http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Evolutionary design for multi-domain engineering system air pump redesign
서기성 한국지능시스템학회 2006 한국지능시스템학회논문지 Vol.16 No.2
This paper introduces design method for air pump system using bond graph and genetic programming to maximize outflow subject to a constraint specifying maximum power consumption. The air pump system is a mixed domain system which includes electromagnetic, mechanical and pneumatic elements. Therefore an appropriate approach for a better system for synthesis is required. Bond graphs are domain independent, allow free composition, and are efficient for classification and analysis of models. Genetic programming is well recognized as a powerful tool for open-ended search. The combination of these two powerful methods, BG/GP, was tested for redesign of air pump system.
서기성,박세현,권혁수,이정환,노석호 한국정보통신학회 2002 한국정보통신학회논문지 Vol.6 No.5
본 논문은 무어 머신을 복제하는 새로운 진화 하드웨어를 제안하였다. 제안된 진화 하드웨어는 FPGA 상에서 효과적인 파이프라인, 병렬처리와 Handshaking을 구현했다. 유전자 알고리즘은 다양한 응용 분야의 NP 문제를 해결하는 방법으로 알려져 있으나 긴 계산 시간이 요구되기 때문에 하드웨어 유전자 알고리즘이 최근 관심사가 되고 있다. 기존의 하드웨어 유전자 알고리즘은 고정 길이의 염색체를 사용하지만 제안된 진화 하드웨어는 가변 길이의 염색체를 사용한다. 실험 결과는 제안된 진화 하드웨어가 무어 머신을 복제하는데 있어 적합함을 알 수 있다. This paper proposes a new type of evolvable hardware for implementing the clone of Moore State machine. The proposed Evolvable Hardware is employed efficient pipeline parallelization, handshaking mechanism and fitness function in FPGA. Genetic Algorithm(GA) has known as a method of solving NP problem in various applications. Since a major drawback of the GA is that it needs a long computation time, the hardware implementation of Genetic Algorithm is focused on in recent studies. Conventional hardware GA uses the fixed length of chromosome but the proposed Evolvable Hardware uses the variable length of chromosome by the efficient 16 bit Pipeline Unit. Experimental results show that the proposed evolvable hardware is applicable to the implementation of the clone for Moore State machine.
서기성,박세현 한국정보통신학회 2004 한국정보통신학회논문지 Vol.8 No.6
멀티-도메인 공학시스템은 전기, 기계, 유압, 열등의 도메인을 포함하며, 시스템 구성이 복잡하여 설계에 많은 어려움을 가지고 있다. 최적의 설계를 위해서는 각 도메인에 대한 통합된 설계 방법과 자동적이고 효율적인 탐색방법이 요구된다. 본 논문은 도메인에 독립적인 본드 그래프(bond graph)와 대규모 공간 해의 탐색에 접합한 진화 알고리즘의 일종인 Genetic Programming(유전 프로그래밍, GP)를 결합하여 멀티 도메인 동적 시스템에 대한 디자인 해를 자동적으로 생성해주는 설계 방법을 제시하였다. 제안된 설계방법의 효용성을 입증하기 위해서 고유값(eigenvalue) 설계 문제가 실험되었고, 서로 다른 태아모델을 가진 고유값의 집합이 사용되었다. Multi-domain engineering systems include electrical, mechanical, hydraulic, pneumatic, and thermal components, making it difficult to design a system because of their complexity and inter domain nature. In order to obtain an optimal design, a unified design approach for each domain and an automated search method are required. This paper suggests a method for automatically synthesizing designs for multi-domain systems using the combination of bond graph that is domain independent and genetic programming that is well recognized as a powerful tool for open-ended search. To investigate the effect of proposed approach, an eigenvalue design problem is tested for some sample target sets of eigenvalues with different embryos.
가려진 사람의 재식별을 위한 트랜스포머 기반 교차 어텐션과 특징 다양화
서기성,강성재,김세준 대한전기학회 2023 전기학회논문지 Vol.72 No.1
Occluded person re-identification is a very difficult because the specific person is occluded by obstacles or other persons or by oneself. Major works adopt transformer-based approach show excellent performances, but they used a basic transformer only. In this paper, we suggest the various techniques to improve the transformer-based Re-ID method for the occluded person as follows. First, after extracting the heatmap and then deleting random body parts on the heat map, accurate keypoint information is obtained in data augmentation. Second, Cross-attention between the keypoint heatmap and the output of the transformer's middle layer is provided to focus more on the non-occluded person area. Third K-menas clustering is utilized to enhance the representation of local features, and the structure of the network is proposed to improve the diversity of the features. We evaluate mAP and Rank-1 performance on the Occluded-Duke and Market-1501 dataset and compare the proposed model with existing state-of-the-art techniques. Experimental results show that our method outperforms state-of-the-art methods.