• 한국전산구조공학회논문집
• /
• 제21권5호
• /
• pp.439-450
• /
• 2008

잘 구축된 축소시스템은 동하중을 받는 구조물의 거동을 정확하게 계산할 수 있으며, 유한요소 기반 동적해석에서 문제가 될 수 있는 계산시간과 전산자원의 문제를 해결할 수 있다. 본 연구에서는 축소모델 기반 동적해석 알고리즘을 개발하였고, 동적 축소모델의 구축을 위한 주자유도 선정방법을 제안하였다. 이 과정에서 기존 연구에서 신뢰성이 검증된 2단계 축소기법을 사용하여 중요 자유도를 선정하고, IRS 방법에 의해 최종 축소모델을 구축하였다. 이를 임의의 동하중을 받는 수치예제에 적용하고 전체시스템의 동적해석 결과와 비교하여 제안 방법의 신뢰성을 검증하였다.

• 대한전자공학회논문지SD
• /
• 제39권9호
• /
• pp.37-44
• /
• 2002

학습된 신경망(Pre-trained neural network)은 고정된 가중치(weight)를 갖는다. 이 논문에서는 이러한 특성을 이용하여 신경망의 효과적인 디지털 하드웨어의 설계방법을 제안한다. 이를 위해 신경망의 PEs(Processing Elements)연산은 행렬-벡터 곱셈으로 표하고 고정된 가중치와 입력 데이터의 관계를 비트-레벨 어레이(array) 구조로 표현하여, 노드 소거와 가중치 비트 패턴에 따른 공유 노드 설정을 통한 최적화로 연산에 필요한 노드를 최소화한다. FPGA 시뮬레이션 결과, 완전한 정확성에 기반한 하드웨어를 설계하는 경우, 하드웨어 비용을 상당부분 줄였고 동작 주파수가 높다는 것을 확인하였다. 또한, 제안한 설계방법은 한정된 공간 내에서 많은 수의 PEs 구현이 가능함으로, 큰 신경망 모델에 대한 온-칩(on-chip) 구현이 가능하다.

• 한국방송∙미디어공학회:학술대회논문집
• /
• 한국방송공학회 2013년도 추계학술대회
• /
• pp.54-57
• /
• 2013

최근 다양한 방송 및 영상 분야에서 사람의 행동을 인식하여는 연구들이 많이 이루어지고 있다. 영상은 다양한 형태를 가질 수 있기 때문에 제약된 환경에서 유용한 템플릿 방법들보다 특징점에 기반한 연구들이 실제 사용자 환경에서 더욱 관심을 받고 있다. 특징점 기반의 연구들은 영상에서 움직임이 발생하는 지점들을 찾아내어 이를 3차원 패치들로 생성한다. 이를 이용하여 영상의 움직임을 히스토그램에 기반한 descriptor(서술자)로 표현하고 학습기반의 판별기(classifier)로 최종적으로 영상 내에 존재하는 행동들을 인식하였다. 그러나 단일 판별기를 이용한 다양한 영상 인식을 수용하기에는 힘들다. 최근에 이를 개선하기 위하여 다중 판별기를 활용한 연구들이 영상 판별 및 물체 검출 영역에서 사용되고 있다. 따라서 본 논문에서는 행동 인식을 위하여 support vector machine과 spare representation을 이용한 decision-level fusion 방법을 제안하고자 한다. 제안된 논문의 방법은 영상에서 특징점 기반의 descriptor를 추출하고 이를 각각의 판별기를 통하여 판별 결과들을 획득한다. 이 후 학습단계에서 획득된 가중치를 활용하여 각 결과들을 융합하여 최종 결과를 도출하였다. 본 논문에 실험에서 제안된 방법은 기존의 융합 방법보다 높은 행동 인식 성능을 보여 주었다.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2004년도 하계학술대회 논문집 B
• /
• pp.1251-1253
• /
• 2004

This paper is proposed a fuzzy neural network controller based on the vector controlled surface permanent magnet synchronous motor(SPMSM) drive system. The hybrid combination of neural network and fuzzy control will produce a powerful representation flexibility and numerical processing capability. Also, this paper is proposed speed control of SPMSM using neuro-fuzzy control(NFC) and estimation of speed using artificial neural network(ANN) Controller. The back propagation neural network technique is used to provide a real time adaptive estimation of the motor speed. The error between the desired state variable and the actual one is back-propagated to adjust the rotor speed, so that the actual state variable will coincide with the desired one. The back propagation mechanism is easy to derive and the estimated speed tracks precisely the actual motor speed. This paper is proposed the theoretical analysis as well as the simulation results to verify the effectiveness of the new method.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2005년도 추계학술대회 논문집 전기기기 및 에너지변환시스템부문
• /
• pp.195-197
• /
• 2005

This paper is proposed adaptive fuzzy neural network(AFNN) and artificial neural network(ANN) based on the vector controlled induction motor drive system. The hybrid combination of fuzzy control and neural network will produce a powerful representation flexibility and numerical processing capability. Also, this paper is proposed control and estimation of speed of induction motor using fuzzy and neural network. The back propagation neural network technique is used to provide a real time adaptive estimation of the motor speed. The error between the desired state variable and the actual one is back-propagated to adjust the rotor speed. so that the actual state variable will coincide with the desired one. This paper is proposed the experimental results to verify the effectiveness of the new method.

• Smart Structures and Systems
• /
• 제29권4호
• /
• pp.617-624
• /
• 2022

A new intelligent adaptive control scheme was proposed that combines the control based on interference observer and fuzzy adaptive s-curve for flight path tracking control of unmanned aerial vehicle (UAV). The most important contribution is that the control configurations don't need to know the uncertainty limit of the vehicle and the influence of interference is removed. The proposed control law is an integration of fuzzy control estimator and adaptive proportional integral (PI) compensator with input. The rated feedback drive specifies the desired dynamic properties of the closed control loop based on the known properties of the preferred acceleration vector. At the same time, the adaptive PI control compensate for the unknown of perturbation. Additional terms such as s-surface control can ensure rapid convergence due to the non-linear representation on the surface and also improve the stability. In addition, the observer improves the robustness of the adaptive fuzzy system. It has been proven that the stability of the regulatory system can be ensured according to linear matrix equality based Lyapunov's theory. In summary, the numerical simulation results show the efficiency and the feasibility by the use of the robust control methodology.

• International Journal of Computer Science & Network Security
• /
• 제23권9호
• /
• pp.47-54
• /
• 2023

For Bengali music emotion classification, deep learning models, particularly CNN and RNN are frequently used. But previous researches had the flaws of low accuracy and overfitting problem. In this research, attention-based Conv1D and BiGRU model is designed for music emotion classification and comparative experimentation shows that the proposed model is classifying emotions more accurate. We have proposed a Conv1D and Bi-GRU with the attention-based model for emotion classification of our Bengali music dataset. The model integrates attention-based. Wav preprocessing makes use of MFCCs. To reduce the dimensionality of the feature space, contextual features were extracted from two Conv1D layers. In order to solve the overfitting problems, dropouts are utilized. Two bidirectional GRUs networks are used to update previous and future emotion representation of the output from the Conv1D layers. Two BiGRU layers are conntected to an attention mechanism to give various MFCC feature vectors more attention. Moreover, the attention mechanism has increased the accuracy of the proposed classification model. The vector is finally classified into four emotion classes: Angry, Happy, Relax, Sad; using a dense, fully connected layer with softmax activation. The proposed Conv1D+BiGRU+Attention model is efficient at classifying emotions in the Bengali music dataset than baseline methods. For our Bengali music dataset, the performance of our proposed model is 95%.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제17권8호
• /
• pp.2101-2123
• /
• 2023

Recent studies have shown that the neural network-based binary code similarity detection technology performs well in vulnerability mining, plagiarism detection, and malicious code analysis. However, existing cross-architecture methods still suffer from insufficient feature characterization and low discrimination accuracy. To address these issues, this paper proposes a cross-architecture binary function similarity detection method based on composite feature model (SDCFM). Firstly, the binary function is converted into vector representation according to the proposed composite feature model, which is composed of instruction statistical features, control flow graph structural features, and application program interface calling behavioral features. Then, the composite features are embedded by the proposed hierarchical embedding network based on a graph neural network. In which, the block-level features and the function-level features are processed separately and finally fused into the embedding. In addition, to make the trained model more accurate and stable, our method utilizes the embeddings of predecessor nodes to modify the node embedding in the iterative updating process of the graph neural network. To assess the effectiveness of composite feature model, we contrast SDCFM with the state of art method on benchmark datasets. The experimental results show that SDCFM has good performance both on the area under the curve in the binary function similarity detection task and the vulnerable candidate function ranking in vulnerability search task.

• Kyungpook Mathematical Journal
• /
• 제63권1호
• /
• pp.123-129
• /
• 2023

For an ordered set W = {w₁, w₂, . . . , w_k} of vertices and a vertex v in a connected graph G, the k-vector r(v|W) = (d(v, w₁), d(v, w₂), . . . , d(v, w_k)) is called the metric representation of v with respect to W, where d(x, y) is the distance between the vertices x and y. A set W is called a resolving set for G if distinct vertices of G have distinct metric representations with respect to W. The minimum cardinality of a resolving set for G is its metric dimension dim(G), and a resolving set of minimum cardinality is a basis of G. The corona product, G ⊙ H of graphs G and H is obtained by taking one copy of G and n(G) copies of H, and by joining each vertex of the ith copy of H to the ith vertex of G. In this paper, we obtain bounds for dim(G ⊙ K₁), characterize all graphs G with dim(G ⊙ K₁) = dim(G), and prove that dim(G ⊙ K₁) = n - 1 if and only if G is the complete graph K_n or the star graph K_1,n-1.

• Wind and Structures
• /
• 제38권1호
• /
• pp.1-13
• /
• 2024

In order to effectively simulate nonstationary stochastic turbulent wind fields, four separation hybrid (SEP-H) models are proposed in the present study. Based on the assumption that the lateral turbulence component at one single-point is uncorrelated with the longitudinal and vertical turbulence components, the fluctuating wind is separated into 2nV-1D and nV1D nonstationary stochastic vector processes. The first process can be expressed as double proper orthogonal decomposition (DPOD) or proper orthogonal decomposition and spectral representation method (POD-SRM), and the second process can be expressed as POD or SRM. On this basis, four SEP-H models of nonstationary stochastic turbulent wind fields are developed. In addition, the orthogonal random variables in the SEP-H models are presented as random orthogonal functions of elementary random variables. Meanwhile, the number theoretical method (NTM) is conveniently adopted to select representative points set of the elementary random variables. The POD-FFT (Fast Fourier transform) technique is introduced in frequency to give full play to the computational efficiency of the SEP-H models. Finally, taking a long-span bridge as the engineering background, the SEP-H models are compared with the dimension-reduction DPOD (DR-DPOD) model to verify the effectiveness and superiority of the proposed models.