach and a mean field annealing approach. We consistently detected corners from the perprocessed smooth boundary obtained by either the constrained eglarization approach or the mean field annealing approach. Moreover, we defined corner sharpness to increase the robustness of both approaches. We evaluate the performance of those methods proposed in this paper. In addition, we show some matching results using a two-dimensional Hopfield neural network in the presence of occlusion as a demonstration of the power of our proposed methods.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1991년도 하계학술대회 논문집
• /
• pp.333-336
• /
• 1991

In this paper, Hopfield & Tank model-like artificial neural network structure is proposed, which can be used for the optimal path planning problems such as the unit commitment problems or the maintenance scheduling problems which have been solved by the dynamic programming method or the branch and bound method. To construct the structure of the neural network, an energy function is defined, of which the global minimum means the optimal path of the problem. To avoid falling into one of the local minima during the optimization process, the simulated annealing method is applied via making the slope of the sigmoid transfer functions steeper gradually while the process progresses. As a result, computer(IBM 386-AT 34MHz) simulations can finish the optimal unit commitment problem with 10 power units and 24 hour periods (1 hour factor) in 5 minites. Furthermore, if the full parallel neural network hardware is contructed, the optimization time will be reduced remarkably.

• 전자공학회논문지CI
• /
• 제37권4호
• /
• pp.26-36
• /
• 2000

본 논문은 실시간 멀티프로세서 스케줄링 문제를 효과적으로 해결하는 신경망 알고리즘을 제안한다. 제안된 알고리즘은 대표적인 신경망 모델인 홉 필드 네트워크를 근간으로 태스크의 처리요구에 대해 지정된 시간이내에 처리할 수 있는 실시간 시스템을 신경망의 장점인 병렬처리가 가능하도록 구현하였다. 본 알고리즘의 성능을 비교하기 위하여 기존에 실시간 멀티프로세서 스케줄링을 위해 연구되는 EDA와 LLA의 두 알고리즘과 비교한다. 제안된 알고리즘은 VHDL을 이용하여 하드웨어로 설계한다.

• 제어로봇시스템학회논문지
• /
• 제13권9호
• /
• pp.825-831
• /
• 2007

In this paper, an interface between a human and a computer is presented. The human and computer interface(HCI) serves as another area of human and machine interfaces. Methods for the HCI we used are voice recognition and image recognition for detecting human's emotional feelings. The idea is that the computer can recognize the present emotional state of the human operator, and amuses him/her in various ways such as turning on musics, searching webs, and talking. For the image recognition process, the human face is captured, and eye and mouth are selected from the facial image for recognition. To train images of the mouth, we use the Hopfield Net. The results show 88%$\sim$92% recognition of the emotion. For the vocal recognition, neural network shows 80%$\sim$98% recognition of voice.

• 전자공학회논문지S
• /
• 제35S권6호
• /
• pp.101-108
• /
• 1998

물체를 인식하기 위한 효율적인 방법 중의 하나는 물체의 경계선에서 가장 적절한 특징들을 추출해 내어 인식에 사용하는 것이다. 본 논문에서는 경계선 위의 각 화소에서 주변 화소들과의 관계를 이용해 코너점, 접점, 변곡점을 추출하여 물체의 특징점으로 사용하였다. 기존에 주로 사용되던 중요한 특징점의 하나인 코너점은 곡률 함수상에서 찾고, 또한 물체가 직선과 곡선으로 이루어져 있을 경우 코너점만으로 물체를 표현하기에 부족하므로 곡률 함수를 미디안 필터링하여 양자화 잡음을 제거함으로써 접점과 변곡점을 찾는 새로운 방법을 제안하였다. 그리고 이 세 가지 특징점을 물체 정합의 요소로 사용하여 물체를 정합하였다. 정합 방법으로는 Discrete Hopfield Neural Network을 사용하였으며, 성능 분석 결과 곡선이 섞인 물체에서 코너점만으로 물체를 정합한 경우보다 특징점으로 물체를 정합한 경우 우수한 정합 성능을 나타내었다.

• 대한산업공학회지
• /
• 제19권1호
• /
• pp.1-17
• /
• 1993

It is well known that an assembly operation is usually constrained by the geometric interference between parts. These constraints are normally presented as AND/OR precedence relationships. To find a feasible assembly sequence which satisfies the geometric constraints is not an easy task because of the TSP(Traveling Salesman Problem) nature with precedence constraints. In this paper, we developed an automated system based on Neural Network for generating feasible assembly sequences. Modified Hopfield and Tank network is used to solve the problem of AND/OR precedence-constrained assembly sequences. An economic assembly sequence can be also obtained by applying the cost matrix that contains cost-reducing factors. To evaluate the performance and effectiveness of the developed system, a case of automobile generator is tested. The results show that the developed system can provide a "good" planning tool for an assembly planner within a reasonable computation time period.

• 대한용접접합학회:학술대회논문집
• /
• 대한용접접합학회 2000년도 특별강연 및 춘계학술발표대회 개요집
• /
• pp.303-306
• /
• 2000

Quality estimation of the weld has been one of the important issues in RSW which is a main process of the sheep metal fabrication in auto-body industry, It was well known that among the various welding process variables, dynamic resistance has a close relation with nugget formation. With this variable, it is possible to estimate the weld quality in real time. In this study, a new quality estimation algorithm is developed with the primary dynamic resistance measured at welding machine timer. For this, feature recognition method of Hopfield neural network is used. Primary resistance patterns are vectorized and classified with five patterns. The network trained by these patterns recognizes the dynamic resistance pattern and estimates the weld quality Because the process variable monitored at the primary circuit is used, it is possible to apply this system to real time application without any consideration of electrode wear or shunt effect.