• The Journal of Korean Institute of Information Technology
• /
• v.16 no.12
• /
• pp.75-83
• /
• 2018

Recently, autonomous vehicles have been actively studied. Pedestrian detection and recognition technology is important in autonomous vehicles. Pedestrian detection using CNN(Convolutional Neural Netwrok), which is mainly used recently, generally shows good performance, but there is a performance degradation depending on the environment of the image. In this paper, we propose a pedestrian detection system applying long-term memory structure of hippocampal neural network based on CNN network with LGP-FL (Local Gradient Pattern-Feature Layer) added. First, change the input image to a size of $227{\times}227$. Then, the feature is extracted through a total of 5 layers of convolution layer. In the process, LGP-FL adds the LGP feature pattern and stores the high-frequency pattern in the long-term memory. In the detection process, it is possible to detect the pedestrian more accurately by detecting using the LGP feature pattern information robust to brightness and color change. A comparison of the existing methods and the proposed method confirmed the increase of detection rate of about 1~4%.

• Smart Structures and Systems
• /
• v.18 no.6
• /
• pp.1233-1250
• /
• 2016

Cracks in plate-like structures are some of the main reasons for destruction of the entire structure. In this study, a novel two-stage methodology is proposed for damage detection of flexural plates using an optimized artificial neural network. In the first stage, location of damages in plates is investigated using curvature-moment and curvature-moment derivative concepts. After detecting the damaged areas, the equations for damage severity detection are solved via Bat Algorithm (BA). In the second stage, in order to efficiently reduce the computational cost of model updating during the optimization process of damage severity detection, multiple damage location assurance criterion index based on the frequency change vector of structures are evaluated using properly trained cascade feed-forward neural network (CFNN) as a surrogate model. In order to achieve the most generalized neural network as a surrogate model, its structure is optimized using binary version of BA. To validate this proposed solution method, two examples are presented. The results indicate that after determining the damage location based on curvature-moment derivative concept, the proposed solution method for damage severity detection leads to significant reduction of computational time compared with direct finite element method. Furthermore, integrating BA with the efficient approximation mechanism of finite element model, maintains the acceptable accuracy of damage severity detection.

• Proceedings of the Korean Magnetic Resonance Society Conference
• /
• 2002.08a
• /
• pp.90-90
• /
• 2002

Syndecans, transmembrane heparan sulfate proteoglycans, are coreceptors with integrin in cell adhesion process. It forms a ternary signaling complex with protein kinase C and phosphatidylinositol 4,5 bisphosphate (PIP2) for integrin signaling. NMR data indicates that cytoplasmic domain of syndecan-4 (4L) undergoes a conformational transition in the presence of PIP2, forming oligomeric conformation. The structure based on NMR data demonstrated that syndecan-4L itself forms a compact intertwined symmetric dimer with an unusual clamp shape for residues Leu$^{186}$ -Ala$^{195}$ . The molecular surface of the syndecan-4L dimer is highly positively charged. In addition, no inter-subunit NOEs in membrane proximal amino acid resides (Cl region) has been observed, demonstrating that the Cl region is mostly unstructured in syndecan-4L dimmer. However, the complex structure in the presence of PIP2 induced a high order multimeric conformation in solution. In addition, phosphorylation of cytoplasmic domain induces conformational change of syndecan-4, resulting inhibition of PKC signaling. The NMR structural data strongly suggest that PIP2 promotes oligomerization of syndecan-4 cytoplasmic domain for PKC activation and further induces structural reorganization of syndecan for mediating signaling network in cell adhesion procedure.

• The Journal of Information Technology
• /
• v.8 no.2
• /
• pp.1-16
• /
• 2005

In high education, knowledge is the very important resource and output. Knowledge management is the art of creating additional value from academic knowledge assets. The e-community is the network community, newly created by information communication technology, and the critical place that share and transfer people's the various knowledge and thinking and the effective space that can apply knowledge management. If we construct a convenient and useful e-community in our university, we will be able to increase the student's intellectual needs, and create new knowledge. Web-portal will promote the activity of e-community providing easy to access and personalized web circumstance. The emergence of Internet and Web access to all university services will force institutions to rethik everything. We should reconfigure our academic environment in order to adapt information technology and educational change, and must do change management based on their structure and culture.

• Proceedings of the KIEE Conference
• /
• 2001.07b
• /
• pp.1182-1184
• /
• 2001

This paper presents a neural network based self-tuning fuzzy PID control scheme for induction motor speed control. The PID controller is being widely used in industrial applications. When continuously used long time, the electric and mechanical parameters of induction motor change, degrading the performance of PID controller considerably. This paper re-analyzes the fuzzy controller as conventional PID controller structure, and proposes a neural network based self-tuning fuzzy PID controller whose scaling factors are adjusted automatically. Proposed scheme is simple in structure and computational burden is small. The simulation using Matlab/Simulink is performed to verify the effectiveness of the proposed scheme.

• The Journal of the Acoustical Society of Korea
• /
• v.21 no.3E
• /
• pp.126-131
• /
• 2002

Ultrasonic measurements are made in egg white to study the properties of the solution of the natural protein. The high-Q ultrasonic resonator method is used to get the ultrasonic absorption spectra over the range 0.2-10 ㎒ at 20℃. It is proportional to the 1.25th power of the frequency. The gelation process caused by heat is studied from the change in the velocity and the absorption. at 3 ㎒ using the pulse echo overlap technique over the range of 10-80℃. The absorption decreases with increasing temperature up to 60℃ where it turns up sharply and rapidly increases thereafter. The strong absorption in the gel region is described by the interaction between the solution and the network structure made of protein. Very slow variation in time elapse is observed after the temperature is quickly raised. It would be a real-time observation of the network building process and the characteristic time for the process is shown to be 400 min. A hysteresis phenomenon with respect to the temperature is observed. This phenomenon is associated with the memorizing effect of the network structure of protein of the gel.

• Proceedings of the Korean Society of Machine Tool Engineers Conference
• /
• 2002.10a
• /
• pp.41-46
• /
• 2002

In the face of global competition, the requirements fer the continuously increasing productivity, flexibility and quality (dimensional accuracy, mechanical properties and surface properties) have imposed a major change on steel manufacturing industries. The automation of hot rolling process requires the developments of several mathematical models for simulation and quantitative description of the industrial operations involved. To achieve this objectives, a new loaming method with neural network to improve the accuracy of rolling force prediction in hot rolling mill is developed. Also, Genetic Algorithm(GA) is applied to select the optimal structure of the neural network and compared with that of engineers experience. It is shown from this research that both structure selection methods can lead to similar results.

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.67.1-67
• /
• 2001

This paper presents a neural network based self-tuning fuzzy PI-PD control scheme for robust speed control of induction motor. The PID controller is being widely used in industrial applications. When continuously used long time, the electric and mechanical parameters of induction motor change, degrading the performance of PID controller considerably. This paper re-analyzes the fuzzy controller as conventional PID controller structure, and proposes a neural network based self-tuning fuzzy PI-PD controller whose scaling factors are adjusted automatically. Proposed scheme is simple in structure and computational burden is small ...

• International Journal of Internet, Broadcasting and Communication
• /
• v.15 no.3
• /
• pp.118-121
• /
• 2023

Recently, green energy support policies have been announced around the world in accordance with environmental regulations, and asthe market grows rapidly, demand for batteries is also increasing. Therefore, various methodologies for battery diagnosis and recycling methods are being discussed, but current accurate life prediction of batteries has limitations due to the nonlinear form according to the internal structure or chemical change of the battery. In this paper, CS2 lithium-ion battery measurement data measured at the A. James Clark School of Engineering, University of Marylan was used to predict battery performance with high accuracy using a convolutional neural network (CNN) model among deep learning-based models. As a result, the battery performance was predicted with high accuracy. A data structure with a matrix of total data 3,931 ☓ 19 was designed as test data for the CS2 battery and checking the result values, the MAE was 0.8451, the RMSE was 1.3448, and the accuracy was 0.984, confirming excellent performance.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.8
• /
• pp.124-133
• /
• 1996

This paper presents the implementation of variable structure control system for a linear or nonlinear system using neural networks. The overall control system consists of neural network controller and a reaching mode controller. While the former approximates the equivalent control input on the sliding surface, the latter is used to bring the entire system trajectories toward the sliding surface. No supervised learning procedures are needed and the weights of the neural network are tuned on-line automatically. The neural netowrk-based variable structure control system is applied to a nonlinare unstable inverted pendulum system through computer simulations, and implemented using a microcomputer (80486-50MHz) and applied to the DC servomotor position control system. Simulation and experimental results show the expected approximation sliding property is occurred. The proposed controller is compared with a PID controller and shows better performance than the PID controller in abrupt plant parameter change.